algorithms.statistics.formula.formulae

Module: algorithms.statistics.formula.formulae

Inheritance diagram for nipy.algorithms.statistics.formula.formulae:

Inheritance diagram of nipy.algorithms.statistics.formula.formulae

Formula objects

A formula is basically a sympy expression for the mean of something of the form:

mean = sum([Beta(e)*e for e in expr])

Or, a linear combination of sympy expressions, with each one multiplied by its own “Beta”. The elements of expr can be instances of Term (for a linear regression formula, they would all be instances of Term). But, in general, there might be some other parameters (i.e. sympy.Symbol instances) that are not Terms.

The design matrix is made up of columns that are the derivatives of mean with respect to everything that is not a Term, evaluated at a recarray that has field names given by [str(t) for t in self.terms].

For those familiar with R’s formula syntax, if we wanted a design matrix like the following:

> s.table = read.table("http://www-stat.stanford.edu/~jtaylo/courses/stats191/data/supervisor.table", header=T)
> d = model.matrix(lm(Y ~ X1*X3, s.table)
)
> d
   (Intercept) X1 X3 X1:X3
1            1 51 39  1989
2            1 64 54  3456
3            1 70 69  4830
4            1 63 47  2961
5            1 78 66  5148
6            1 55 44  2420
7            1 67 56  3752
8            1 75 55  4125
9            1 82 67  5494
10           1 61 47  2867
11           1 53 58  3074
12           1 60 39  2340
13           1 62 42  2604
14           1 83 45  3735
15           1 77 72  5544
16           1 90 72  6480
17           1 85 69  5865
18           1 60 75  4500
19           1 70 57  3990
20           1 58 54  3132
21           1 40 34  1360
22           1 61 62  3782
23           1 66 50  3300
24           1 37 58  2146
25           1 54 48  2592
26           1 77 63  4851
27           1 75 74  5550
28           1 57 45  2565
29           1 85 71  6035
30           1 82 59  4838
attr(,"assign")
[1] 0 1 2 3
>

With the Formula, it looks like this:

>>> r = np.rec.array([
...     (43, 51, 30, 39, 61, 92, 45), (63, 64, 51, 54, 63, 73, 47),
...     (71, 70, 68, 69, 76, 86, 48), (61, 63, 45, 47, 54, 84, 35),
...     (81, 78, 56, 66, 71, 83, 47), (43, 55, 49, 44, 54, 49, 34),
...     (58, 67, 42, 56, 66, 68, 35), (71, 75, 50, 55, 70, 66, 41),
...     (72, 82, 72, 67, 71, 83, 31), (67, 61, 45, 47, 62, 80, 41),
...     (64, 53, 53, 58, 58, 67, 34), (67, 60, 47, 39, 59, 74, 41),
...     (69, 62, 57, 42, 55, 63, 25), (68, 83, 83, 45, 59, 77, 35),
...     (77, 77, 54, 72, 79, 77, 46), (81, 90, 50, 72, 60, 54, 36),
...     (74, 85, 64, 69, 79, 79, 63), (65, 60, 65, 75, 55, 80, 60),
...     (65, 70, 46, 57, 75, 85, 46), (50, 58, 68, 54, 64, 78, 52),
...     (50, 40, 33, 34, 43, 64, 33), (64, 61, 52, 62, 66, 80, 41),
...     (53, 66, 52, 50, 63, 80, 37), (40, 37, 42, 58, 50, 57, 49),
...     (63, 54, 42, 48, 66, 75, 33), (66, 77, 66, 63, 88, 76, 72),
...     (78, 75, 58, 74, 80, 78, 49), (48, 57, 44, 45, 51, 83, 38),
...     (85, 85, 71, 71, 77, 74, 55), (82, 82, 39, 59, 64, 78, 39)],
...              dtype=[('y', '<i8'), ('x1', '<i8'), ('x2', '<i8'),
...                     ('x3', '<i8'), ('x4', '<i8'), ('x5', '<i8'),
...                     ('x6', '<i8')])
>>> x1 = Term('x1'); x3 = Term('x3')
>>> f = Formula([x1, x3, x1*x3]) + I
>>> f.mean
_b0*x1 + _b1*x3 + _b2*x1*x3 + _b3

The I is the “intercept” term, I have explicitly not used R’s default of adding it to everything.

>>> f.design(r)  
array([(51.0, 39.0, 1989.0, 1.0), (64.0, 54.0, 3456.0, 1.0),
       (70.0, 69.0, 4830.0, 1.0), (63.0, 47.0, 2961.0, 1.0),
       (78.0, 66.0, 5148.0, 1.0), (55.0, 44.0, 2420.0, 1.0),
       (67.0, 56.0, 3752.0, 1.0), (75.0, 55.0, 4125.0, 1.0),
       (82.0, 67.0, 5494.0, 1.0), (61.0, 47.0, 2867.0, 1.0),
       (53.0, 58.0, 3074.0, 1.0), (60.0, 39.0, 2340.0, 1.0),
       (62.0, 42.0, 2604.0, 1.0), (83.0, 45.0, 3735.0, 1.0),
       (77.0, 72.0, 5544.0, 1.0), (90.0, 72.0, 6480.0, 1.0),
       (85.0, 69.0, 5865.0, 1.0), (60.0, 75.0, 4500.0, 1.0),
       (70.0, 57.0, 3990.0, 1.0), (58.0, 54.0, 3132.0, 1.0),
       (40.0, 34.0, 1360.0, 1.0), (61.0, 62.0, 3782.0, 1.0),
       (66.0, 50.0, 3300.0, 1.0), (37.0, 58.0, 2146.0, 1.0),
       (54.0, 48.0, 2592.0, 1.0), (77.0, 63.0, 4851.0, 1.0),
       (75.0, 74.0, 5550.0, 1.0), (57.0, 45.0, 2565.0, 1.0),
       (85.0, 71.0, 6035.0, 1.0), (82.0, 59.0, 4838.0, 1.0)],
      dtype=[('x1', '<f8'), ('x3', '<f8'), ('x1*x3', '<f8'), ('1', '<f8')])

Classes

Beta

class nipy.algorithms.statistics.formula.formulae.Beta(name, term)

Bases: Dummy

A symbol tied to a Term term

__init__(*args, **kwargs)
adjoint()
apart(x=None, **args)

See the apart function in sympy.polys

property args: tuple[Basic, ...]

Returns a tuple of arguments of ‘self’.

Notes

Never use self._args, always use self.args. Only use _args in __new__ when creating a new function. Do not override .args() from Basic (so that it is easy to change the interface in the future if needed).

Examples

>>> from sympy import cot
>>> from sympy.abc import x, y
>>> cot(x).args
(x,)
>>> cot(x).args[0]
x
>>> (x*y).args
(x, y)
>>> (x*y).args[1]
y
args_cnc(cset=False, warn=True, split_1=True)

Return [commutative factors, non-commutative factors] of self.

Examples

>>> from sympy import symbols, oo
>>> A, B = symbols('A B', commutative=0)
>>> x, y = symbols('x y')
>>> (-2*x*y).args_cnc()
[[-1, 2, x, y], []]
>>> (-2.5*x).args_cnc()
[[-1, 2.5, x], []]
>>> (-2*x*A*B*y).args_cnc()
[[-1, 2, x, y], [A, B]]
>>> (-2*x*A*B*y).args_cnc(split_1=False)
[[-2, x, y], [A, B]]
>>> (-2*x*y).args_cnc(cset=True)
[{-1, 2, x, y}, []]

The arg is always treated as a Mul:

>>> (-2 + x + A).args_cnc()
[[], [x - 2 + A]]
>>> (-oo).args_cnc() # -oo is a singleton
[[-1, oo], []]
as_base_exp() tuple[Expr, Expr]
as_coeff_Add(rational=False) tuple[Number, Expr]

Efficiently extract the coefficient of a summation.

as_coeff_Mul(rational: bool = False) tuple[Number, Expr]

Efficiently extract the coefficient of a product.

as_coeff_add(*deps) tuple[Expr, tuple[Expr, ...]]

Return the tuple (c, args) where self is written as an Add, a.

c should be a Rational added to any terms of the Add that are independent of deps.

args should be a tuple of all other terms of a; args is empty if self is a Number or if self is independent of deps (when given).

This should be used when you do not know if self is an Add or not but you want to treat self as an Add or if you want to process the individual arguments of the tail of self as an Add.

  • if you know self is an Add and want only the head, use self.args[0];

  • if you do not want to process the arguments of the tail but need the tail then use self.as_two_terms() which gives the head and tail.

  • if you want to split self into an independent and dependent parts use self.as_independent(*deps)

>>> from sympy import S
>>> from sympy.abc import x, y
>>> (S(3)).as_coeff_add()
(3, ())
>>> (3 + x).as_coeff_add()
(3, (x,))
>>> (3 + x + y).as_coeff_add(x)
(y + 3, (x,))
>>> (3 + y).as_coeff_add(x)
(y + 3, ())
as_coeff_exponent(x) tuple[Expr, Expr]

c*x**e -> c,e where x can be any symbolic expression.

as_coeff_mul(*deps, **kwargs) tuple[Expr, tuple[Expr, ...]]

Return the tuple (c, args) where self is written as a Mul, m.

c should be a Rational multiplied by any factors of the Mul that are independent of deps.

args should be a tuple of all other factors of m; args is empty if self is a Number or if self is independent of deps (when given).

This should be used when you do not know if self is a Mul or not but you want to treat self as a Mul or if you want to process the individual arguments of the tail of self as a Mul.

  • if you know self is a Mul and want only the head, use self.args[0];

  • if you do not want to process the arguments of the tail but need the tail then use self.as_two_terms() which gives the head and tail;

  • if you want to split self into an independent and dependent parts use self.as_independent(*deps)

>>> from sympy import S
>>> from sympy.abc import x, y
>>> (S(3)).as_coeff_mul()
(3, ())
>>> (3*x*y).as_coeff_mul()
(3, (x, y))
>>> (3*x*y).as_coeff_mul(x)
(3*y, (x,))
>>> (3*y).as_coeff_mul(x)
(3*y, ())
as_coefficient(expr)

Extracts symbolic coefficient at the given expression. In other words, this functions separates ‘self’ into the product of ‘expr’ and ‘expr’-free coefficient. If such separation is not possible it will return None.

See also

coeff

return sum of terms have a given factor

as_coeff_Add

separate the additive constant from an expression

as_coeff_Mul

separate the multiplicative constant from an expression

as_independent

separate x-dependent terms/factors from others

sympy.polys.polytools.Poly.coeff_monomial

efficiently find the single coefficient of a monomial in Poly

sympy.polys.polytools.Poly.nth

like coeff_monomial but powers of monomial terms are used

Examples

>>> from sympy import E, pi, sin, I, Poly
>>> from sympy.abc import x
>>> E.as_coefficient(E)
1
>>> (2*E).as_coefficient(E)
2
>>> (2*sin(E)*E).as_coefficient(E)

Two terms have E in them so a sum is returned. (If one were desiring the coefficient of the term exactly matching E then the constant from the returned expression could be selected. Or, for greater precision, a method of Poly can be used to indicate the desired term from which the coefficient is desired.)

>>> (2*E + x*E).as_coefficient(E)
x + 2
>>> _.args[0]  # just want the exact match
2
>>> p = Poly(2*E + x*E); p
Poly(x*E + 2*E, x, E, domain='ZZ')
>>> p.coeff_monomial(E)
2
>>> p.nth(0, 1)
2

Since the following cannot be written as a product containing E as a factor, None is returned. (If the coefficient 2*x is desired then the coeff method should be used.)

>>> (2*E*x + x).as_coefficient(E)
>>> (2*E*x + x).coeff(E)
2*x
>>> (E*(x + 1) + x).as_coefficient(E)
>>> (2*pi*I).as_coefficient(pi*I)
2
>>> (2*I).as_coefficient(pi*I)
as_coefficients_dict(*syms)

Return a dictionary mapping terms to their Rational coefficient. Since the dictionary is a defaultdict, inquiries about terms which were not present will return a coefficient of 0.

If symbols syms are provided, any multiplicative terms independent of them will be considered a coefficient and a regular dictionary of syms-dependent generators as keys and their corresponding coefficients as values will be returned.

Examples

>>> from sympy.abc import a, x, y
>>> (3*x + a*x + 4).as_coefficients_dict()
{1: 4, x: 3, a*x: 1}
>>> _[a]
0
>>> (3*a*x).as_coefficients_dict()
{a*x: 3}
>>> (3*a*x).as_coefficients_dict(x)
{x: 3*a}
>>> (3*a*x).as_coefficients_dict(y)
{1: 3*a*x}
as_content_primitive(radical=False, clear=True)

This method should recursively remove a Rational from all arguments and return that (content) and the new self (primitive). The content should always be positive and Mul(*foo.as_content_primitive()) == foo. The primitive need not be in canonical form and should try to preserve the underlying structure if possible (i.e. expand_mul should not be applied to self).

Examples

>>> from sympy import sqrt
>>> from sympy.abc import x, y, z
>>> eq = 2 + 2*x + 2*y*(3 + 3*y)

The as_content_primitive function is recursive and retains structure:

>>> eq.as_content_primitive()
(2, x + 3*y*(y + 1) + 1)

Integer powers will have Rationals extracted from the base:

>>> ((2 + 6*x)**2).as_content_primitive()
(4, (3*x + 1)**2)
>>> ((2 + 6*x)**(2*y)).as_content_primitive()
(1, (2*(3*x + 1))**(2*y))

Terms may end up joining once their as_content_primitives are added:

>>> ((5*(x*(1 + y)) + 2*x*(3 + 3*y))).as_content_primitive()
(11, x*(y + 1))
>>> ((3*(x*(1 + y)) + 2*x*(3 + 3*y))).as_content_primitive()
(9, x*(y + 1))
>>> ((3*(z*(1 + y)) + 2.0*x*(3 + 3*y))).as_content_primitive()
(1, 6.0*x*(y + 1) + 3*z*(y + 1))
>>> ((5*(x*(1 + y)) + 2*x*(3 + 3*y))**2).as_content_primitive()
(121, x**2*(y + 1)**2)
>>> ((x*(1 + y) + 0.4*x*(3 + 3*y))**2).as_content_primitive()
(1, 4.84*x**2*(y + 1)**2)

Radical content can also be factored out of the primitive:

>>> (2*sqrt(2) + 4*sqrt(10)).as_content_primitive(radical=True)
(2, sqrt(2)*(1 + 2*sqrt(5)))

If clear=False (default is True) then content will not be removed from an Add if it can be distributed to leave one or more terms with integer coefficients.

>>> (x/2 + y).as_content_primitive()
(1/2, x + 2*y)
>>> (x/2 + y).as_content_primitive(clear=False)
(1, x/2 + y)
as_dummy()

Return the expression with any objects having structurally bound symbols replaced with unique, canonical symbols within the object in which they appear and having only the default assumption for commutativity being True. When applied to a symbol a new symbol having only the same commutativity will be returned.

Notes

Any object that has structurally bound variables should have a property, bound_symbols that returns those symbols appearing in the object.

Examples

>>> from sympy import Integral, Symbol
>>> from sympy.abc import x
>>> r = Symbol('r', real=True)
>>> Integral(r, (r, x)).as_dummy()
Integral(_0, (_0, x))
>>> _.variables[0].is_real is None
True
>>> r.as_dummy()
_r
as_expr(*gens)

Convert a polynomial to a SymPy expression.

Examples

>>> from sympy import sin
>>> from sympy.abc import x, y
>>> f = (x**2 + x*y).as_poly(x, y)
>>> f.as_expr()
x**2 + x*y
>>> sin(x).as_expr()
sin(x)
as_independent(*deps, **hint) tuple[Expr, Expr]

A mostly naive separation of a Mul or Add into arguments that are not are dependent on deps. To obtain as complete a separation of variables as possible, use a separation method first, e.g.:

  • separatevars() to change Mul, Add and Pow (including exp) into Mul

  • .expand(mul=True) to change Add or Mul into Add

  • .expand(log=True) to change log expr into an Add

The only non-naive thing that is done here is to respect noncommutative ordering of variables and to always return (0, 0) for self of zero regardless of hints.

For nonzero self, the returned tuple (i, d) has the following interpretation:

  • i will has no variable that appears in deps

  • d will either have terms that contain variables that are in deps, or be equal to 0 (when self is an Add) or 1 (when self is a Mul)

  • if self is an Add then self = i + d

  • if self is a Mul then self = i*d

  • otherwise (self, S.One) or (S.One, self) is returned.

To force the expression to be treated as an Add, use the hint as_Add=True

See also

separatevars
expand_log
sympy.core.add.Add.as_two_terms
sympy.core.mul.Mul.as_two_terms
as_coeff_mul

Examples

– self is an Add

>>> from sympy import sin, cos, exp
>>> from sympy.abc import x, y, z
>>> (x + x*y).as_independent(x)
(0, x*y + x)
>>> (x + x*y).as_independent(y)
(x, x*y)
>>> (2*x*sin(x) + y + x + z).as_independent(x)
(y + z, 2*x*sin(x) + x)
>>> (2*x*sin(x) + y + x + z).as_independent(x, y)
(z, 2*x*sin(x) + x + y)

– self is a Mul

>>> (x*sin(x)*cos(y)).as_independent(x)
(cos(y), x*sin(x))

non-commutative terms cannot always be separated out when self is a Mul

>>> from sympy import symbols
>>> n1, n2, n3 = symbols('n1 n2 n3', commutative=False)
>>> (n1 + n1*n2).as_independent(n2)
(n1, n1*n2)
>>> (n2*n1 + n1*n2).as_independent(n2)
(0, n1*n2 + n2*n1)
>>> (n1*n2*n3).as_independent(n1)
(1, n1*n2*n3)
>>> (n1*n2*n3).as_independent(n2)
(n1, n2*n3)
>>> ((x-n1)*(x-y)).as_independent(x)
(1, (x - y)*(x - n1))

– self is anything else:

>>> (sin(x)).as_independent(x)
(1, sin(x))
>>> (sin(x)).as_independent(y)
(sin(x), 1)
>>> exp(x+y).as_independent(x)
(1, exp(x + y))

– force self to be treated as an Add:

>>> (3*x).as_independent(x, as_Add=True)
(0, 3*x)

– force self to be treated as a Mul:

>>> (3+x).as_independent(x, as_Add=False)
(1, x + 3)
>>> (-3+x).as_independent(x, as_Add=False)
(1, x - 3)

Note how the below differs from the above in making the constant on the dep term positive.

>>> (y*(-3+x)).as_independent(x)
(y, x - 3)
– use .as_independent() for true independence testing instead

of .has(). The former considers only symbols in the free symbols while the latter considers all symbols

>>> from sympy import Integral
>>> I = Integral(x, (x, 1, 2))
>>> I.has(x)
True
>>> x in I.free_symbols
False
>>> I.as_independent(x) == (I, 1)
True
>>> (I + x).as_independent(x) == (I, x)
True

Note: when trying to get independent terms, a separation method might need to be used first. In this case, it is important to keep track of what you send to this routine so you know how to interpret the returned values

>>> from sympy import separatevars, log
>>> separatevars(exp(x+y)).as_independent(x)
(exp(y), exp(x))
>>> (x + x*y).as_independent(y)
(x, x*y)
>>> separatevars(x + x*y).as_independent(y)
(x, y + 1)
>>> (x*(1 + y)).as_independent(y)
(x, y + 1)
>>> (x*(1 + y)).expand(mul=True).as_independent(y)
(x, x*y)
>>> a, b=symbols('a b', positive=True)
>>> (log(a*b).expand(log=True)).as_independent(b)
(log(a), log(b))
as_leading_term(*symbols, logx=None, cdir=0)

Returns the leading (nonzero) term of the series expansion of self.

The _eval_as_leading_term routines are used to do this, and they must always return a non-zero value.

Examples

>>> from sympy.abc import x
>>> (1 + x + x**2).as_leading_term(x)
1
>>> (1/x**2 + x + x**2).as_leading_term(x)
x**(-2)
as_numer_denom()

Return the numerator and the denominator of an expression.

expression -> a/b -> a, b

This is just a stub that should be defined by an object’s class methods to get anything else.

See also

normal

return a/b instead of (a, b)

as_ordered_factors(order=None)

Return list of ordered factors (if Mul) else [self].

as_ordered_terms(order=None, data=False)

Transform an expression to an ordered list of terms.

Examples

>>> from sympy import sin, cos
>>> from sympy.abc import x
>>> (sin(x)**2*cos(x) + sin(x)**2 + 1).as_ordered_terms()
[sin(x)**2*cos(x), sin(x)**2, 1]
as_poly(*gens, **args)

Converts self to a polynomial or returns None.

as_powers_dict()

Return self as a dictionary of factors with each factor being treated as a power. The keys are the bases of the factors and the values, the corresponding exponents. The resulting dictionary should be used with caution if the expression is a Mul and contains non- commutative factors since the order that they appeared will be lost in the dictionary.

See also

as_ordered_factors

An alternative for noncommutative applications, returning an ordered list of factors.

args_cnc

Similar to as_ordered_factors, but guarantees separation of commutative and noncommutative factors.

as_real_imag(deep=True, **hints)

Performs complex expansion on ‘self’ and returns a tuple containing collected both real and imaginary parts. This method cannot be confused with re() and im() functions, which does not perform complex expansion at evaluation.

However it is possible to expand both re() and im() functions and get exactly the same results as with a single call to this function.

>>> from sympy import symbols, I
>>> x, y = symbols('x,y', real=True)
>>> (x + y*I).as_real_imag()
(x, y)
>>> from sympy.abc import z, w
>>> (z + w*I).as_real_imag()
(re(z) - im(w), re(w) + im(z))
as_set()

Rewrites Boolean expression in terms of real sets.

Examples

>>> from sympy import Symbol, Eq, Or, And
>>> x = Symbol('x', real=True)
>>> Eq(x, 0).as_set()
{0}
>>> (x > 0).as_set()
Interval.open(0, oo)
>>> And(-2 < x, x < 2).as_set()
Interval.open(-2, 2)
>>> Or(x < -2, 2 < x).as_set()
Union(Interval.open(-oo, -2), Interval.open(2, oo))
as_terms()

Transform an expression to a list of terms.

aseries(x=None, n=6, bound=0, hir=False)

Asymptotic Series expansion of self. This is equivalent to self.series(x, oo, n).

Parameters:
selfExpression

The expression whose series is to be expanded.

xSymbol

It is the variable of the expression to be calculated.

nValue

The value used to represent the order in terms of x**n, up to which the series is to be expanded.

hirBoolean

Set this parameter to be True to produce hierarchical series. It stops the recursion at an early level and may provide nicer and more useful results.

boundValue, Integer

Use the bound parameter to give limit on rewriting coefficients in its normalised form.

Returns:
Expr

Asymptotic series expansion of the expression.

See also

Expr.aseries

See the docstring of this function for complete details of this wrapper.

Notes

This algorithm is directly induced from the limit computational algorithm provided by Gruntz. It majorly uses the mrv and rewrite sub-routines. The overall idea of this algorithm is first to look for the most rapidly varying subexpression w of a given expression f and then expands f in a series in w. Then same thing is recursively done on the leading coefficient till we get constant coefficients.

If the most rapidly varying subexpression of a given expression f is f itself, the algorithm tries to find a normalised representation of the mrv set and rewrites f using this normalised representation.

If the expansion contains an order term, it will be either O(x ** (-n)) or O(w ** (-n)) where w belongs to the most rapidly varying expression of self.

References

[1]

Gruntz, Dominik. A new algorithm for computing asymptotic series. In: Proc. 1993 Int. Symp. Symbolic and Algebraic Computation. 1993. pp. 239-244.

[2]

Gruntz thesis - p90

Examples

>>> from sympy import sin, exp
>>> from sympy.abc import x
>>> e = sin(1/x + exp(-x)) - sin(1/x)
>>> e.aseries(x)
(1/(24*x**4) - 1/(2*x**2) + 1 + O(x**(-6), (x, oo)))*exp(-x)
>>> e.aseries(x, n=3, hir=True)
-exp(-2*x)*sin(1/x)/2 + exp(-x)*cos(1/x) + O(exp(-3*x), (x, oo))
>>> e = exp(exp(x)/(1 - 1/x))
>>> e.aseries(x)
exp(exp(x)/(1 - 1/x))
>>> e.aseries(x, bound=3) 
exp(exp(x)/x**2)*exp(exp(x)/x)*exp(-exp(x) + exp(x)/(1 - 1/x) - exp(x)/x - exp(x)/x**2)*exp(exp(x))

For rational expressions this method may return original expression without the Order term. >>> (1/x).aseries(x, n=8) 1/x

property assumptions0

Return object type assumptions.

For example:

Symbol(‘x’, real=True) Symbol(‘x’, integer=True)

are different objects. In other words, besides Python type (Symbol in this case), the initial assumptions are also forming their typeinfo.

Examples

>>> from sympy import Symbol
>>> from sympy.abc import x
>>> x.assumptions0
{'commutative': True}
>>> x = Symbol("x", positive=True)
>>> x.assumptions0
{'commutative': True, 'complex': True, 'extended_negative': False,
 'extended_nonnegative': True, 'extended_nonpositive': False,
 'extended_nonzero': True, 'extended_positive': True, 'extended_real':
 True, 'finite': True, 'hermitian': True, 'imaginary': False,
 'infinite': False, 'negative': False, 'nonnegative': True,
 'nonpositive': False, 'nonzero': True, 'positive': True, 'real':
 True, 'zero': False}
atoms(*types)

Returns the atoms that form the current object.

By default, only objects that are truly atomic and cannot be divided into smaller pieces are returned: symbols, numbers, and number symbols like I and pi. It is possible to request atoms of any type, however, as demonstrated below.

Examples

>>> from sympy import I, pi, sin
>>> from sympy.abc import x, y
>>> (1 + x + 2*sin(y + I*pi)).atoms()
{1, 2, I, pi, x, y}

If one or more types are given, the results will contain only those types of atoms.

>>> from sympy import Number, NumberSymbol, Symbol
>>> (1 + x + 2*sin(y + I*pi)).atoms(Symbol)
{x, y}
>>> (1 + x + 2*sin(y + I*pi)).atoms(Number)
{1, 2}
>>> (1 + x + 2*sin(y + I*pi)).atoms(Number, NumberSymbol)
{1, 2, pi}
>>> (1 + x + 2*sin(y + I*pi)).atoms(Number, NumberSymbol, I)
{1, 2, I, pi}

Note that I (imaginary unit) and zoo (complex infinity) are special types of number symbols and are not part of the NumberSymbol class.

The type can be given implicitly, too:

>>> (1 + x + 2*sin(y + I*pi)).atoms(x) # x is a Symbol
{x, y}

Be careful to check your assumptions when using the implicit option since S(1).is_Integer = True but type(S(1)) is One, a special type of SymPy atom, while type(S(2)) is type Integer and will find all integers in an expression:

>>> from sympy import S
>>> (1 + x + 2*sin(y + I*pi)).atoms(S(1))
{1}
>>> (1 + x + 2*sin(y + I*pi)).atoms(S(2))
{1, 2}

Finally, arguments to atoms() can select more than atomic atoms: any SymPy type (loaded in core/__init__.py) can be listed as an argument and those types of “atoms” as found in scanning the arguments of the expression recursively:

>>> from sympy import Function, Mul
>>> from sympy.core.function import AppliedUndef
>>> f = Function('f')
>>> (1 + f(x) + 2*sin(y + I*pi)).atoms(Function)
{f(x), sin(y + I*pi)}
>>> (1 + f(x) + 2*sin(y + I*pi)).atoms(AppliedUndef)
{f(x)}
>>> (1 + x + 2*sin(y + I*pi)).atoms(Mul)
{I*pi, 2*sin(y + I*pi)}
property binary_symbols

Return from the atoms of self those which are free symbols.

Not all free symbols are Symbol. Eg: IndexedBase(‘I’)[0].free_symbols

For most expressions, all symbols are free symbols. For some classes this is not true. e.g. Integrals use Symbols for the dummy variables which are bound variables, so Integral has a method to return all symbols except those. Derivative keeps track of symbols with respect to which it will perform a derivative; those are bound variables, too, so it has its own free_symbols method.

Any other method that uses bound variables should implement a free_symbols method.

cancel(*gens, **args)

See the cancel function in sympy.polys

property canonical_variables

Return a dictionary mapping any variable defined in self.bound_symbols to Symbols that do not clash with any free symbols in the expression.

Examples

>>> from sympy import Lambda
>>> from sympy.abc import x
>>> Lambda(x, 2*x).canonical_variables
{x: _0}
classmethod class_key()

Nice order of classes.

coeff(x, n=1, right=False, _first=True)

Returns the coefficient from the term(s) containing x**n. If n is zero then all terms independent of x will be returned.

See also

as_coefficient

separate the expression into a coefficient and factor

as_coeff_Add

separate the additive constant from an expression

as_coeff_Mul

separate the multiplicative constant from an expression

as_independent

separate x-dependent terms/factors from others

sympy.polys.polytools.Poly.coeff_monomial

efficiently find the single coefficient of a monomial in Poly

sympy.polys.polytools.Poly.nth

like coeff_monomial but powers of monomial terms are used

Examples

>>> from sympy import symbols
>>> from sympy.abc import x, y, z

You can select terms that have an explicit negative in front of them:

>>> (-x + 2*y).coeff(-1)
x
>>> (x - 2*y).coeff(-1)
2*y

You can select terms with no Rational coefficient:

>>> (x + 2*y).coeff(1)
x
>>> (3 + 2*x + 4*x**2).coeff(1)
0

You can select terms independent of x by making n=0; in this case expr.as_independent(x)[0] is returned (and 0 will be returned instead of None):

>>> (3 + 2*x + 4*x**2).coeff(x, 0)
3
>>> eq = ((x + 1)**3).expand() + 1
>>> eq
x**3 + 3*x**2 + 3*x + 2
>>> [eq.coeff(x, i) for i in reversed(range(4))]
[1, 3, 3, 2]
>>> eq -= 2
>>> [eq.coeff(x, i) for i in reversed(range(4))]
[1, 3, 3, 0]

You can select terms that have a numerical term in front of them:

>>> (-x - 2*y).coeff(2)
-y
>>> from sympy import sqrt
>>> (x + sqrt(2)*x).coeff(sqrt(2))
x

The matching is exact:

>>> (3 + 2*x + 4*x**2).coeff(x)
2
>>> (3 + 2*x + 4*x**2).coeff(x**2)
4
>>> (3 + 2*x + 4*x**2).coeff(x**3)
0
>>> (z*(x + y)**2).coeff((x + y)**2)
z
>>> (z*(x + y)**2).coeff(x + y)
0

In addition, no factoring is done, so 1 + z*(1 + y) is not obtained from the following:

>>> (x + z*(x + x*y)).coeff(x)
1

If such factoring is desired, factor_terms can be used first:

>>> from sympy import factor_terms
>>> factor_terms(x + z*(x + x*y)).coeff(x)
z*(y + 1) + 1
>>> n, m, o = symbols('n m o', commutative=False)
>>> n.coeff(n)
1
>>> (3*n).coeff(n)
3
>>> (n*m + m*n*m).coeff(n) # = (1 + m)*n*m
1 + m
>>> (n*m + m*n*m).coeff(n, right=True) # = (1 + m)*n*m
m

If there is more than one possible coefficient 0 is returned:

>>> (n*m + m*n).coeff(n)
0

If there is only one possible coefficient, it is returned:

>>> (n*m + x*m*n).coeff(m*n)
x
>>> (n*m + x*m*n).coeff(m*n, right=1)
1
collect(syms, func=None, evaluate=True, exact=False, distribute_order_term=True)

See the collect function in sympy.simplify

combsimp()

See the combsimp function in sympy.simplify

compare(other)

Return -1, 0, 1 if the object is smaller, equal, or greater than other.

Not in the mathematical sense. If the object is of a different type from the “other” then their classes are ordered according to the sorted_classes list.

Examples

>>> from sympy.abc import x, y
>>> x.compare(y)
-1
>>> x.compare(x)
0
>>> y.compare(x)
1
compute_leading_term(x, logx=None)

Deprecated function to compute the leading term of a series.

as_leading_term is only allowed for results of .series() This is a wrapper to compute a series first.

conjugate()

Returns the complex conjugate of ‘self’.

copy()
could_extract_minus_sign()

Return True if self has -1 as a leading factor or has more literal negative signs than positive signs in a sum, otherwise False.

Examples

>>> from sympy.abc import x, y
>>> e = x - y
>>> {i.could_extract_minus_sign() for i in (e, -e)}
{False, True}

Though the y - x is considered like -(x - y), since it is in a product without a leading factor of -1, the result is false below:

>>> (x*(y - x)).could_extract_minus_sign()
False

To put something in canonical form wrt to sign, use signsimp:

>>> from sympy import signsimp
>>> signsimp(x*(y - x))
-x*(x - y)
>>> _.could_extract_minus_sign()
True
count(query)

Count the number of matching subexpressions.

count_ops(visual=None)

Wrapper for count_ops that returns the operation count.

default_assumptions = {}
diff(*symbols, **assumptions)
dir(x, cdir)
doit(**hints)

Evaluate objects that are not evaluated by default like limits, integrals, sums and products. All objects of this kind will be evaluated recursively, unless some species were excluded via ‘hints’ or unless the ‘deep’ hint was set to ‘False’.

>>> from sympy import Integral
>>> from sympy.abc import x
>>> 2*Integral(x, x)
2*Integral(x, x)
>>> (2*Integral(x, x)).doit()
x**2
>>> (2*Integral(x, x)).doit(deep=False)
2*Integral(x, x)
dummy_eq(other, symbol=None)

Compare two expressions and handle dummy symbols.

Examples

>>> from sympy import Dummy
>>> from sympy.abc import x, y
>>> u = Dummy('u')
>>> (u**2 + 1).dummy_eq(x**2 + 1)
True
>>> (u**2 + 1) == (x**2 + 1)
False
>>> (u**2 + y).dummy_eq(x**2 + y, x)
True
>>> (u**2 + y).dummy_eq(x**2 + y, y)
False
dummy_index
equals(other, failing_expression=False)

Return True if self == other, False if it does not, or None. If failing_expression is True then the expression which did not simplify to a 0 will be returned instead of None.

evalf(n=15, subs=None, maxn=100, chop=False, strict=False, quad=None, verbose=False)

Evaluate the given formula to an accuracy of n digits.

Parameters:
subsdict, optional

Substitute numerical values for symbols, e.g. subs={x:3, y:1+pi}. The substitutions must be given as a dictionary.

maxnint, optional

Allow a maximum temporary working precision of maxn digits.

chopbool or number, optional

Specifies how to replace tiny real or imaginary parts in subresults by exact zeros.

When True the chop value defaults to standard precision.

Otherwise the chop value is used to determine the magnitude of “small” for purposes of chopping.

>>> from sympy import N
>>> x = 1e-4
>>> N(x, chop=True)
0.000100000000000000
>>> N(x, chop=1e-5)
0.000100000000000000
>>> N(x, chop=1e-4)
0
strictbool, optional

Raise PrecisionExhausted if any subresult fails to evaluate to full accuracy, given the available maxprec.

quadstr, optional

Choose algorithm for numerical quadrature. By default, tanh-sinh quadrature is used. For oscillatory integrals on an infinite interval, try quad='osc'.

verbosebool, optional

Print debug information.

Notes

When Floats are naively substituted into an expression, precision errors may adversely affect the result. For example, adding 1e16 (a Float) to 1 will truncate to 1e16; if 1e16 is then subtracted, the result will be 0. That is exactly what happens in the following:

>>> from sympy.abc import x, y, z
>>> values = {x: 1e16, y: 1, z: 1e16}
>>> (x + y - z).subs(values)
0

Using the subs argument for evalf is the accurate way to evaluate such an expression:

>>> (x + y - z).evalf(subs=values)
1.00000000000000
expand(deep=True, modulus=None, power_base=True, power_exp=True, mul=True, log=True, multinomial=True, basic=True, **hints)

Expand an expression using hints.

See the docstring of the expand() function in sympy.core.function for more information.

property expr_free_symbols

Like free_symbols, but returns the free symbols only if they are contained in an expression node.

Examples

>>> from sympy.abc import x, y
>>> (x + y).expr_free_symbols 
{x, y}

If the expression is contained in a non-expression object, do not return the free symbols. Compare:

>>> from sympy import Tuple
>>> t = Tuple(x + y)
>>> t.expr_free_symbols 
set()
>>> t.free_symbols
{x, y}
extract_additively(c)

Return self - c if it’s possible to subtract c from self and make all matching coefficients move towards zero, else return None.

Examples

>>> from sympy.abc import x, y
>>> e = 2*x + 3
>>> e.extract_additively(x + 1)
x + 2
>>> e.extract_additively(3*x)
>>> e.extract_additively(4)
>>> (y*(x + 1)).extract_additively(x + 1)
>>> ((x + 1)*(x + 2*y + 1) + 3).extract_additively(x + 1)
(x + 1)*(x + 2*y) + 3
extract_branch_factor(allow_half=False)

Try to write self as exp_polar(2*pi*I*n)*z in a nice way. Return (z, n).

>>> from sympy import exp_polar, I, pi
>>> from sympy.abc import x, y
>>> exp_polar(I*pi).extract_branch_factor()
(exp_polar(I*pi), 0)
>>> exp_polar(2*I*pi).extract_branch_factor()
(1, 1)
>>> exp_polar(-pi*I).extract_branch_factor()
(exp_polar(I*pi), -1)
>>> exp_polar(3*pi*I + x).extract_branch_factor()
(exp_polar(x + I*pi), 1)
>>> (y*exp_polar(-5*pi*I)*exp_polar(3*pi*I + 2*pi*x)).extract_branch_factor()
(y*exp_polar(2*pi*x), -1)
>>> exp_polar(-I*pi/2).extract_branch_factor()
(exp_polar(-I*pi/2), 0)

If allow_half is True, also extract exp_polar(I*pi):

>>> exp_polar(I*pi).extract_branch_factor(allow_half=True)
(1, 1/2)
>>> exp_polar(2*I*pi).extract_branch_factor(allow_half=True)
(1, 1)
>>> exp_polar(3*I*pi).extract_branch_factor(allow_half=True)
(1, 3/2)
>>> exp_polar(-I*pi).extract_branch_factor(allow_half=True)
(1, -1/2)
extract_multiplicatively(c)

Return None if it’s not possible to make self in the form c * something in a nice way, i.e. preserving the properties of arguments of self.

Examples

>>> from sympy import symbols, Rational
>>> x, y = symbols('x,y', real=True)
>>> ((x*y)**3).extract_multiplicatively(x**2 * y)
x*y**2
>>> ((x*y)**3).extract_multiplicatively(x**4 * y)
>>> (2*x).extract_multiplicatively(2)
x
>>> (2*x).extract_multiplicatively(3)
>>> (Rational(1, 2)*x).extract_multiplicatively(3)
x/6
factor(*gens, **args)

See the factor() function in sympy.polys.polytools

find(query, group=False)

Find all subexpressions matching a query.

fourier_series(limits=None)

Compute fourier sine/cosine series of self.

See the docstring of the fourier_series() in sympy.series.fourier for more information.

fps(x=None, x0=0, dir=1, hyper=True, order=4, rational=True, full=False)

Compute formal power power series of self.

See the docstring of the fps() function in sympy.series.formal for more information.

property free_symbols

Return from the atoms of self those which are free symbols.

Not all free symbols are Symbol. Eg: IndexedBase(‘I’)[0].free_symbols

For most expressions, all symbols are free symbols. For some classes this is not true. e.g. Integrals use Symbols for the dummy variables which are bound variables, so Integral has a method to return all symbols except those. Derivative keeps track of symbols with respect to which it will perform a derivative; those are bound variables, too, so it has its own free_symbols method.

Any other method that uses bound variables should implement a free_symbols method.

classmethod fromiter(args, **assumptions)

Create a new object from an iterable.

This is a convenience function that allows one to create objects from any iterable, without having to convert to a list or tuple first.

Examples

>>> from sympy import Tuple
>>> Tuple.fromiter(i for i in range(5))
(0, 1, 2, 3, 4)
property func

The top-level function in an expression.

The following should hold for all objects:

>> x == x.func(*x.args)

Examples

>>> from sympy.abc import x
>>> a = 2*x
>>> a.func
<class 'sympy.core.mul.Mul'>
>>> a.args
(2, x)
>>> a.func(*a.args)
2*x
>>> a == a.func(*a.args)
True
gammasimp()

See the gammasimp function in sympy.simplify

getO()

Returns the additive O(..) symbol if there is one, else None.

getn()

Returns the order of the expression.

Examples

>>> from sympy import O
>>> from sympy.abc import x
>>> (1 + x + O(x**2)).getn()
2
>>> (1 + x).getn()
has(*patterns)

Test whether any subexpression matches any of the patterns.

Examples

>>> from sympy import sin
>>> from sympy.abc import x, y, z
>>> (x**2 + sin(x*y)).has(z)
False
>>> (x**2 + sin(x*y)).has(x, y, z)
True
>>> x.has(x)
True

Note has is a structural algorithm with no knowledge of mathematics. Consider the following half-open interval:

>>> from sympy import Interval
>>> i = Interval.Lopen(0, 5); i
Interval.Lopen(0, 5)
>>> i.args
(0, 5, True, False)
>>> i.has(4)  # there is no "4" in the arguments
False
>>> i.has(0)  # there *is* a "0" in the arguments
True

Instead, use contains to determine whether a number is in the interval or not:

>>> i.contains(4)
True
>>> i.contains(0)
False

Note that expr.has(*patterns) is exactly equivalent to any(expr.has(p) for p in patterns). In particular, False is returned when the list of patterns is empty.

>>> x.has()
False
has_free(*patterns)

Return True if self has object(s) x as a free expression else False.

Examples

>>> from sympy import Integral, Function
>>> from sympy.abc import x, y
>>> f = Function('f')
>>> g = Function('g')
>>> expr = Integral(f(x), (f(x), 1, g(y)))
>>> expr.free_symbols
{y}
>>> expr.has_free(g(y))
True
>>> expr.has_free(*(x, f(x)))
False

This works for subexpressions and types, too:

>>> expr.has_free(g)
True
>>> (x + y + 1).has_free(y + 1)
True
has_xfree(s: set[Basic])

Return True if self has any of the patterns in s as a free argument, else False. This is like Basic.has_free but this will only report exact argument matches.

Examples

>>> from sympy import Function
>>> from sympy.abc import x, y
>>> f = Function('f')
>>> f(x).has_xfree({f})
False
>>> f(x).has_xfree({f(x)})
True
>>> f(x + 1).has_xfree({x})
True
>>> f(x + 1).has_xfree({x + 1})
True
>>> f(x + y + 1).has_xfree({x + 1})
False
integrate(*args, **kwargs)

See the integrate function in sympy.integrals

invert(g, *gens, **args)

Return the multiplicative inverse of self mod g where self (and g) may be symbolic expressions).

See also

sympy.core.numbers.mod_inverse, sympy.polys.polytools.invert
is_Add = False
is_AlgebraicNumber = False
is_Atom = True
is_Boolean = False
is_Derivative = False
is_Dummy = True
is_Equality = False
is_Float = False
is_Function = False
is_Indexed = False
is_Integer = False
is_MatAdd = False
is_MatMul = False
is_Matrix = False
is_Mul = False
is_Not = False
is_Number = False
is_NumberSymbol = False
is_Order = False
is_Piecewise = False
is_Point = False
is_Poly = False
is_Pow = False
is_Rational = False
is_Relational = False
is_Symbol = True
is_Vector = False
is_Wild = False
property is_algebraic
is_algebraic_expr(*syms)

This tests whether a given expression is algebraic or not, in the given symbols, syms. When syms is not given, all free symbols will be used. The rational function does not have to be in expanded or in any kind of canonical form.

This function returns False for expressions that are “algebraic expressions” with symbolic exponents. This is a simple extension to the is_rational_function, including rational exponentiation.

References

Examples

>>> from sympy import Symbol, sqrt
>>> x = Symbol('x', real=True)
>>> sqrt(1 + x).is_rational_function()
False
>>> sqrt(1 + x).is_algebraic_expr()
True

This function does not attempt any nontrivial simplifications that may result in an expression that does not appear to be an algebraic expression to become one.

>>> from sympy import exp, factor
>>> a = sqrt(exp(x)**2 + 2*exp(x) + 1)/(exp(x) + 1)
>>> a.is_algebraic_expr(x)
False
>>> factor(a).is_algebraic_expr()
True
property is_antihermitian
property is_commutative
is_comparable = False
property is_complex
property is_composite
is_constant(*wrt, **flags)

Return True if self is constant, False if not, or None if the constancy could not be determined conclusively.

Examples

>>> from sympy import cos, sin, Sum, S, pi
>>> from sympy.abc import a, n, x, y
>>> x.is_constant()
False
>>> S(2).is_constant()
True
>>> Sum(x, (x, 1, 10)).is_constant()
True
>>> Sum(x, (x, 1, n)).is_constant()
False
>>> Sum(x, (x, 1, n)).is_constant(y)
True
>>> Sum(x, (x, 1, n)).is_constant(n)
False
>>> Sum(x, (x, 1, n)).is_constant(x)
True
>>> eq = a*cos(x)**2 + a*sin(x)**2 - a
>>> eq.is_constant()
True
>>> eq.subs({x: pi, a: 2}) == eq.subs({x: pi, a: 3}) == 0
True
>>> (0**x).is_constant()
False
>>> x.is_constant()
False
>>> (x**x).is_constant()
False
>>> one = cos(x)**2 + sin(x)**2
>>> one.is_constant()
True
>>> ((one - 1)**(x + 1)).is_constant() in (True, False) # could be 0 or 1
True
property is_even
property is_extended_negative
property is_extended_nonnegative
property is_extended_nonpositive
property is_extended_nonzero
property is_extended_positive
property is_extended_real
property is_finite
property is_hermitian
is_hypergeometric(k)
property is_imaginary
property is_infinite
property is_integer
property is_irrational
is_meromorphic(x, a)

This tests whether an expression is meromorphic as a function of the given symbol x at the point a.

This method is intended as a quick test that will return None if no decision can be made without simplification or more detailed analysis.

Examples

>>> from sympy import zoo, log, sin, sqrt
>>> from sympy.abc import x
>>> f = 1/x**2 + 1 - 2*x**3
>>> f.is_meromorphic(x, 0)
True
>>> f.is_meromorphic(x, 1)
True
>>> f.is_meromorphic(x, zoo)
True
>>> g = x**log(3)
>>> g.is_meromorphic(x, 0)
False
>>> g.is_meromorphic(x, 1)
True
>>> g.is_meromorphic(x, zoo)
False
>>> h = sin(1/x)*x**2
>>> h.is_meromorphic(x, 0)
False
>>> h.is_meromorphic(x, 1)
True
>>> h.is_meromorphic(x, zoo)
True

Multivalued functions are considered meromorphic when their branches are meromorphic. Thus most functions are meromorphic everywhere except at essential singularities and branch points. In particular, they will be meromorphic also on branch cuts except at their endpoints.

>>> log(x).is_meromorphic(x, -1)
True
>>> log(x).is_meromorphic(x, 0)
False
>>> sqrt(x).is_meromorphic(x, -1)
True
>>> sqrt(x).is_meromorphic(x, 0)
False
property is_negative
property is_noninteger
property is_nonnegative
property is_nonpositive
property is_nonzero
is_number = False
property is_odd
property is_polar
is_polynomial(*syms)

Return True if self is a polynomial in syms and False otherwise.

This checks if self is an exact polynomial in syms. This function returns False for expressions that are “polynomials” with symbolic exponents. Thus, you should be able to apply polynomial algorithms to expressions for which this returns True, and Poly(expr, *syms) should work if and only if expr.is_polynomial(*syms) returns True. The polynomial does not have to be in expanded form. If no symbols are given, all free symbols in the expression will be used.

This is not part of the assumptions system. You cannot do Symbol(‘z’, polynomial=True).

Examples

>>> from sympy import Symbol, Function
>>> x = Symbol('x')
>>> ((x**2 + 1)**4).is_polynomial(x)
True
>>> ((x**2 + 1)**4).is_polynomial()
True
>>> (2**x + 1).is_polynomial(x)
False
>>> (2**x + 1).is_polynomial(2**x)
True
>>> f = Function('f')
>>> (f(x) + 1).is_polynomial(x)
False
>>> (f(x) + 1).is_polynomial(f(x))
True
>>> (1/f(x) + 1).is_polynomial(f(x))
False
>>> n = Symbol('n', nonnegative=True, integer=True)
>>> (x**n + 1).is_polynomial(x)
False

This function does not attempt any nontrivial simplifications that may result in an expression that does not appear to be a polynomial to become one.

>>> from sympy import sqrt, factor, cancel
>>> y = Symbol('y', positive=True)
>>> a = sqrt(y**2 + 2*y + 1)
>>> a.is_polynomial(y)
False
>>> factor(a)
y + 1
>>> factor(a).is_polynomial(y)
True
>>> b = (y**2 + 2*y + 1)/(y + 1)
>>> b.is_polynomial(y)
False
>>> cancel(b)
y + 1
>>> cancel(b).is_polynomial(y)
True

See also .is_rational_function()

property is_positive
property is_prime
property is_rational
is_rational_function(*syms)

Test whether function is a ratio of two polynomials in the given symbols, syms. When syms is not given, all free symbols will be used. The rational function does not have to be in expanded or in any kind of canonical form.

This function returns False for expressions that are “rational functions” with symbolic exponents. Thus, you should be able to call .as_numer_denom() and apply polynomial algorithms to the result for expressions for which this returns True.

This is not part of the assumptions system. You cannot do Symbol(‘z’, rational_function=True).

Examples

>>> from sympy import Symbol, sin
>>> from sympy.abc import x, y
>>> (x/y).is_rational_function()
True
>>> (x**2).is_rational_function()
True
>>> (x/sin(y)).is_rational_function(y)
False
>>> n = Symbol('n', integer=True)
>>> (x**n + 1).is_rational_function(x)
False

This function does not attempt any nontrivial simplifications that may result in an expression that does not appear to be a rational function to become one.

>>> from sympy import sqrt, factor
>>> y = Symbol('y', positive=True)
>>> a = sqrt(y**2 + 2*y + 1)/y
>>> a.is_rational_function(y)
False
>>> factor(a)
(y + 1)/y
>>> factor(a).is_rational_function(y)
True

See also is_algebraic_expr().

property is_real
is_scalar = True
is_symbol = True
property is_transcendental
property is_zero
property kind

Default kind for all SymPy object. If the kind is not defined for the object, or if the object cannot infer the kind from its arguments, this will be returned.

Examples

>>> from sympy import Expr
>>> Expr().kind
UndefinedKind
leadterm(x, logx=None, cdir=0)

Returns the leading term a*x**b as a tuple (a, b).

Examples

>>> from sympy.abc import x
>>> (1+x+x**2).leadterm(x)
(1, 0)
>>> (1/x**2+x+x**2).leadterm(x)
(1, -2)
limit(x, xlim, dir='+')

Compute limit x->xlim.

lseries(x=None, x0=0, dir='+', logx=None, cdir=0)

Wrapper for series yielding an iterator of the terms of the series.

Note: an infinite series will yield an infinite iterator. The following, for exaxmple, will never terminate. It will just keep printing terms of the sin(x) series:

for term in sin(x).lseries(x):
    print term

The advantage of lseries() over nseries() is that many times you are just interested in the next term in the series (i.e. the first term for example), but you do not know how many you should ask for in nseries() using the “n” parameter.

See also nseries().

match(pattern, old=False)

Pattern matching.

Wild symbols match all.

Return None when expression (self) does not match with pattern. Otherwise return a dictionary such that:

pattern.xreplace(self.match(pattern)) == self

Examples

>>> from sympy import Wild, Sum
>>> from sympy.abc import x, y
>>> p = Wild("p")
>>> q = Wild("q")
>>> r = Wild("r")
>>> e = (x+y)**(x+y)
>>> e.match(p**p)
{p_: x + y}
>>> e.match(p**q)
{p_: x + y, q_: x + y}
>>> e = (2*x)**2
>>> e.match(p*q**r)
{p_: 4, q_: x, r_: 2}
>>> (p*q**r).xreplace(e.match(p*q**r))
4*x**2

Structurally bound symbols are ignored during matching:

>>> Sum(x, (x, 1, 2)).match(Sum(y, (y, 1, p)))
{p_: 2}

But they can be identified if desired:

>>> Sum(x, (x, 1, 2)).match(Sum(q, (q, 1, p)))
{p_: 2, q_: x}

The old flag will give the old-style pattern matching where expressions and patterns are essentially solved to give the match. Both of the following give None unless old=True:

>>> (x - 2).match(p - x, old=True)
{p_: 2*x - 2}
>>> (2/x).match(p*x, old=True)
{p_: 2/x**2}
matches(expr, repl_dict=None, old=False)

Helper method for match() that looks for a match between Wild symbols in self and expressions in expr.

Examples

>>> from sympy import symbols, Wild, Basic
>>> a, b, c = symbols('a b c')
>>> x = Wild('x')
>>> Basic(a + x, x).matches(Basic(a + b, c)) is None
True
>>> Basic(a + x, x).matches(Basic(a + b + c, b + c))
{x_: b + c}
n(n=15, subs=None, maxn=100, chop=False, strict=False, quad=None, verbose=False)

Evaluate the given formula to an accuracy of n digits.

Parameters:
subsdict, optional

Substitute numerical values for symbols, e.g. subs={x:3, y:1+pi}. The substitutions must be given as a dictionary.

maxnint, optional

Allow a maximum temporary working precision of maxn digits.

chopbool or number, optional

Specifies how to replace tiny real or imaginary parts in subresults by exact zeros.

When True the chop value defaults to standard precision.

Otherwise the chop value is used to determine the magnitude of “small” for purposes of chopping.

>>> from sympy import N
>>> x = 1e-4
>>> N(x, chop=True)
0.000100000000000000
>>> N(x, chop=1e-5)
0.000100000000000000
>>> N(x, chop=1e-4)
0
strictbool, optional

Raise PrecisionExhausted if any subresult fails to evaluate to full accuracy, given the available maxprec.

quadstr, optional

Choose algorithm for numerical quadrature. By default, tanh-sinh quadrature is used. For oscillatory integrals on an infinite interval, try quad='osc'.

verbosebool, optional

Print debug information.

Notes

When Floats are naively substituted into an expression, precision errors may adversely affect the result. For example, adding 1e16 (a Float) to 1 will truncate to 1e16; if 1e16 is then subtracted, the result will be 0. That is exactly what happens in the following:

>>> from sympy.abc import x, y, z
>>> values = {x: 1e16, y: 1, z: 1e16}
>>> (x + y - z).subs(values)
0

Using the subs argument for evalf is the accurate way to evaluate such an expression:

>>> (x + y - z).evalf(subs=values)
1.00000000000000
name: str
normal()

Return the expression as a fraction.

expression -> a/b

See also

as_numer_denom

return (a, b) instead of a/b

nseries(x=None, x0=0, n=6, dir='+', logx=None, cdir=0)

Wrapper to _eval_nseries if assumptions allow, else to series.

If x is given, x0 is 0, dir=’+’, and self has x, then _eval_nseries is called. This calculates “n” terms in the innermost expressions and then builds up the final series just by “cross-multiplying” everything out.

The optional logx parameter can be used to replace any log(x) in the returned series with a symbolic value to avoid evaluating log(x) at 0. A symbol to use in place of log(x) should be provided.

Advantage – it’s fast, because we do not have to determine how many terms we need to calculate in advance.

Disadvantage – you may end up with less terms than you may have expected, but the O(x**n) term appended will always be correct and so the result, though perhaps shorter, will also be correct.

If any of those assumptions is not met, this is treated like a wrapper to series which will try harder to return the correct number of terms.

See also lseries().

Examples

>>> from sympy import sin, log, Symbol
>>> from sympy.abc import x, y
>>> sin(x).nseries(x, 0, 6)
x - x**3/6 + x**5/120 + O(x**6)
>>> log(x+1).nseries(x, 0, 5)
x - x**2/2 + x**3/3 - x**4/4 + O(x**5)

Handling of the logx parameter — in the following example the expansion fails since sin does not have an asymptotic expansion at -oo (the limit of log(x) as x approaches 0):

>>> e = sin(log(x))
>>> e.nseries(x, 0, 6)
Traceback (most recent call last):
...
PoleError: ...
...
>>> logx = Symbol('logx')
>>> e.nseries(x, 0, 6, logx=logx)
sin(logx)

In the following example, the expansion works but only returns self unless the logx parameter is used:

>>> e = x**y
>>> e.nseries(x, 0, 2)
x**y
>>> e.nseries(x, 0, 2, logx=logx)
exp(logx*y)
nsimplify(constants=(), tolerance=None, full=False)

See the nsimplify function in sympy.simplify

powsimp(*args, **kwargs)

See the powsimp function in sympy.simplify

primitive()

Return the positive Rational that can be extracted non-recursively from every term of self (i.e., self is treated like an Add). This is like the as_coeff_Mul() method but primitive always extracts a positive Rational (never a negative or a Float).

Examples

>>> from sympy.abc import x
>>> (3*(x + 1)**2).primitive()
(3, (x + 1)**2)
>>> a = (6*x + 2); a.primitive()
(2, 3*x + 1)
>>> b = (x/2 + 3); b.primitive()
(1/2, x + 6)
>>> (a*b).primitive() == (1, a*b)
True
radsimp(**kwargs)

See the radsimp function in sympy.simplify

ratsimp()

See the ratsimp function in sympy.simplify

rcall(*args)

Apply on the argument recursively through the expression tree.

This method is used to simulate a common abuse of notation for operators. For instance, in SymPy the following will not work:

(x+Lambda(y, 2*y))(z) == x+2*z,

however, you can use:

>>> from sympy import Lambda
>>> from sympy.abc import x, y, z
>>> (x + Lambda(y, 2*y)).rcall(z)
x + 2*z
refine(assumption=True)

See the refine function in sympy.assumptions

removeO()

Removes the additive O(..) symbol if there is one

replace(query, value, map=False, simultaneous=True, exact=None)

Replace matching subexpressions of self with value.

If map = True then also return the mapping {old: new} where old was a sub-expression found with query and new is the replacement value for it. If the expression itself does not match the query, then the returned value will be self.xreplace(map) otherwise it should be self.subs(ordered(map.items())).

Traverses an expression tree and performs replacement of matching subexpressions from the bottom to the top of the tree. The default approach is to do the replacement in a simultaneous fashion so changes made are targeted only once. If this is not desired or causes problems, simultaneous can be set to False.

In addition, if an expression containing more than one Wild symbol is being used to match subexpressions and the exact flag is None it will be set to True so the match will only succeed if all non-zero values are received for each Wild that appears in the match pattern. Setting this to False accepts a match of 0; while setting it True accepts all matches that have a 0 in them. See example below for cautions.

The list of possible combinations of queries and replacement values is listed below:

See also

subs

substitution of subexpressions as defined by the objects themselves.

xreplace

exact node replacement in expr tree; also capable of using matching rules

Examples

Initial setup

>>> from sympy import log, sin, cos, tan, Wild, Mul, Add
>>> from sympy.abc import x, y
>>> f = log(sin(x)) + tan(sin(x**2))
1.1. type -> type

obj.replace(type, newtype)

When object of type type is found, replace it with the result of passing its argument(s) to newtype.

>>> f.replace(sin, cos)
log(cos(x)) + tan(cos(x**2))
>>> sin(x).replace(sin, cos, map=True)
(cos(x), {sin(x): cos(x)})
>>> (x*y).replace(Mul, Add)
x + y
1.2. type -> func

obj.replace(type, func)

When object of type type is found, apply func to its argument(s). func must be written to handle the number of arguments of type.

>>> f.replace(sin, lambda arg: sin(2*arg))
log(sin(2*x)) + tan(sin(2*x**2))
>>> (x*y).replace(Mul, lambda *args: sin(2*Mul(*args)))
sin(2*x*y)
2.1. pattern -> expr

obj.replace(pattern(wild), expr(wild))

Replace subexpressions matching pattern with the expression written in terms of the Wild symbols in pattern.

>>> a, b = map(Wild, 'ab')
>>> f.replace(sin(a), tan(a))
log(tan(x)) + tan(tan(x**2))
>>> f.replace(sin(a), tan(a/2))
log(tan(x/2)) + tan(tan(x**2/2))
>>> f.replace(sin(a), a)
log(x) + tan(x**2)
>>> (x*y).replace(a*x, a)
y

Matching is exact by default when more than one Wild symbol is used: matching fails unless the match gives non-zero values for all Wild symbols:

>>> (2*x + y).replace(a*x + b, b - a)
y - 2
>>> (2*x).replace(a*x + b, b - a)
2*x

When set to False, the results may be non-intuitive:

>>> (2*x).replace(a*x + b, b - a, exact=False)
2/x
2.2. pattern -> func

obj.replace(pattern(wild), lambda wild: expr(wild))

All behavior is the same as in 2.1 but now a function in terms of pattern variables is used rather than an expression:

>>> f.replace(sin(a), lambda a: sin(2*a))
log(sin(2*x)) + tan(sin(2*x**2))
3.1. func -> func

obj.replace(filter, func)

Replace subexpression e with func(e) if filter(e) is True.

>>> g = 2*sin(x**3)
>>> g.replace(lambda expr: expr.is_Number, lambda expr: expr**2)
4*sin(x**9)

The expression itself is also targeted by the query but is done in such a fashion that changes are not made twice.

>>> e = x*(x*y + 1)
>>> e.replace(lambda x: x.is_Mul, lambda x: 2*x)
2*x*(2*x*y + 1)

When matching a single symbol, exact will default to True, but this may or may not be the behavior that is desired:

Here, we want exact=False:

>>> from sympy import Function
>>> f = Function('f')
>>> e = f(1) + f(0)
>>> q = f(a), lambda a: f(a + 1)
>>> e.replace(*q, exact=False)
f(1) + f(2)
>>> e.replace(*q, exact=True)
f(0) + f(2)

But here, the nature of matching makes selecting the right setting tricky:

>>> e = x**(1 + y)
>>> (x**(1 + y)).replace(x**(1 + a), lambda a: x**-a, exact=False)
x
>>> (x**(1 + y)).replace(x**(1 + a), lambda a: x**-a, exact=True)
x**(-x - y + 1)
>>> (x**y).replace(x**(1 + a), lambda a: x**-a, exact=False)
x
>>> (x**y).replace(x**(1 + a), lambda a: x**-a, exact=True)
x**(1 - y)

It is probably better to use a different form of the query that describes the target expression more precisely:

>>> (1 + x**(1 + y)).replace(
... lambda x: x.is_Pow and x.exp.is_Add and x.exp.args[0] == 1,
... lambda x: x.base**(1 - (x.exp - 1)))
...
x**(1 - y) + 1
rewrite(*args, deep=True, **hints)

Rewrite self using a defined rule.

Rewriting transforms an expression to another, which is mathematically equivalent but structurally different. For example you can rewrite trigonometric functions as complex exponentials or combinatorial functions as gamma function.

This method takes a pattern and a rule as positional arguments. pattern is optional parameter which defines the types of expressions that will be transformed. If it is not passed, all possible expressions will be rewritten. rule defines how the expression will be rewritten.

Parameters:
argsExpr

A rule, or pattern and rule. - pattern is a type or an iterable of types. - rule can be any object.

deepbool, optional

If True, subexpressions are recursively transformed. Default is True.

Examples

If pattern is unspecified, all possible expressions are transformed.

>>> from sympy import cos, sin, exp, I
>>> from sympy.abc import x
>>> expr = cos(x) + I*sin(x)
>>> expr.rewrite(exp)
exp(I*x)

Pattern can be a type or an iterable of types.

>>> expr.rewrite(sin, exp)
exp(I*x)/2 + cos(x) - exp(-I*x)/2
>>> expr.rewrite([cos,], exp)
exp(I*x)/2 + I*sin(x) + exp(-I*x)/2
>>> expr.rewrite([cos, sin], exp)
exp(I*x)

Rewriting behavior can be implemented by defining _eval_rewrite() method.

>>> from sympy import Expr, sqrt, pi
>>> class MySin(Expr):
...     def _eval_rewrite(self, rule, args, **hints):
...         x, = args
...         if rule == cos:
...             return cos(pi/2 - x, evaluate=False)
...         if rule == sqrt:
...             return sqrt(1 - cos(x)**2)
>>> MySin(MySin(x)).rewrite(cos)
cos(-cos(-x + pi/2) + pi/2)
>>> MySin(x).rewrite(sqrt)
sqrt(1 - cos(x)**2)

Defining _eval_rewrite_as_[...]() method is supported for backwards compatibility reason. This may be removed in the future and using it is discouraged.

>>> class MySin(Expr):
...     def _eval_rewrite_as_cos(self, *args, **hints):
...         x, = args
...         return cos(pi/2 - x, evaluate=False)
>>> MySin(x).rewrite(cos)
cos(-x + pi/2)
round(n=None)

Return x rounded to the given decimal place.

If a complex number would results, apply round to the real and imaginary components of the number.

Notes

The Python round function uses the SymPy round method so it will always return a SymPy number (not a Python float or int):

>>> isinstance(round(S(123), -2), Number)
True

Examples

>>> from sympy import pi, E, I, S, Number
>>> pi.round()
3
>>> pi.round(2)
3.14
>>> (2*pi + E*I).round()
6 + 3*I

The round method has a chopping effect:

>>> (2*pi + I/10).round()
6
>>> (pi/10 + 2*I).round()
2*I
>>> (pi/10 + E*I).round(2)
0.31 + 2.72*I
separate(deep=False, force=False)

See the separate function in sympy.simplify

series(x=None, x0=0, n=6, dir='+', logx=None, cdir=0)

Series expansion of “self” around x = x0 yielding either terms of the series one by one (the lazy series given when n=None), else all the terms at once when n != None.

Returns the series expansion of “self” around the point x = x0 with respect to x up to O((x - x0)**n, x, x0) (default n is 6).

If x=None and self is univariate, the univariate symbol will be supplied, otherwise an error will be raised.

Parameters:
exprExpression

The expression whose series is to be expanded.

xSymbol

It is the variable of the expression to be calculated.

x0Value

The value around which x is calculated. Can be any value from -oo to oo.

nValue

The value used to represent the order in terms of x**n, up to which the series is to be expanded.

dirString, optional

The series-expansion can be bi-directional. If dir="+", then (x->x0+). If dir="-", then (x->x0-). For infinite ``x0 (oo or -oo), the dir argument is determined from the direction of the infinity (i.e., dir="-" for oo).

logxoptional

It is used to replace any log(x) in the returned series with a symbolic value rather than evaluating the actual value.

cdiroptional

It stands for complex direction, and indicates the direction from which the expansion needs to be evaluated.

Returns:
ExprExpression

Series expansion of the expression about x0

Raises:
TypeError

If “n” and “x0” are infinity objects

PoleError

If “x0” is an infinity object

Examples

>>> from sympy import cos, exp, tan
>>> from sympy.abc import x, y
>>> cos(x).series()
1 - x**2/2 + x**4/24 + O(x**6)
>>> cos(x).series(n=4)
1 - x**2/2 + O(x**4)
>>> cos(x).series(x, x0=1, n=2)
cos(1) - (x - 1)*sin(1) + O((x - 1)**2, (x, 1))
>>> e = cos(x + exp(y))
>>> e.series(y, n=2)
cos(x + 1) - y*sin(x + 1) + O(y**2)
>>> e.series(x, n=2)
cos(exp(y)) - x*sin(exp(y)) + O(x**2)

If n=None then a generator of the series terms will be returned.

>>> term=cos(x).series(n=None)
>>> [next(term) for i in range(2)]
[1, -x**2/2]

For dir=+ (default) the series is calculated from the right and for dir=- the series from the left. For smooth functions this flag will not alter the results.

>>> abs(x).series(dir="+")
x
>>> abs(x).series(dir="-")
-x
>>> f = tan(x)
>>> f.series(x, 2, 6, "+")
tan(2) + (1 + tan(2)**2)*(x - 2) + (x - 2)**2*(tan(2)**3 + tan(2)) +
(x - 2)**3*(1/3 + 4*tan(2)**2/3 + tan(2)**4) + (x - 2)**4*(tan(2)**5 +
5*tan(2)**3/3 + 2*tan(2)/3) + (x - 2)**5*(2/15 + 17*tan(2)**2/15 +
2*tan(2)**4 + tan(2)**6) + O((x - 2)**6, (x, 2))
>>> f.series(x, 2, 3, "-")
tan(2) + (2 - x)*(-tan(2)**2 - 1) + (2 - x)**2*(tan(2)**3 + tan(2))
+ O((x - 2)**3, (x, 2))

For rational expressions this method may return original expression without the Order term. >>> (1/x).series(x, n=8) 1/x

simplify(**kwargs)

See the simplify function in sympy.simplify

sort_key(order=None)

Return a sort key.

Examples

>>> from sympy import S, I
>>> sorted([S(1)/2, I, -I], key=lambda x: x.sort_key())
[1/2, -I, I]
>>> S("[x, 1/x, 1/x**2, x**2, x**(1/2), x**(1/4), x**(3/2)]")
[x, 1/x, x**(-2), x**2, sqrt(x), x**(1/4), x**(3/2)]
>>> sorted(_, key=lambda x: x.sort_key())
[x**(-2), 1/x, x**(1/4), sqrt(x), x, x**(3/2), x**2]
subs(*args, **kwargs)

Substitutes old for new in an expression after sympifying args.

args is either:
  • two arguments, e.g. foo.subs(old, new)

  • one iterable argument, e.g. foo.subs(iterable). The iterable may be
    o an iterable container with (old, new) pairs. In this case the

    replacements are processed in the order given with successive patterns possibly affecting replacements already made.

    o a dict or set whose key/value items correspond to old/new pairs.

    In this case the old/new pairs will be sorted by op count and in case of a tie, by number of args and the default_sort_key. The resulting sorted list is then processed as an iterable container (see previous).

If the keyword simultaneous is True, the subexpressions will not be evaluated until all the substitutions have been made.

See also

replace

replacement capable of doing wildcard-like matching, parsing of match, and conditional replacements

xreplace

exact node replacement in expr tree; also capable of using matching rules

sympy.core.evalf.EvalfMixin.evalf

calculates the given formula to a desired level of precision

Examples

>>> from sympy import pi, exp, limit, oo
>>> from sympy.abc import x, y
>>> (1 + x*y).subs(x, pi)
pi*y + 1
>>> (1 + x*y).subs({x:pi, y:2})
1 + 2*pi
>>> (1 + x*y).subs([(x, pi), (y, 2)])
1 + 2*pi
>>> reps = [(y, x**2), (x, 2)]
>>> (x + y).subs(reps)
6
>>> (x + y).subs(reversed(reps))
x**2 + 2
>>> (x**2 + x**4).subs(x**2, y)
y**2 + y

To replace only the x**2 but not the x**4, use xreplace:

>>> (x**2 + x**4).xreplace({x**2: y})
x**4 + y

To delay evaluation until all substitutions have been made, set the keyword simultaneous to True:

>>> (x/y).subs([(x, 0), (y, 0)])
0
>>> (x/y).subs([(x, 0), (y, 0)], simultaneous=True)
nan

This has the added feature of not allowing subsequent substitutions to affect those already made:

>>> ((x + y)/y).subs({x + y: y, y: x + y})
1
>>> ((x + y)/y).subs({x + y: y, y: x + y}, simultaneous=True)
y/(x + y)

In order to obtain a canonical result, unordered iterables are sorted by count_op length, number of arguments and by the default_sort_key to break any ties. All other iterables are left unsorted.

>>> from sympy import sqrt, sin, cos
>>> from sympy.abc import a, b, c, d, e
>>> A = (sqrt(sin(2*x)), a)
>>> B = (sin(2*x), b)
>>> C = (cos(2*x), c)
>>> D = (x, d)
>>> E = (exp(x), e)
>>> expr = sqrt(sin(2*x))*sin(exp(x)*x)*cos(2*x) + sin(2*x)
>>> expr.subs(dict([A, B, C, D, E]))
a*c*sin(d*e) + b

The resulting expression represents a literal replacement of the old arguments with the new arguments. This may not reflect the limiting behavior of the expression:

>>> (x**3 - 3*x).subs({x: oo})
nan
>>> limit(x**3 - 3*x, x, oo)
oo

If the substitution will be followed by numerical evaluation, it is better to pass the substitution to evalf as

>>> (1/x).evalf(subs={x: 3.0}, n=21)
0.333333333333333333333

rather than

>>> (1/x).subs({x: 3.0}).evalf(21)
0.333333333333333314830

as the former will ensure that the desired level of precision is obtained.

taylor_term(n, x, *previous_terms)

General method for the taylor term.

This method is slow, because it differentiates n-times. Subclasses can redefine it to make it faster by using the “previous_terms”.

to_nnf(simplify=True)
together(*args, **kwargs)

See the together function in sympy.polys

transpose()
trigsimp(**args)

See the trigsimp function in sympy.simplify

xreplace(rule, hack2=False)

Replace occurrences of objects within the expression.

Parameters:
ruledict-like

Expresses a replacement rule

Returns:
xreplacethe result of the replacement

See also

replace

replacement capable of doing wildcard-like matching, parsing of match, and conditional replacements

subs

substitution of subexpressions as defined by the objects themselves.

Examples

>>> from sympy import symbols, pi, exp
>>> x, y, z = symbols('x y z')
>>> (1 + x*y).xreplace({x: pi})
pi*y + 1
>>> (1 + x*y).xreplace({x: pi, y: 2})
1 + 2*pi

Replacements occur only if an entire node in the expression tree is matched:

>>> (x*y + z).xreplace({x*y: pi})
z + pi
>>> (x*y*z).xreplace({x*y: pi})
x*y*z
>>> (2*x).xreplace({2*x: y, x: z})
y
>>> (2*2*x).xreplace({2*x: y, x: z})
4*z
>>> (x + y + 2).xreplace({x + y: 2})
x + y + 2
>>> (x + 2 + exp(x + 2)).xreplace({x + 2: y})
x + exp(y) + 2

xreplace does not differentiate between free and bound symbols. In the following, subs(x, y) would not change x since it is a bound symbol, but xreplace does:

>>> from sympy import Integral
>>> Integral(x, (x, 1, 2*x)).xreplace({x: y})
Integral(y, (y, 1, 2*y))

Trying to replace x with an expression raises an error:

>>> Integral(x, (x, 1, 2*x)).xreplace({x: 2*y}) 
ValueError: Invalid limits given: ((2*y, 1, 4*y),)

Factor

class nipy.algorithms.statistics.formula.formulae.Factor(name, levels, char='b')

Bases: Formula

A qualitative variable in a regression model

A Factor is similar to R’s factor. The levels of the Factor can be either strings or ints.

__init__(name, levels, char='b')

Initialize Factor

Parameters:
namestr
levels[str or int]

A sequence of strings or ints.

charstr, optional

prefix character for regression coefficients

property coefs

Coefficients in the linear regression formula.

design(input, param=None, return_float=False, contrasts=None)

Construct the design matrix, and optional contrast matrices.

Parameters:
inputnp.recarray

Recarray including fields needed to compute the Terms in getparams(self.design_expr).

paramNone or np.recarray

Recarray including fields that are not Terms in getparams(self.design_expr)

return_floatbool, optional

If True, return a np.float64 array rather than a np.recarray

contrastsNone or dict, optional

Contrasts. The items in this dictionary should be (str, Formula) pairs where a contrast matrix is constructed for each Formula by evaluating its design at the same parameters as self.design. If not None, then the return_float is set to True.

Returns:
des2D array

design matrix

cmatricesdict, optional

Dictionary with keys from contrasts input, and contrast matrices corresponding to des design matrix. Returned only if contrasts input is not None

property design_expr
property dtype

The dtype of the design matrix of the Formula.

static fromcol(col, name)

Create a Factor from a column array.

Parameters:
colndarray

an array with ndim==1

namestr

name of the Factor

Returns:
factorFactor

Examples

>>> data = np.array([(3,'a'),(4,'a'),(5,'b'),(3,'b')], np.dtype([('x', np.float64), ('y', 'S1')]))
>>> f1 = Factor.fromcol(data['y'], 'y')
>>> f2 = Factor.fromcol(data['x'], 'x')
>>> d = f1.design(data)
>>> print(d.dtype.descr)
[('y_a', '<f8'), ('y_b', '<f8')]
>>> d = f2.design(data)
>>> print(d.dtype.descr)
[('x_3', '<f8'), ('x_4', '<f8'), ('x_5', '<f8')]
static fromrec(rec, keep=[], drop=[])

Construct Formula from recarray

For fields with a string-dtype, it is assumed that these are qualtiatitve regressors, i.e. Factors.

Parameters:
rec: recarray

Recarray whose field names will be used to create a formula.

keep: []

Field names to explicitly keep, dropping all others.

drop: []

Field names to drop.

get_term(level)

Retrieve a term of the Factor…

property main_effect
property mean

Expression for the mean, expressed as a linear combination of terms, each with dummy variables in front.

property params

The parameters in the Formula.

stratify(variable)

Create a new variable, stratified by the levels of a Factor.

Parameters:
variablestr or simple sympy expression

If sympy expression, then string representation must be all lower or upper case letters, i.e. it can be interpreted as a name.

Returns:
formulaFormula

Formula whose mean has one parameter named variable%d, for each level in self.levels

Examples

>>> f = Factor('a', ['x','y'])
>>> sf = f.stratify('theta')
>>> sf.mean
_theta0*a_x + _theta1*a_y
subs(old, new)

Perform a sympy substitution on all terms in the Formula

Returns a new instance of the same class

Parameters:
oldsympy.Basic

The expression to be changed

newsympy.Basic

The value to change it to.

Returns:
newfFormula

Examples

>>> s, t = [Term(l) for l in 'st']
>>> f, g = [sympy.Function(l) for l in 'fg']
>>> form = Formula([f(t),g(s)])
>>> newform = form.subs(g, sympy.Function('h'))
>>> newform.terms
array([f(t), h(s)], dtype=object)
>>> form.terms
array([f(t), g(s)], dtype=object)
property terms

Terms in the linear regression formula.

FactorTerm

class nipy.algorithms.statistics.formula.formulae.FactorTerm(name, level)

Bases: Term

Boolean Term derived from a Factor.

Its properties are the same as a Term except that its product with itself is itself.

__init__(*args, **kwargs)
adjoint()
apart(x=None, **args)

See the apart function in sympy.polys

property args: tuple[Basic, ...]

Returns a tuple of arguments of ‘self’.

Notes

Never use self._args, always use self.args. Only use _args in __new__ when creating a new function. Do not override .args() from Basic (so that it is easy to change the interface in the future if needed).

Examples

>>> from sympy import cot
>>> from sympy.abc import x, y
>>> cot(x).args
(x,)
>>> cot(x).args[0]
x
>>> (x*y).args
(x, y)
>>> (x*y).args[1]
y
args_cnc(cset=False, warn=True, split_1=True)

Return [commutative factors, non-commutative factors] of self.

Examples

>>> from sympy import symbols, oo
>>> A, B = symbols('A B', commutative=0)
>>> x, y = symbols('x y')
>>> (-2*x*y).args_cnc()
[[-1, 2, x, y], []]
>>> (-2.5*x).args_cnc()
[[-1, 2.5, x], []]
>>> (-2*x*A*B*y).args_cnc()
[[-1, 2, x, y], [A, B]]
>>> (-2*x*A*B*y).args_cnc(split_1=False)
[[-2, x, y], [A, B]]
>>> (-2*x*y).args_cnc(cset=True)
[{-1, 2, x, y}, []]

The arg is always treated as a Mul:

>>> (-2 + x + A).args_cnc()
[[], [x - 2 + A]]
>>> (-oo).args_cnc() # -oo is a singleton
[[-1, oo], []]
as_base_exp() tuple[Expr, Expr]
as_coeff_Add(rational=False) tuple[Number, Expr]

Efficiently extract the coefficient of a summation.

as_coeff_Mul(rational: bool = False) tuple[Number, Expr]

Efficiently extract the coefficient of a product.

as_coeff_add(*deps) tuple[Expr, tuple[Expr, ...]]

Return the tuple (c, args) where self is written as an Add, a.

c should be a Rational added to any terms of the Add that are independent of deps.

args should be a tuple of all other terms of a; args is empty if self is a Number or if self is independent of deps (when given).

This should be used when you do not know if self is an Add or not but you want to treat self as an Add or if you want to process the individual arguments of the tail of self as an Add.

  • if you know self is an Add and want only the head, use self.args[0];

  • if you do not want to process the arguments of the tail but need the tail then use self.as_two_terms() which gives the head and tail.

  • if you want to split self into an independent and dependent parts use self.as_independent(*deps)

>>> from sympy import S
>>> from sympy.abc import x, y
>>> (S(3)).as_coeff_add()
(3, ())
>>> (3 + x).as_coeff_add()
(3, (x,))
>>> (3 + x + y).as_coeff_add(x)
(y + 3, (x,))
>>> (3 + y).as_coeff_add(x)
(y + 3, ())
as_coeff_exponent(x) tuple[Expr, Expr]

c*x**e -> c,e where x can be any symbolic expression.

as_coeff_mul(*deps, **kwargs) tuple[Expr, tuple[Expr, ...]]

Return the tuple (c, args) where self is written as a Mul, m.

c should be a Rational multiplied by any factors of the Mul that are independent of deps.

args should be a tuple of all other factors of m; args is empty if self is a Number or if self is independent of deps (when given).

This should be used when you do not know if self is a Mul or not but you want to treat self as a Mul or if you want to process the individual arguments of the tail of self as a Mul.

  • if you know self is a Mul and want only the head, use self.args[0];

  • if you do not want to process the arguments of the tail but need the tail then use self.as_two_terms() which gives the head and tail;

  • if you want to split self into an independent and dependent parts use self.as_independent(*deps)

>>> from sympy import S
>>> from sympy.abc import x, y
>>> (S(3)).as_coeff_mul()
(3, ())
>>> (3*x*y).as_coeff_mul()
(3, (x, y))
>>> (3*x*y).as_coeff_mul(x)
(3*y, (x,))
>>> (3*y).as_coeff_mul(x)
(3*y, ())
as_coefficient(expr)

Extracts symbolic coefficient at the given expression. In other words, this functions separates ‘self’ into the product of ‘expr’ and ‘expr’-free coefficient. If such separation is not possible it will return None.

See also

coeff

return sum of terms have a given factor

as_coeff_Add

separate the additive constant from an expression

as_coeff_Mul

separate the multiplicative constant from an expression

as_independent

separate x-dependent terms/factors from others

sympy.polys.polytools.Poly.coeff_monomial

efficiently find the single coefficient of a monomial in Poly

sympy.polys.polytools.Poly.nth

like coeff_monomial but powers of monomial terms are used

Examples

>>> from sympy import E, pi, sin, I, Poly
>>> from sympy.abc import x
>>> E.as_coefficient(E)
1
>>> (2*E).as_coefficient(E)
2
>>> (2*sin(E)*E).as_coefficient(E)

Two terms have E in them so a sum is returned. (If one were desiring the coefficient of the term exactly matching E then the constant from the returned expression could be selected. Or, for greater precision, a method of Poly can be used to indicate the desired term from which the coefficient is desired.)

>>> (2*E + x*E).as_coefficient(E)
x + 2
>>> _.args[0]  # just want the exact match
2
>>> p = Poly(2*E + x*E); p
Poly(x*E + 2*E, x, E, domain='ZZ')
>>> p.coeff_monomial(E)
2
>>> p.nth(0, 1)
2

Since the following cannot be written as a product containing E as a factor, None is returned. (If the coefficient 2*x is desired then the coeff method should be used.)

>>> (2*E*x + x).as_coefficient(E)
>>> (2*E*x + x).coeff(E)
2*x
>>> (E*(x + 1) + x).as_coefficient(E)
>>> (2*pi*I).as_coefficient(pi*I)
2
>>> (2*I).as_coefficient(pi*I)
as_coefficients_dict(*syms)

Return a dictionary mapping terms to their Rational coefficient. Since the dictionary is a defaultdict, inquiries about terms which were not present will return a coefficient of 0.

If symbols syms are provided, any multiplicative terms independent of them will be considered a coefficient and a regular dictionary of syms-dependent generators as keys and their corresponding coefficients as values will be returned.

Examples

>>> from sympy.abc import a, x, y
>>> (3*x + a*x + 4).as_coefficients_dict()
{1: 4, x: 3, a*x: 1}
>>> _[a]
0
>>> (3*a*x).as_coefficients_dict()
{a*x: 3}
>>> (3*a*x).as_coefficients_dict(x)
{x: 3*a}
>>> (3*a*x).as_coefficients_dict(y)
{1: 3*a*x}
as_content_primitive(radical=False, clear=True)

This method should recursively remove a Rational from all arguments and return that (content) and the new self (primitive). The content should always be positive and Mul(*foo.as_content_primitive()) == foo. The primitive need not be in canonical form and should try to preserve the underlying structure if possible (i.e. expand_mul should not be applied to self).

Examples

>>> from sympy import sqrt
>>> from sympy.abc import x, y, z
>>> eq = 2 + 2*x + 2*y*(3 + 3*y)

The as_content_primitive function is recursive and retains structure:

>>> eq.as_content_primitive()
(2, x + 3*y*(y + 1) + 1)

Integer powers will have Rationals extracted from the base:

>>> ((2 + 6*x)**2).as_content_primitive()
(4, (3*x + 1)**2)
>>> ((2 + 6*x)**(2*y)).as_content_primitive()
(1, (2*(3*x + 1))**(2*y))

Terms may end up joining once their as_content_primitives are added:

>>> ((5*(x*(1 + y)) + 2*x*(3 + 3*y))).as_content_primitive()
(11, x*(y + 1))
>>> ((3*(x*(1 + y)) + 2*x*(3 + 3*y))).as_content_primitive()
(9, x*(y + 1))
>>> ((3*(z*(1 + y)) + 2.0*x*(3 + 3*y))).as_content_primitive()
(1, 6.0*x*(y + 1) + 3*z*(y + 1))
>>> ((5*(x*(1 + y)) + 2*x*(3 + 3*y))**2).as_content_primitive()
(121, x**2*(y + 1)**2)
>>> ((x*(1 + y) + 0.4*x*(3 + 3*y))**2).as_content_primitive()
(1, 4.84*x**2*(y + 1)**2)

Radical content can also be factored out of the primitive:

>>> (2*sqrt(2) + 4*sqrt(10)).as_content_primitive(radical=True)
(2, sqrt(2)*(1 + 2*sqrt(5)))

If clear=False (default is True) then content will not be removed from an Add if it can be distributed to leave one or more terms with integer coefficients.

>>> (x/2 + y).as_content_primitive()
(1/2, x + 2*y)
>>> (x/2 + y).as_content_primitive(clear=False)
(1, x/2 + y)
as_dummy()

Return the expression with any objects having structurally bound symbols replaced with unique, canonical symbols within the object in which they appear and having only the default assumption for commutativity being True. When applied to a symbol a new symbol having only the same commutativity will be returned.

Notes

Any object that has structurally bound variables should have a property, bound_symbols that returns those symbols appearing in the object.

Examples

>>> from sympy import Integral, Symbol
>>> from sympy.abc import x
>>> r = Symbol('r', real=True)
>>> Integral(r, (r, x)).as_dummy()
Integral(_0, (_0, x))
>>> _.variables[0].is_real is None
True
>>> r.as_dummy()
_r
as_expr(*gens)

Convert a polynomial to a SymPy expression.

Examples

>>> from sympy import sin
>>> from sympy.abc import x, y
>>> f = (x**2 + x*y).as_poly(x, y)
>>> f.as_expr()
x**2 + x*y
>>> sin(x).as_expr()
sin(x)
as_independent(*deps, **hint) tuple[Expr, Expr]

A mostly naive separation of a Mul or Add into arguments that are not are dependent on deps. To obtain as complete a separation of variables as possible, use a separation method first, e.g.:

  • separatevars() to change Mul, Add and Pow (including exp) into Mul

  • .expand(mul=True) to change Add or Mul into Add

  • .expand(log=True) to change log expr into an Add

The only non-naive thing that is done here is to respect noncommutative ordering of variables and to always return (0, 0) for self of zero regardless of hints.

For nonzero self, the returned tuple (i, d) has the following interpretation:

  • i will has no variable that appears in deps

  • d will either have terms that contain variables that are in deps, or be equal to 0 (when self is an Add) or 1 (when self is a Mul)

  • if self is an Add then self = i + d

  • if self is a Mul then self = i*d

  • otherwise (self, S.One) or (S.One, self) is returned.

To force the expression to be treated as an Add, use the hint as_Add=True

See also

separatevars
expand_log
sympy.core.add.Add.as_two_terms
sympy.core.mul.Mul.as_two_terms
as_coeff_mul

Examples

– self is an Add

>>> from sympy import sin, cos, exp
>>> from sympy.abc import x, y, z
>>> (x + x*y).as_independent(x)
(0, x*y + x)
>>> (x + x*y).as_independent(y)
(x, x*y)
>>> (2*x*sin(x) + y + x + z).as_independent(x)
(y + z, 2*x*sin(x) + x)
>>> (2*x*sin(x) + y + x + z).as_independent(x, y)
(z, 2*x*sin(x) + x + y)

– self is a Mul

>>> (x*sin(x)*cos(y)).as_independent(x)
(cos(y), x*sin(x))

non-commutative terms cannot always be separated out when self is a Mul

>>> from sympy import symbols
>>> n1, n2, n3 = symbols('n1 n2 n3', commutative=False)
>>> (n1 + n1*n2).as_independent(n2)
(n1, n1*n2)
>>> (n2*n1 + n1*n2).as_independent(n2)
(0, n1*n2 + n2*n1)
>>> (n1*n2*n3).as_independent(n1)
(1, n1*n2*n3)
>>> (n1*n2*n3).as_independent(n2)
(n1, n2*n3)
>>> ((x-n1)*(x-y)).as_independent(x)
(1, (x - y)*(x - n1))

– self is anything else:

>>> (sin(x)).as_independent(x)
(1, sin(x))
>>> (sin(x)).as_independent(y)
(sin(x), 1)
>>> exp(x+y).as_independent(x)
(1, exp(x + y))

– force self to be treated as an Add:

>>> (3*x).as_independent(x, as_Add=True)
(0, 3*x)

– force self to be treated as a Mul:

>>> (3+x).as_independent(x, as_Add=False)
(1, x + 3)
>>> (-3+x).as_independent(x, as_Add=False)
(1, x - 3)

Note how the below differs from the above in making the constant on the dep term positive.

>>> (y*(-3+x)).as_independent(x)
(y, x - 3)
– use .as_independent() for true independence testing instead

of .has(). The former considers only symbols in the free symbols while the latter considers all symbols

>>> from sympy import Integral
>>> I = Integral(x, (x, 1, 2))
>>> I.has(x)
True
>>> x in I.free_symbols
False
>>> I.as_independent(x) == (I, 1)
True
>>> (I + x).as_independent(x) == (I, x)
True

Note: when trying to get independent terms, a separation method might need to be used first. In this case, it is important to keep track of what you send to this routine so you know how to interpret the returned values

>>> from sympy import separatevars, log
>>> separatevars(exp(x+y)).as_independent(x)
(exp(y), exp(x))
>>> (x + x*y).as_independent(y)
(x, x*y)
>>> separatevars(x + x*y).as_independent(y)
(x, y + 1)
>>> (x*(1 + y)).as_independent(y)
(x, y + 1)
>>> (x*(1 + y)).expand(mul=True).as_independent(y)
(x, x*y)
>>> a, b=symbols('a b', positive=True)
>>> (log(a*b).expand(log=True)).as_independent(b)
(log(a), log(b))
as_leading_term(*symbols, logx=None, cdir=0)

Returns the leading (nonzero) term of the series expansion of self.

The _eval_as_leading_term routines are used to do this, and they must always return a non-zero value.

Examples

>>> from sympy.abc import x
>>> (1 + x + x**2).as_leading_term(x)
1
>>> (1/x**2 + x + x**2).as_leading_term(x)
x**(-2)
as_numer_denom()

Return the numerator and the denominator of an expression.

expression -> a/b -> a, b

This is just a stub that should be defined by an object’s class methods to get anything else.

See also

normal

return a/b instead of (a, b)

as_ordered_factors(order=None)

Return list of ordered factors (if Mul) else [self].

as_ordered_terms(order=None, data=False)

Transform an expression to an ordered list of terms.

Examples

>>> from sympy import sin, cos
>>> from sympy.abc import x
>>> (sin(x)**2*cos(x) + sin(x)**2 + 1).as_ordered_terms()
[sin(x)**2*cos(x), sin(x)**2, 1]
as_poly(*gens, **args)

Converts self to a polynomial or returns None.

as_powers_dict()

Return self as a dictionary of factors with each factor being treated as a power. The keys are the bases of the factors and the values, the corresponding exponents. The resulting dictionary should be used with caution if the expression is a Mul and contains non- commutative factors since the order that they appeared will be lost in the dictionary.

See also

as_ordered_factors

An alternative for noncommutative applications, returning an ordered list of factors.

args_cnc

Similar to as_ordered_factors, but guarantees separation of commutative and noncommutative factors.

as_real_imag(deep=True, **hints)

Performs complex expansion on ‘self’ and returns a tuple containing collected both real and imaginary parts. This method cannot be confused with re() and im() functions, which does not perform complex expansion at evaluation.

However it is possible to expand both re() and im() functions and get exactly the same results as with a single call to this function.

>>> from sympy import symbols, I
>>> x, y = symbols('x,y', real=True)
>>> (x + y*I).as_real_imag()
(x, y)
>>> from sympy.abc import z, w
>>> (z + w*I).as_real_imag()
(re(z) - im(w), re(w) + im(z))
as_set()

Rewrites Boolean expression in terms of real sets.

Examples

>>> from sympy import Symbol, Eq, Or, And
>>> x = Symbol('x', real=True)
>>> Eq(x, 0).as_set()
{0}
>>> (x > 0).as_set()
Interval.open(0, oo)
>>> And(-2 < x, x < 2).as_set()
Interval.open(-2, 2)
>>> Or(x < -2, 2 < x).as_set()
Union(Interval.open(-oo, -2), Interval.open(2, oo))
as_terms()

Transform an expression to a list of terms.

aseries(x=None, n=6, bound=0, hir=False)

Asymptotic Series expansion of self. This is equivalent to self.series(x, oo, n).

Parameters:
selfExpression

The expression whose series is to be expanded.

xSymbol

It is the variable of the expression to be calculated.

nValue

The value used to represent the order in terms of x**n, up to which the series is to be expanded.

hirBoolean

Set this parameter to be True to produce hierarchical series. It stops the recursion at an early level and may provide nicer and more useful results.

boundValue, Integer

Use the bound parameter to give limit on rewriting coefficients in its normalised form.

Returns:
Expr

Asymptotic series expansion of the expression.

See also

Expr.aseries

See the docstring of this function for complete details of this wrapper.

Notes

This algorithm is directly induced from the limit computational algorithm provided by Gruntz. It majorly uses the mrv and rewrite sub-routines. The overall idea of this algorithm is first to look for the most rapidly varying subexpression w of a given expression f and then expands f in a series in w. Then same thing is recursively done on the leading coefficient till we get constant coefficients.

If the most rapidly varying subexpression of a given expression f is f itself, the algorithm tries to find a normalised representation of the mrv set and rewrites f using this normalised representation.

If the expansion contains an order term, it will be either O(x ** (-n)) or O(w ** (-n)) where w belongs to the most rapidly varying expression of self.

References

[1]

Gruntz, Dominik. A new algorithm for computing asymptotic series. In: Proc. 1993 Int. Symp. Symbolic and Algebraic Computation. 1993. pp. 239-244.

[2]

Gruntz thesis - p90

Examples

>>> from sympy import sin, exp
>>> from sympy.abc import x
>>> e = sin(1/x + exp(-x)) - sin(1/x)
>>> e.aseries(x)
(1/(24*x**4) - 1/(2*x**2) + 1 + O(x**(-6), (x, oo)))*exp(-x)
>>> e.aseries(x, n=3, hir=True)
-exp(-2*x)*sin(1/x)/2 + exp(-x)*cos(1/x) + O(exp(-3*x), (x, oo))
>>> e = exp(exp(x)/(1 - 1/x))
>>> e.aseries(x)
exp(exp(x)/(1 - 1/x))
>>> e.aseries(x, bound=3) 
exp(exp(x)/x**2)*exp(exp(x)/x)*exp(-exp(x) + exp(x)/(1 - 1/x) - exp(x)/x - exp(x)/x**2)*exp(exp(x))

For rational expressions this method may return original expression without the Order term. >>> (1/x).aseries(x, n=8) 1/x

property assumptions0

Return object type assumptions.

For example:

Symbol(‘x’, real=True) Symbol(‘x’, integer=True)

are different objects. In other words, besides Python type (Symbol in this case), the initial assumptions are also forming their typeinfo.

Examples

>>> from sympy import Symbol
>>> from sympy.abc import x
>>> x.assumptions0
{'commutative': True}
>>> x = Symbol("x", positive=True)
>>> x.assumptions0
{'commutative': True, 'complex': True, 'extended_negative': False,
 'extended_nonnegative': True, 'extended_nonpositive': False,
 'extended_nonzero': True, 'extended_positive': True, 'extended_real':
 True, 'finite': True, 'hermitian': True, 'imaginary': False,
 'infinite': False, 'negative': False, 'nonnegative': True,
 'nonpositive': False, 'nonzero': True, 'positive': True, 'real':
 True, 'zero': False}
atoms(*types)

Returns the atoms that form the current object.

By default, only objects that are truly atomic and cannot be divided into smaller pieces are returned: symbols, numbers, and number symbols like I and pi. It is possible to request atoms of any type, however, as demonstrated below.

Examples

>>> from sympy import I, pi, sin
>>> from sympy.abc import x, y
>>> (1 + x + 2*sin(y + I*pi)).atoms()
{1, 2, I, pi, x, y}

If one or more types are given, the results will contain only those types of atoms.

>>> from sympy import Number, NumberSymbol, Symbol
>>> (1 + x + 2*sin(y + I*pi)).atoms(Symbol)
{x, y}
>>> (1 + x + 2*sin(y + I*pi)).atoms(Number)
{1, 2}
>>> (1 + x + 2*sin(y + I*pi)).atoms(Number, NumberSymbol)
{1, 2, pi}
>>> (1 + x + 2*sin(y + I*pi)).atoms(Number, NumberSymbol, I)
{1, 2, I, pi}

Note that I (imaginary unit) and zoo (complex infinity) are special types of number symbols and are not part of the NumberSymbol class.

The type can be given implicitly, too:

>>> (1 + x + 2*sin(y + I*pi)).atoms(x) # x is a Symbol
{x, y}

Be careful to check your assumptions when using the implicit option since S(1).is_Integer = True but type(S(1)) is One, a special type of SymPy atom, while type(S(2)) is type Integer and will find all integers in an expression:

>>> from sympy import S
>>> (1 + x + 2*sin(y + I*pi)).atoms(S(1))
{1}
>>> (1 + x + 2*sin(y + I*pi)).atoms(S(2))
{1, 2}

Finally, arguments to atoms() can select more than atomic atoms: any SymPy type (loaded in core/__init__.py) can be listed as an argument and those types of “atoms” as found in scanning the arguments of the expression recursively:

>>> from sympy import Function, Mul
>>> from sympy.core.function import AppliedUndef
>>> f = Function('f')
>>> (1 + f(x) + 2*sin(y + I*pi)).atoms(Function)
{f(x), sin(y + I*pi)}
>>> (1 + f(x) + 2*sin(y + I*pi)).atoms(AppliedUndef)
{f(x)}
>>> (1 + x + 2*sin(y + I*pi)).atoms(Mul)
{I*pi, 2*sin(y + I*pi)}
property binary_symbols

Return from the atoms of self those which are free symbols.

Not all free symbols are Symbol. Eg: IndexedBase(‘I’)[0].free_symbols

For most expressions, all symbols are free symbols. For some classes this is not true. e.g. Integrals use Symbols for the dummy variables which are bound variables, so Integral has a method to return all symbols except those. Derivative keeps track of symbols with respect to which it will perform a derivative; those are bound variables, too, so it has its own free_symbols method.

Any other method that uses bound variables should implement a free_symbols method.

cancel(*gens, **args)

See the cancel function in sympy.polys

property canonical_variables

Return a dictionary mapping any variable defined in self.bound_symbols to Symbols that do not clash with any free symbols in the expression.

Examples

>>> from sympy import Lambda
>>> from sympy.abc import x
>>> Lambda(x, 2*x).canonical_variables
{x: _0}
classmethod class_key()

Nice order of classes.

coeff(x, n=1, right=False, _first=True)

Returns the coefficient from the term(s) containing x**n. If n is zero then all terms independent of x will be returned.

See also

as_coefficient

separate the expression into a coefficient and factor

as_coeff_Add

separate the additive constant from an expression

as_coeff_Mul

separate the multiplicative constant from an expression

as_independent

separate x-dependent terms/factors from others

sympy.polys.polytools.Poly.coeff_monomial

efficiently find the single coefficient of a monomial in Poly

sympy.polys.polytools.Poly.nth

like coeff_monomial but powers of monomial terms are used

Examples

>>> from sympy import symbols
>>> from sympy.abc import x, y, z

You can select terms that have an explicit negative in front of them:

>>> (-x + 2*y).coeff(-1)
x
>>> (x - 2*y).coeff(-1)
2*y

You can select terms with no Rational coefficient:

>>> (x + 2*y).coeff(1)
x
>>> (3 + 2*x + 4*x**2).coeff(1)
0

You can select terms independent of x by making n=0; in this case expr.as_independent(x)[0] is returned (and 0 will be returned instead of None):

>>> (3 + 2*x + 4*x**2).coeff(x, 0)
3
>>> eq = ((x + 1)**3).expand() + 1
>>> eq
x**3 + 3*x**2 + 3*x + 2
>>> [eq.coeff(x, i) for i in reversed(range(4))]
[1, 3, 3, 2]
>>> eq -= 2
>>> [eq.coeff(x, i) for i in reversed(range(4))]
[1, 3, 3, 0]

You can select terms that have a numerical term in front of them:

>>> (-x - 2*y).coeff(2)
-y
>>> from sympy import sqrt
>>> (x + sqrt(2)*x).coeff(sqrt(2))
x

The matching is exact:

>>> (3 + 2*x + 4*x**2).coeff(x)
2
>>> (3 + 2*x + 4*x**2).coeff(x**2)
4
>>> (3 + 2*x + 4*x**2).coeff(x**3)
0
>>> (z*(x + y)**2).coeff((x + y)**2)
z
>>> (z*(x + y)**2).coeff(x + y)
0

In addition, no factoring is done, so 1 + z*(1 + y) is not obtained from the following:

>>> (x + z*(x + x*y)).coeff(x)
1

If such factoring is desired, factor_terms can be used first:

>>> from sympy import factor_terms
>>> factor_terms(x + z*(x + x*y)).coeff(x)
z*(y + 1) + 1
>>> n, m, o = symbols('n m o', commutative=False)
>>> n.coeff(n)
1
>>> (3*n).coeff(n)
3
>>> (n*m + m*n*m).coeff(n) # = (1 + m)*n*m
1 + m
>>> (n*m + m*n*m).coeff(n, right=True) # = (1 + m)*n*m
m

If there is more than one possible coefficient 0 is returned:

>>> (n*m + m*n).coeff(n)
0

If there is only one possible coefficient, it is returned:

>>> (n*m + x*m*n).coeff(m*n)
x
>>> (n*m + x*m*n).coeff(m*n, right=1)
1
collect(syms, func=None, evaluate=True, exact=False, distribute_order_term=True)

See the collect function in sympy.simplify

combsimp()

See the combsimp function in sympy.simplify

compare(other)

Return -1, 0, 1 if the object is smaller, equal, or greater than other.

Not in the mathematical sense. If the object is of a different type from the “other” then their classes are ordered according to the sorted_classes list.

Examples

>>> from sympy.abc import x, y
>>> x.compare(y)
-1
>>> x.compare(x)
0
>>> y.compare(x)
1
compute_leading_term(x, logx=None)

Deprecated function to compute the leading term of a series.

as_leading_term is only allowed for results of .series() This is a wrapper to compute a series first.

conjugate()

Returns the complex conjugate of ‘self’.

copy()
could_extract_minus_sign()

Return True if self has -1 as a leading factor or has more literal negative signs than positive signs in a sum, otherwise False.

Examples

>>> from sympy.abc import x, y
>>> e = x - y
>>> {i.could_extract_minus_sign() for i in (e, -e)}
{False, True}

Though the y - x is considered like -(x - y), since it is in a product without a leading factor of -1, the result is false below:

>>> (x*(y - x)).could_extract_minus_sign()
False

To put something in canonical form wrt to sign, use signsimp:

>>> from sympy import signsimp
>>> signsimp(x*(y - x))
-x*(x - y)
>>> _.could_extract_minus_sign()
True
count(query)

Count the number of matching subexpressions.

count_ops(visual=None)

Wrapper for count_ops that returns the operation count.

default_assumptions = {}
diff(*symbols, **assumptions)
dir(x, cdir)
doit(**hints)

Evaluate objects that are not evaluated by default like limits, integrals, sums and products. All objects of this kind will be evaluated recursively, unless some species were excluded via ‘hints’ or unless the ‘deep’ hint was set to ‘False’.

>>> from sympy import Integral
>>> from sympy.abc import x
>>> 2*Integral(x, x)
2*Integral(x, x)
>>> (2*Integral(x, x)).doit()
x**2
>>> (2*Integral(x, x)).doit(deep=False)
2*Integral(x, x)
dummy_eq(other, symbol=None)

Compare two expressions and handle dummy symbols.

Examples

>>> from sympy import Dummy
>>> from sympy.abc import x, y
>>> u = Dummy('u')
>>> (u**2 + 1).dummy_eq(x**2 + 1)
True
>>> (u**2 + 1) == (x**2 + 1)
False
>>> (u**2 + y).dummy_eq(x**2 + y, x)
True
>>> (u**2 + y).dummy_eq(x**2 + y, y)
False
equals(other, failing_expression=False)

Return True if self == other, False if it does not, or None. If failing_expression is True then the expression which did not simplify to a 0 will be returned instead of None.

evalf(n=15, subs=None, maxn=100, chop=False, strict=False, quad=None, verbose=False)

Evaluate the given formula to an accuracy of n digits.

Parameters:
subsdict, optional

Substitute numerical values for symbols, e.g. subs={x:3, y:1+pi}. The substitutions must be given as a dictionary.

maxnint, optional

Allow a maximum temporary working precision of maxn digits.

chopbool or number, optional

Specifies how to replace tiny real or imaginary parts in subresults by exact zeros.

When True the chop value defaults to standard precision.

Otherwise the chop value is used to determine the magnitude of “small” for purposes of chopping.

>>> from sympy import N
>>> x = 1e-4
>>> N(x, chop=True)
0.000100000000000000
>>> N(x, chop=1e-5)
0.000100000000000000
>>> N(x, chop=1e-4)
0
strictbool, optional

Raise PrecisionExhausted if any subresult fails to evaluate to full accuracy, given the available maxprec.

quadstr, optional

Choose algorithm for numerical quadrature. By default, tanh-sinh quadrature is used. For oscillatory integrals on an infinite interval, try quad='osc'.

verbosebool, optional

Print debug information.

Notes

When Floats are naively substituted into an expression, precision errors may adversely affect the result. For example, adding 1e16 (a Float) to 1 will truncate to 1e16; if 1e16 is then subtracted, the result will be 0. That is exactly what happens in the following:

>>> from sympy.abc import x, y, z
>>> values = {x: 1e16, y: 1, z: 1e16}
>>> (x + y - z).subs(values)
0

Using the subs argument for evalf is the accurate way to evaluate such an expression:

>>> (x + y - z).evalf(subs=values)
1.00000000000000
expand(deep=True, modulus=None, power_base=True, power_exp=True, mul=True, log=True, multinomial=True, basic=True, **hints)

Expand an expression using hints.

See the docstring of the expand() function in sympy.core.function for more information.

property expr_free_symbols

Like free_symbols, but returns the free symbols only if they are contained in an expression node.

Examples

>>> from sympy.abc import x, y
>>> (x + y).expr_free_symbols 
{x, y}

If the expression is contained in a non-expression object, do not return the free symbols. Compare:

>>> from sympy import Tuple
>>> t = Tuple(x + y)
>>> t.expr_free_symbols 
set()
>>> t.free_symbols
{x, y}
extract_additively(c)

Return self - c if it’s possible to subtract c from self and make all matching coefficients move towards zero, else return None.

Examples

>>> from sympy.abc import x, y
>>> e = 2*x + 3
>>> e.extract_additively(x + 1)
x + 2
>>> e.extract_additively(3*x)
>>> e.extract_additively(4)
>>> (y*(x + 1)).extract_additively(x + 1)
>>> ((x + 1)*(x + 2*y + 1) + 3).extract_additively(x + 1)
(x + 1)*(x + 2*y) + 3
extract_branch_factor(allow_half=False)

Try to write self as exp_polar(2*pi*I*n)*z in a nice way. Return (z, n).

>>> from sympy import exp_polar, I, pi
>>> from sympy.abc import x, y
>>> exp_polar(I*pi).extract_branch_factor()
(exp_polar(I*pi), 0)
>>> exp_polar(2*I*pi).extract_branch_factor()
(1, 1)
>>> exp_polar(-pi*I).extract_branch_factor()
(exp_polar(I*pi), -1)
>>> exp_polar(3*pi*I + x).extract_branch_factor()
(exp_polar(x + I*pi), 1)
>>> (y*exp_polar(-5*pi*I)*exp_polar(3*pi*I + 2*pi*x)).extract_branch_factor()
(y*exp_polar(2*pi*x), -1)
>>> exp_polar(-I*pi/2).extract_branch_factor()
(exp_polar(-I*pi/2), 0)

If allow_half is True, also extract exp_polar(I*pi):

>>> exp_polar(I*pi).extract_branch_factor(allow_half=True)
(1, 1/2)
>>> exp_polar(2*I*pi).extract_branch_factor(allow_half=True)
(1, 1)
>>> exp_polar(3*I*pi).extract_branch_factor(allow_half=True)
(1, 3/2)
>>> exp_polar(-I*pi).extract_branch_factor(allow_half=True)
(1, -1/2)
extract_multiplicatively(c)

Return None if it’s not possible to make self in the form c * something in a nice way, i.e. preserving the properties of arguments of self.

Examples

>>> from sympy import symbols, Rational
>>> x, y = symbols('x,y', real=True)
>>> ((x*y)**3).extract_multiplicatively(x**2 * y)
x*y**2
>>> ((x*y)**3).extract_multiplicatively(x**4 * y)
>>> (2*x).extract_multiplicatively(2)
x
>>> (2*x).extract_multiplicatively(3)
>>> (Rational(1, 2)*x).extract_multiplicatively(3)
x/6
factor(*gens, **args)

See the factor() function in sympy.polys.polytools

find(query, group=False)

Find all subexpressions matching a query.

property formula

Return a Formula with only terms=[self].

fourier_series(limits=None)

Compute fourier sine/cosine series of self.

See the docstring of the fourier_series() in sympy.series.fourier for more information.

fps(x=None, x0=0, dir=1, hyper=True, order=4, rational=True, full=False)

Compute formal power power series of self.

See the docstring of the fps() function in sympy.series.formal for more information.

property free_symbols

Return from the atoms of self those which are free symbols.

Not all free symbols are Symbol. Eg: IndexedBase(‘I’)[0].free_symbols

For most expressions, all symbols are free symbols. For some classes this is not true. e.g. Integrals use Symbols for the dummy variables which are bound variables, so Integral has a method to return all symbols except those. Derivative keeps track of symbols with respect to which it will perform a derivative; those are bound variables, too, so it has its own free_symbols method.

Any other method that uses bound variables should implement a free_symbols method.

classmethod fromiter(args, **assumptions)

Create a new object from an iterable.

This is a convenience function that allows one to create objects from any iterable, without having to convert to a list or tuple first.

Examples

>>> from sympy import Tuple
>>> Tuple.fromiter(i for i in range(5))
(0, 1, 2, 3, 4)
property func

The top-level function in an expression.

The following should hold for all objects:

>> x == x.func(*x.args)

Examples

>>> from sympy.abc import x
>>> a = 2*x
>>> a.func
<class 'sympy.core.mul.Mul'>
>>> a.args
(2, x)
>>> a.func(*a.args)
2*x
>>> a == a.func(*a.args)
True
gammasimp()

See the gammasimp function in sympy.simplify

getO()

Returns the additive O(..) symbol if there is one, else None.

getn()

Returns the order of the expression.

Examples

>>> from sympy import O
>>> from sympy.abc import x
>>> (1 + x + O(x**2)).getn()
2
>>> (1 + x).getn()
has(*patterns)

Test whether any subexpression matches any of the patterns.

Examples

>>> from sympy import sin
>>> from sympy.abc import x, y, z
>>> (x**2 + sin(x*y)).has(z)
False
>>> (x**2 + sin(x*y)).has(x, y, z)
True
>>> x.has(x)
True

Note has is a structural algorithm with no knowledge of mathematics. Consider the following half-open interval:

>>> from sympy import Interval
>>> i = Interval.Lopen(0, 5); i
Interval.Lopen(0, 5)
>>> i.args
(0, 5, True, False)
>>> i.has(4)  # there is no "4" in the arguments
False
>>> i.has(0)  # there *is* a "0" in the arguments
True

Instead, use contains to determine whether a number is in the interval or not:

>>> i.contains(4)
True
>>> i.contains(0)
False

Note that expr.has(*patterns) is exactly equivalent to any(expr.has(p) for p in patterns). In particular, False is returned when the list of patterns is empty.

>>> x.has()
False
has_free(*patterns)

Return True if self has object(s) x as a free expression else False.

Examples

>>> from sympy import Integral, Function
>>> from sympy.abc import x, y
>>> f = Function('f')
>>> g = Function('g')
>>> expr = Integral(f(x), (f(x), 1, g(y)))
>>> expr.free_symbols
{y}
>>> expr.has_free(g(y))
True
>>> expr.has_free(*(x, f(x)))
False

This works for subexpressions and types, too:

>>> expr.has_free(g)
True
>>> (x + y + 1).has_free(y + 1)
True
has_xfree(s: set[Basic])

Return True if self has any of the patterns in s as a free argument, else False. This is like Basic.has_free but this will only report exact argument matches.

Examples

>>> from sympy import Function
>>> from sympy.abc import x, y
>>> f = Function('f')
>>> f(x).has_xfree({f})
False
>>> f(x).has_xfree({f(x)})
True
>>> f(x + 1).has_xfree({x})
True
>>> f(x + 1).has_xfree({x + 1})
True
>>> f(x + y + 1).has_xfree({x + 1})
False
integrate(*args, **kwargs)

See the integrate function in sympy.integrals

invert(g, *gens, **args)

Return the multiplicative inverse of self mod g where self (and g) may be symbolic expressions).

See also

sympy.core.numbers.mod_inverse, sympy.polys.polytools.invert
is_Add = False
is_AlgebraicNumber = False
is_Atom = True
is_Boolean = False
is_Derivative = False
is_Dummy = False
is_Equality = False
is_Float = False
is_Function = False
is_Indexed = False
is_Integer = False
is_MatAdd = False
is_MatMul = False
is_Matrix = False
is_Mul = False
is_Not = False
is_Number = False
is_NumberSymbol = False
is_Order = False
is_Piecewise = False
is_Point = False
is_Poly = False
is_Pow = False
is_Rational = False
is_Relational = False
is_Symbol = True
is_Vector = False
is_Wild = False
property is_algebraic
is_algebraic_expr(*syms)

This tests whether a given expression is algebraic or not, in the given symbols, syms. When syms is not given, all free symbols will be used. The rational function does not have to be in expanded or in any kind of canonical form.

This function returns False for expressions that are “algebraic expressions” with symbolic exponents. This is a simple extension to the is_rational_function, including rational exponentiation.

References

Examples

>>> from sympy import Symbol, sqrt
>>> x = Symbol('x', real=True)
>>> sqrt(1 + x).is_rational_function()
False
>>> sqrt(1 + x).is_algebraic_expr()
True

This function does not attempt any nontrivial simplifications that may result in an expression that does not appear to be an algebraic expression to become one.

>>> from sympy import exp, factor
>>> a = sqrt(exp(x)**2 + 2*exp(x) + 1)/(exp(x) + 1)
>>> a.is_algebraic_expr(x)
False
>>> factor(a).is_algebraic_expr()
True
property is_antihermitian
property is_commutative
is_comparable = False
property is_complex
property is_composite
is_constant(*wrt, **flags)

Return True if self is constant, False if not, or None if the constancy could not be determined conclusively.

Examples

>>> from sympy import cos, sin, Sum, S, pi
>>> from sympy.abc import a, n, x, y
>>> x.is_constant()
False
>>> S(2).is_constant()
True
>>> Sum(x, (x, 1, 10)).is_constant()
True
>>> Sum(x, (x, 1, n)).is_constant()
False
>>> Sum(x, (x, 1, n)).is_constant(y)
True
>>> Sum(x, (x, 1, n)).is_constant(n)
False
>>> Sum(x, (x, 1, n)).is_constant(x)
True
>>> eq = a*cos(x)**2 + a*sin(x)**2 - a
>>> eq.is_constant()
True
>>> eq.subs({x: pi, a: 2}) == eq.subs({x: pi, a: 3}) == 0
True
>>> (0**x).is_constant()
False
>>> x.is_constant()
False
>>> (x**x).is_constant()
False
>>> one = cos(x)**2 + sin(x)**2
>>> one.is_constant()
True
>>> ((one - 1)**(x + 1)).is_constant() in (True, False) # could be 0 or 1
True
property is_even
property is_extended_negative
property is_extended_nonnegative
property is_extended_nonpositive
property is_extended_nonzero
property is_extended_positive
property is_extended_real
property is_finite
property is_hermitian
is_hypergeometric(k)
property is_imaginary
property is_infinite
property is_integer
property is_irrational
is_meromorphic(x, a)

This tests whether an expression is meromorphic as a function of the given symbol x at the point a.

This method is intended as a quick test that will return None if no decision can be made without simplification or more detailed analysis.

Examples

>>> from sympy import zoo, log, sin, sqrt
>>> from sympy.abc import x
>>> f = 1/x**2 + 1 - 2*x**3
>>> f.is_meromorphic(x, 0)
True
>>> f.is_meromorphic(x, 1)
True
>>> f.is_meromorphic(x, zoo)
True
>>> g = x**log(3)
>>> g.is_meromorphic(x, 0)
False
>>> g.is_meromorphic(x, 1)
True
>>> g.is_meromorphic(x, zoo)
False
>>> h = sin(1/x)*x**2
>>> h.is_meromorphic(x, 0)
False
>>> h.is_meromorphic(x, 1)
True
>>> h.is_meromorphic(x, zoo)
True

Multivalued functions are considered meromorphic when their branches are meromorphic. Thus most functions are meromorphic everywhere except at essential singularities and branch points. In particular, they will be meromorphic also on branch cuts except at their endpoints.

>>> log(x).is_meromorphic(x, -1)
True
>>> log(x).is_meromorphic(x, 0)
False
>>> sqrt(x).is_meromorphic(x, -1)
True
>>> sqrt(x).is_meromorphic(x, 0)
False
property is_negative
property is_noninteger
property is_nonnegative
property is_nonpositive
property is_nonzero
is_number = False
property is_odd
property is_polar
is_polynomial(*syms)

Return True if self is a polynomial in syms and False otherwise.

This checks if self is an exact polynomial in syms. This function returns False for expressions that are “polynomials” with symbolic exponents. Thus, you should be able to apply polynomial algorithms to expressions for which this returns True, and Poly(expr, *syms) should work if and only if expr.is_polynomial(*syms) returns True. The polynomial does not have to be in expanded form. If no symbols are given, all free symbols in the expression will be used.

This is not part of the assumptions system. You cannot do Symbol(‘z’, polynomial=True).

Examples

>>> from sympy import Symbol, Function
>>> x = Symbol('x')
>>> ((x**2 + 1)**4).is_polynomial(x)
True
>>> ((x**2 + 1)**4).is_polynomial()
True
>>> (2**x + 1).is_polynomial(x)
False
>>> (2**x + 1).is_polynomial(2**x)
True
>>> f = Function('f')
>>> (f(x) + 1).is_polynomial(x)
False
>>> (f(x) + 1).is_polynomial(f(x))
True
>>> (1/f(x) + 1).is_polynomial(f(x))
False
>>> n = Symbol('n', nonnegative=True, integer=True)
>>> (x**n + 1).is_polynomial(x)
False

This function does not attempt any nontrivial simplifications that may result in an expression that does not appear to be a polynomial to become one.

>>> from sympy import sqrt, factor, cancel
>>> y = Symbol('y', positive=True)
>>> a = sqrt(y**2 + 2*y + 1)
>>> a.is_polynomial(y)
False
>>> factor(a)
y + 1
>>> factor(a).is_polynomial(y)
True
>>> b = (y**2 + 2*y + 1)/(y + 1)
>>> b.is_polynomial(y)
False
>>> cancel(b)
y + 1
>>> cancel(b).is_polynomial(y)
True

See also .is_rational_function()

property is_positive
property is_prime
property is_rational
is_rational_function(*syms)

Test whether function is a ratio of two polynomials in the given symbols, syms. When syms is not given, all free symbols will be used. The rational function does not have to be in expanded or in any kind of canonical form.

This function returns False for expressions that are “rational functions” with symbolic exponents. Thus, you should be able to call .as_numer_denom() and apply polynomial algorithms to the result for expressions for which this returns True.

This is not part of the assumptions system. You cannot do Symbol(‘z’, rational_function=True).

Examples

>>> from sympy import Symbol, sin
>>> from sympy.abc import x, y
>>> (x/y).is_rational_function()
True
>>> (x**2).is_rational_function()
True
>>> (x/sin(y)).is_rational_function(y)
False
>>> n = Symbol('n', integer=True)
>>> (x**n + 1).is_rational_function(x)
False

This function does not attempt any nontrivial simplifications that may result in an expression that does not appear to be a rational function to become one.

>>> from sympy import sqrt, factor
>>> y = Symbol('y', positive=True)
>>> a = sqrt(y**2 + 2*y + 1)/y
>>> a.is_rational_function(y)
False
>>> factor(a)
(y + 1)/y
>>> factor(a).is_rational_function(y)
True

See also is_algebraic_expr().

property is_real
is_scalar = True
is_symbol = True
property is_transcendental
property is_zero
property kind

Default kind for all SymPy object. If the kind is not defined for the object, or if the object cannot infer the kind from its arguments, this will be returned.

Examples

>>> from sympy import Expr
>>> Expr().kind
UndefinedKind
leadterm(x, logx=None, cdir=0)

Returns the leading term a*x**b as a tuple (a, b).

Examples

>>> from sympy.abc import x
>>> (1+x+x**2).leadterm(x)
(1, 0)
>>> (1/x**2+x+x**2).leadterm(x)
(1, -2)
limit(x, xlim, dir='+')

Compute limit x->xlim.

lseries(x=None, x0=0, dir='+', logx=None, cdir=0)

Wrapper for series yielding an iterator of the terms of the series.

Note: an infinite series will yield an infinite iterator. The following, for exaxmple, will never terminate. It will just keep printing terms of the sin(x) series:

for term in sin(x).lseries(x):
    print term

The advantage of lseries() over nseries() is that many times you are just interested in the next term in the series (i.e. the first term for example), but you do not know how many you should ask for in nseries() using the “n” parameter.

See also nseries().

match(pattern, old=False)

Pattern matching.

Wild symbols match all.

Return None when expression (self) does not match with pattern. Otherwise return a dictionary such that:

pattern.xreplace(self.match(pattern)) == self

Examples

>>> from sympy import Wild, Sum
>>> from sympy.abc import x, y
>>> p = Wild("p")
>>> q = Wild("q")
>>> r = Wild("r")
>>> e = (x+y)**(x+y)
>>> e.match(p**p)
{p_: x + y}
>>> e.match(p**q)
{p_: x + y, q_: x + y}
>>> e = (2*x)**2
>>> e.match(p*q**r)
{p_: 4, q_: x, r_: 2}
>>> (p*q**r).xreplace(e.match(p*q**r))
4*x**2

Structurally bound symbols are ignored during matching:

>>> Sum(x, (x, 1, 2)).match(Sum(y, (y, 1, p)))
{p_: 2}

But they can be identified if desired:

>>> Sum(x, (x, 1, 2)).match(Sum(q, (q, 1, p)))
{p_: 2, q_: x}

The old flag will give the old-style pattern matching where expressions and patterns are essentially solved to give the match. Both of the following give None unless old=True:

>>> (x - 2).match(p - x, old=True)
{p_: 2*x - 2}
>>> (2/x).match(p*x, old=True)
{p_: 2/x**2}
matches(expr, repl_dict=None, old=False)

Helper method for match() that looks for a match between Wild symbols in self and expressions in expr.

Examples

>>> from sympy import symbols, Wild, Basic
>>> a, b, c = symbols('a b c')
>>> x = Wild('x')
>>> Basic(a + x, x).matches(Basic(a + b, c)) is None
True
>>> Basic(a + x, x).matches(Basic(a + b + c, b + c))
{x_: b + c}
n(n=15, subs=None, maxn=100, chop=False, strict=False, quad=None, verbose=False)

Evaluate the given formula to an accuracy of n digits.

Parameters:
subsdict, optional

Substitute numerical values for symbols, e.g. subs={x:3, y:1+pi}. The substitutions must be given as a dictionary.

maxnint, optional

Allow a maximum temporary working precision of maxn digits.

chopbool or number, optional

Specifies how to replace tiny real or imaginary parts in subresults by exact zeros.

When True the chop value defaults to standard precision.

Otherwise the chop value is used to determine the magnitude of “small” for purposes of chopping.

>>> from sympy import N
>>> x = 1e-4
>>> N(x, chop=True)
0.000100000000000000
>>> N(x, chop=1e-5)
0.000100000000000000
>>> N(x, chop=1e-4)
0
strictbool, optional

Raise PrecisionExhausted if any subresult fails to evaluate to full accuracy, given the available maxprec.

quadstr, optional

Choose algorithm for numerical quadrature. By default, tanh-sinh quadrature is used. For oscillatory integrals on an infinite interval, try quad='osc'.

verbosebool, optional

Print debug information.

Notes

When Floats are naively substituted into an expression, precision errors may adversely affect the result. For example, adding 1e16 (a Float) to 1 will truncate to 1e16; if 1e16 is then subtracted, the result will be 0. That is exactly what happens in the following:

>>> from sympy.abc import x, y, z
>>> values = {x: 1e16, y: 1, z: 1e16}
>>> (x + y - z).subs(values)
0

Using the subs argument for evalf is the accurate way to evaluate such an expression:

>>> (x + y - z).evalf(subs=values)
1.00000000000000
name: str
normal()

Return the expression as a fraction.

expression -> a/b

See also

as_numer_denom

return (a, b) instead of a/b

nseries(x=None, x0=0, n=6, dir='+', logx=None, cdir=0)

Wrapper to _eval_nseries if assumptions allow, else to series.

If x is given, x0 is 0, dir=’+’, and self has x, then _eval_nseries is called. This calculates “n” terms in the innermost expressions and then builds up the final series just by “cross-multiplying” everything out.

The optional logx parameter can be used to replace any log(x) in the returned series with a symbolic value to avoid evaluating log(x) at 0. A symbol to use in place of log(x) should be provided.

Advantage – it’s fast, because we do not have to determine how many terms we need to calculate in advance.

Disadvantage – you may end up with less terms than you may have expected, but the O(x**n) term appended will always be correct and so the result, though perhaps shorter, will also be correct.

If any of those assumptions is not met, this is treated like a wrapper to series which will try harder to return the correct number of terms.

See also lseries().

Examples

>>> from sympy import sin, log, Symbol
>>> from sympy.abc import x, y
>>> sin(x).nseries(x, 0, 6)
x - x**3/6 + x**5/120 + O(x**6)
>>> log(x+1).nseries(x, 0, 5)
x - x**2/2 + x**3/3 - x**4/4 + O(x**5)

Handling of the logx parameter — in the following example the expansion fails since sin does not have an asymptotic expansion at -oo (the limit of log(x) as x approaches 0):

>>> e = sin(log(x))
>>> e.nseries(x, 0, 6)
Traceback (most recent call last):
...
PoleError: ...
...
>>> logx = Symbol('logx')
>>> e.nseries(x, 0, 6, logx=logx)
sin(logx)

In the following example, the expansion works but only returns self unless the logx parameter is used:

>>> e = x**y
>>> e.nseries(x, 0, 2)
x**y
>>> e.nseries(x, 0, 2, logx=logx)
exp(logx*y)
nsimplify(constants=(), tolerance=None, full=False)

See the nsimplify function in sympy.simplify

powsimp(*args, **kwargs)

See the powsimp function in sympy.simplify

primitive()

Return the positive Rational that can be extracted non-recursively from every term of self (i.e., self is treated like an Add). This is like the as_coeff_Mul() method but primitive always extracts a positive Rational (never a negative or a Float).

Examples

>>> from sympy.abc import x
>>> (3*(x + 1)**2).primitive()
(3, (x + 1)**2)
>>> a = (6*x + 2); a.primitive()
(2, 3*x + 1)
>>> b = (x/2 + 3); b.primitive()
(1/2, x + 6)
>>> (a*b).primitive() == (1, a*b)
True
radsimp(**kwargs)

See the radsimp function in sympy.simplify

ratsimp()

See the ratsimp function in sympy.simplify

rcall(*args)

Apply on the argument recursively through the expression tree.

This method is used to simulate a common abuse of notation for operators. For instance, in SymPy the following will not work:

(x+Lambda(y, 2*y))(z) == x+2*z,

however, you can use:

>>> from sympy import Lambda
>>> from sympy.abc import x, y, z
>>> (x + Lambda(y, 2*y)).rcall(z)
x + 2*z
refine(assumption=True)

See the refine function in sympy.assumptions

removeO()

Removes the additive O(..) symbol if there is one

replace(query, value, map=False, simultaneous=True, exact=None)

Replace matching subexpressions of self with value.

If map = True then also return the mapping {old: new} where old was a sub-expression found with query and new is the replacement value for it. If the expression itself does not match the query, then the returned value will be self.xreplace(map) otherwise it should be self.subs(ordered(map.items())).

Traverses an expression tree and performs replacement of matching subexpressions from the bottom to the top of the tree. The default approach is to do the replacement in a simultaneous fashion so changes made are targeted only once. If this is not desired or causes problems, simultaneous can be set to False.

In addition, if an expression containing more than one Wild symbol is being used to match subexpressions and the exact flag is None it will be set to True so the match will only succeed if all non-zero values are received for each Wild that appears in the match pattern. Setting this to False accepts a match of 0; while setting it True accepts all matches that have a 0 in them. See example below for cautions.

The list of possible combinations of queries and replacement values is listed below:

See also

subs

substitution of subexpressions as defined by the objects themselves.

xreplace

exact node replacement in expr tree; also capable of using matching rules

Examples

Initial setup

>>> from sympy import log, sin, cos, tan, Wild, Mul, Add
>>> from sympy.abc import x, y
>>> f = log(sin(x)) + tan(sin(x**2))
1.1. type -> type

obj.replace(type, newtype)

When object of type type is found, replace it with the result of passing its argument(s) to newtype.

>>> f.replace(sin, cos)
log(cos(x)) + tan(cos(x**2))
>>> sin(x).replace(sin, cos, map=True)
(cos(x), {sin(x): cos(x)})
>>> (x*y).replace(Mul, Add)
x + y
1.2. type -> func

obj.replace(type, func)

When object of type type is found, apply func to its argument(s). func must be written to handle the number of arguments of type.

>>> f.replace(sin, lambda arg: sin(2*arg))
log(sin(2*x)) + tan(sin(2*x**2))
>>> (x*y).replace(Mul, lambda *args: sin(2*Mul(*args)))
sin(2*x*y)
2.1. pattern -> expr

obj.replace(pattern(wild), expr(wild))

Replace subexpressions matching pattern with the expression written in terms of the Wild symbols in pattern.

>>> a, b = map(Wild, 'ab')
>>> f.replace(sin(a), tan(a))
log(tan(x)) + tan(tan(x**2))
>>> f.replace(sin(a), tan(a/2))
log(tan(x/2)) + tan(tan(x**2/2))
>>> f.replace(sin(a), a)
log(x) + tan(x**2)
>>> (x*y).replace(a*x, a)
y

Matching is exact by default when more than one Wild symbol is used: matching fails unless the match gives non-zero values for all Wild symbols:

>>> (2*x + y).replace(a*x + b, b - a)
y - 2
>>> (2*x).replace(a*x + b, b - a)
2*x

When set to False, the results may be non-intuitive:

>>> (2*x).replace(a*x + b, b - a, exact=False)
2/x
2.2. pattern -> func

obj.replace(pattern(wild), lambda wild: expr(wild))

All behavior is the same as in 2.1 but now a function in terms of pattern variables is used rather than an expression:

>>> f.replace(sin(a), lambda a: sin(2*a))
log(sin(2*x)) + tan(sin(2*x**2))
3.1. func -> func

obj.replace(filter, func)

Replace subexpression e with func(e) if filter(e) is True.

>>> g = 2*sin(x**3)
>>> g.replace(lambda expr: expr.is_Number, lambda expr: expr**2)
4*sin(x**9)

The expression itself is also targeted by the query but is done in such a fashion that changes are not made twice.

>>> e = x*(x*y + 1)
>>> e.replace(lambda x: x.is_Mul, lambda x: 2*x)
2*x*(2*x*y + 1)

When matching a single symbol, exact will default to True, but this may or may not be the behavior that is desired:

Here, we want exact=False:

>>> from sympy import Function
>>> f = Function('f')
>>> e = f(1) + f(0)
>>> q = f(a), lambda a: f(a + 1)
>>> e.replace(*q, exact=False)
f(1) + f(2)
>>> e.replace(*q, exact=True)
f(0) + f(2)

But here, the nature of matching makes selecting the right setting tricky:

>>> e = x**(1 + y)
>>> (x**(1 + y)).replace(x**(1 + a), lambda a: x**-a, exact=False)
x
>>> (x**(1 + y)).replace(x**(1 + a), lambda a: x**-a, exact=True)
x**(-x - y + 1)
>>> (x**y).replace(x**(1 + a), lambda a: x**-a, exact=False)
x
>>> (x**y).replace(x**(1 + a), lambda a: x**-a, exact=True)
x**(1 - y)

It is probably better to use a different form of the query that describes the target expression more precisely:

>>> (1 + x**(1 + y)).replace(
... lambda x: x.is_Pow and x.exp.is_Add and x.exp.args[0] == 1,
... lambda x: x.base**(1 - (x.exp - 1)))
...
x**(1 - y) + 1
rewrite(*args, deep=True, **hints)

Rewrite self using a defined rule.

Rewriting transforms an expression to another, which is mathematically equivalent but structurally different. For example you can rewrite trigonometric functions as complex exponentials or combinatorial functions as gamma function.

This method takes a pattern and a rule as positional arguments. pattern is optional parameter which defines the types of expressions that will be transformed. If it is not passed, all possible expressions will be rewritten. rule defines how the expression will be rewritten.

Parameters:
argsExpr

A rule, or pattern and rule. - pattern is a type or an iterable of types. - rule can be any object.

deepbool, optional

If True, subexpressions are recursively transformed. Default is True.

Examples

If pattern is unspecified, all possible expressions are transformed.

>>> from sympy import cos, sin, exp, I
>>> from sympy.abc import x
>>> expr = cos(x) + I*sin(x)
>>> expr.rewrite(exp)
exp(I*x)

Pattern can be a type or an iterable of types.

>>> expr.rewrite(sin, exp)
exp(I*x)/2 + cos(x) - exp(-I*x)/2
>>> expr.rewrite([cos,], exp)
exp(I*x)/2 + I*sin(x) + exp(-I*x)/2
>>> expr.rewrite([cos, sin], exp)
exp(I*x)

Rewriting behavior can be implemented by defining _eval_rewrite() method.

>>> from sympy import Expr, sqrt, pi
>>> class MySin(Expr):
...     def _eval_rewrite(self, rule, args, **hints):
...         x, = args
...         if rule == cos:
...             return cos(pi/2 - x, evaluate=False)
...         if rule == sqrt:
...             return sqrt(1 - cos(x)**2)
>>> MySin(MySin(x)).rewrite(cos)
cos(-cos(-x + pi/2) + pi/2)
>>> MySin(x).rewrite(sqrt)
sqrt(1 - cos(x)**2)

Defining _eval_rewrite_as_[...]() method is supported for backwards compatibility reason. This may be removed in the future and using it is discouraged.

>>> class MySin(Expr):
...     def _eval_rewrite_as_cos(self, *args, **hints):
...         x, = args
...         return cos(pi/2 - x, evaluate=False)
>>> MySin(x).rewrite(cos)
cos(-x + pi/2)
round(n=None)

Return x rounded to the given decimal place.

If a complex number would results, apply round to the real and imaginary components of the number.

Notes

The Python round function uses the SymPy round method so it will always return a SymPy number (not a Python float or int):

>>> isinstance(round(S(123), -2), Number)
True

Examples

>>> from sympy import pi, E, I, S, Number
>>> pi.round()
3
>>> pi.round(2)
3.14
>>> (2*pi + E*I).round()
6 + 3*I

The round method has a chopping effect:

>>> (2*pi + I/10).round()
6
>>> (pi/10 + 2*I).round()
2*I
>>> (pi/10 + E*I).round(2)
0.31 + 2.72*I
separate(deep=False, force=False)

See the separate function in sympy.simplify

series(x=None, x0=0, n=6, dir='+', logx=None, cdir=0)

Series expansion of “self” around x = x0 yielding either terms of the series one by one (the lazy series given when n=None), else all the terms at once when n != None.

Returns the series expansion of “self” around the point x = x0 with respect to x up to O((x - x0)**n, x, x0) (default n is 6).

If x=None and self is univariate, the univariate symbol will be supplied, otherwise an error will be raised.

Parameters:
exprExpression

The expression whose series is to be expanded.

xSymbol

It is the variable of the expression to be calculated.

x0Value

The value around which x is calculated. Can be any value from -oo to oo.

nValue

The value used to represent the order in terms of x**n, up to which the series is to be expanded.

dirString, optional

The series-expansion can be bi-directional. If dir="+", then (x->x0+). If dir="-", then (x->x0-). For infinite ``x0 (oo or -oo), the dir argument is determined from the direction of the infinity (i.e., dir="-" for oo).

logxoptional

It is used to replace any log(x) in the returned series with a symbolic value rather than evaluating the actual value.

cdiroptional

It stands for complex direction, and indicates the direction from which the expansion needs to be evaluated.

Returns:
ExprExpression

Series expansion of the expression about x0

Raises:
TypeError

If “n” and “x0” are infinity objects

PoleError

If “x0” is an infinity object

Examples

>>> from sympy import cos, exp, tan
>>> from sympy.abc import x, y
>>> cos(x).series()
1 - x**2/2 + x**4/24 + O(x**6)
>>> cos(x).series(n=4)
1 - x**2/2 + O(x**4)
>>> cos(x).series(x, x0=1, n=2)
cos(1) - (x - 1)*sin(1) + O((x - 1)**2, (x, 1))
>>> e = cos(x + exp(y))
>>> e.series(y, n=2)
cos(x + 1) - y*sin(x + 1) + O(y**2)
>>> e.series(x, n=2)
cos(exp(y)) - x*sin(exp(y)) + O(x**2)

If n=None then a generator of the series terms will be returned.

>>> term=cos(x).series(n=None)
>>> [next(term) for i in range(2)]
[1, -x**2/2]

For dir=+ (default) the series is calculated from the right and for dir=- the series from the left. For smooth functions this flag will not alter the results.

>>> abs(x).series(dir="+")
x
>>> abs(x).series(dir="-")
-x
>>> f = tan(x)
>>> f.series(x, 2, 6, "+")
tan(2) + (1 + tan(2)**2)*(x - 2) + (x - 2)**2*(tan(2)**3 + tan(2)) +
(x - 2)**3*(1/3 + 4*tan(2)**2/3 + tan(2)**4) + (x - 2)**4*(tan(2)**5 +
5*tan(2)**3/3 + 2*tan(2)/3) + (x - 2)**5*(2/15 + 17*tan(2)**2/15 +
2*tan(2)**4 + tan(2)**6) + O((x - 2)**6, (x, 2))
>>> f.series(x, 2, 3, "-")
tan(2) + (2 - x)*(-tan(2)**2 - 1) + (2 - x)**2*(tan(2)**3 + tan(2))
+ O((x - 2)**3, (x, 2))

For rational expressions this method may return original expression without the Order term. >>> (1/x).series(x, n=8) 1/x

simplify(**kwargs)

See the simplify function in sympy.simplify

sort_key(order=None)

Return a sort key.

Examples

>>> from sympy import S, I
>>> sorted([S(1)/2, I, -I], key=lambda x: x.sort_key())
[1/2, -I, I]
>>> S("[x, 1/x, 1/x**2, x**2, x**(1/2), x**(1/4), x**(3/2)]")
[x, 1/x, x**(-2), x**2, sqrt(x), x**(1/4), x**(3/2)]
>>> sorted(_, key=lambda x: x.sort_key())
[x**(-2), 1/x, x**(1/4), sqrt(x), x, x**(3/2), x**2]
subs(*args, **kwargs)

Substitutes old for new in an expression after sympifying args.

args is either:
  • two arguments, e.g. foo.subs(old, new)

  • one iterable argument, e.g. foo.subs(iterable). The iterable may be
    o an iterable container with (old, new) pairs. In this case the

    replacements are processed in the order given with successive patterns possibly affecting replacements already made.

    o a dict or set whose key/value items correspond to old/new pairs.

    In this case the old/new pairs will be sorted by op count and in case of a tie, by number of args and the default_sort_key. The resulting sorted list is then processed as an iterable container (see previous).

If the keyword simultaneous is True, the subexpressions will not be evaluated until all the substitutions have been made.

See also

replace

replacement capable of doing wildcard-like matching, parsing of match, and conditional replacements

xreplace

exact node replacement in expr tree; also capable of using matching rules

sympy.core.evalf.EvalfMixin.evalf

calculates the given formula to a desired level of precision

Examples

>>> from sympy import pi, exp, limit, oo
>>> from sympy.abc import x, y
>>> (1 + x*y).subs(x, pi)
pi*y + 1
>>> (1 + x*y).subs({x:pi, y:2})
1 + 2*pi
>>> (1 + x*y).subs([(x, pi), (y, 2)])
1 + 2*pi
>>> reps = [(y, x**2), (x, 2)]
>>> (x + y).subs(reps)
6
>>> (x + y).subs(reversed(reps))
x**2 + 2
>>> (x**2 + x**4).subs(x**2, y)
y**2 + y

To replace only the x**2 but not the x**4, use xreplace:

>>> (x**2 + x**4).xreplace({x**2: y})
x**4 + y

To delay evaluation until all substitutions have been made, set the keyword simultaneous to True:

>>> (x/y).subs([(x, 0), (y, 0)])
0
>>> (x/y).subs([(x, 0), (y, 0)], simultaneous=True)
nan

This has the added feature of not allowing subsequent substitutions to affect those already made:

>>> ((x + y)/y).subs({x + y: y, y: x + y})
1
>>> ((x + y)/y).subs({x + y: y, y: x + y}, simultaneous=True)
y/(x + y)

In order to obtain a canonical result, unordered iterables are sorted by count_op length, number of arguments and by the default_sort_key to break any ties. All other iterables are left unsorted.

>>> from sympy import sqrt, sin, cos
>>> from sympy.abc import a, b, c, d, e
>>> A = (sqrt(sin(2*x)), a)
>>> B = (sin(2*x), b)
>>> C = (cos(2*x), c)
>>> D = (x, d)
>>> E = (exp(x), e)
>>> expr = sqrt(sin(2*x))*sin(exp(x)*x)*cos(2*x) + sin(2*x)
>>> expr.subs(dict([A, B, C, D, E]))
a*c*sin(d*e) + b

The resulting expression represents a literal replacement of the old arguments with the new arguments. This may not reflect the limiting behavior of the expression:

>>> (x**3 - 3*x).subs({x: oo})
nan
>>> limit(x**3 - 3*x, x, oo)
oo

If the substitution will be followed by numerical evaluation, it is better to pass the substitution to evalf as

>>> (1/x).evalf(subs={x: 3.0}, n=21)
0.333333333333333333333

rather than

>>> (1/x).subs({x: 3.0}).evalf(21)
0.333333333333333314830

as the former will ensure that the desired level of precision is obtained.

taylor_term(n, x, *previous_terms)

General method for the taylor term.

This method is slow, because it differentiates n-times. Subclasses can redefine it to make it faster by using the “previous_terms”.

to_nnf(simplify=True)
together(*args, **kwargs)

See the together function in sympy.polys

transpose()
trigsimp(**args)

See the trigsimp function in sympy.simplify

xreplace(rule, hack2=False)

Replace occurrences of objects within the expression.

Parameters:
ruledict-like

Expresses a replacement rule

Returns:
xreplacethe result of the replacement

See also

replace

replacement capable of doing wildcard-like matching, parsing of match, and conditional replacements

subs

substitution of subexpressions as defined by the objects themselves.

Examples

>>> from sympy import symbols, pi, exp
>>> x, y, z = symbols('x y z')
>>> (1 + x*y).xreplace({x: pi})
pi*y + 1
>>> (1 + x*y).xreplace({x: pi, y: 2})
1 + 2*pi

Replacements occur only if an entire node in the expression tree is matched:

>>> (x*y + z).xreplace({x*y: pi})
z + pi
>>> (x*y*z).xreplace({x*y: pi})
x*y*z
>>> (2*x).xreplace({2*x: y, x: z})
y
>>> (2*2*x).xreplace({2*x: y, x: z})
4*z
>>> (x + y + 2).xreplace({x + y: 2})
x + y + 2
>>> (x + 2 + exp(x + 2)).xreplace({x + 2: y})
x + exp(y) + 2

xreplace does not differentiate between free and bound symbols. In the following, subs(x, y) would not change x since it is a bound symbol, but xreplace does:

>>> from sympy import Integral
>>> Integral(x, (x, 1, 2*x)).xreplace({x: y})
Integral(y, (y, 1, 2*y))

Trying to replace x with an expression raises an error:

>>> Integral(x, (x, 1, 2*x)).xreplace({x: 2*y}) 
ValueError: Invalid limits given: ((2*y, 1, 4*y),)

Formula

class nipy.algorithms.statistics.formula.formulae.Formula(seq, char='b')

Bases: object

A Formula is a model for a mean in a regression model.

It is often given by a sequence of sympy expressions, with the mean model being the sum of each term multiplied by a linear regression coefficient.

The expressions may depend on additional Symbol instances, giving a non-linear regression model.

__init__(seq, char='b')
Parameters:
seqsequence of sympy.Basic
charstr, optional

character for regression coefficient

property coefs

Coefficients in the linear regression formula.

design(input, param=None, return_float=False, contrasts=None)

Construct the design matrix, and optional contrast matrices.

Parameters:
inputnp.recarray

Recarray including fields needed to compute the Terms in getparams(self.design_expr).

paramNone or np.recarray

Recarray including fields that are not Terms in getparams(self.design_expr)

return_floatbool, optional

If True, return a np.float64 array rather than a np.recarray

contrastsNone or dict, optional

Contrasts. The items in this dictionary should be (str, Formula) pairs where a contrast matrix is constructed for each Formula by evaluating its design at the same parameters as self.design. If not None, then the return_float is set to True.

Returns:
des2D array

design matrix

cmatricesdict, optional

Dictionary with keys from contrasts input, and contrast matrices corresponding to des design matrix. Returned only if contrasts input is not None

property design_expr
property dtype

The dtype of the design matrix of the Formula.

static fromrec(rec, keep=[], drop=[])

Construct Formula from recarray

For fields with a string-dtype, it is assumed that these are qualtiatitve regressors, i.e. Factors.

Parameters:
rec: recarray

Recarray whose field names will be used to create a formula.

keep: []

Field names to explicitly keep, dropping all others.

drop: []

Field names to drop.

property mean

Expression for the mean, expressed as a linear combination of terms, each with dummy variables in front.

property params

The parameters in the Formula.

subs(old, new)

Perform a sympy substitution on all terms in the Formula

Returns a new instance of the same class

Parameters:
oldsympy.Basic

The expression to be changed

newsympy.Basic

The value to change it to.

Returns:
newfFormula

Examples

>>> s, t = [Term(l) for l in 'st']
>>> f, g = [sympy.Function(l) for l in 'fg']
>>> form = Formula([f(t),g(s)])
>>> newform = form.subs(g, sympy.Function('h'))
>>> newform.terms
array([f(t), h(s)], dtype=object)
>>> form.terms
array([f(t), g(s)], dtype=object)
property terms

Terms in the linear regression formula.

RandomEffects

class nipy.algorithms.statistics.formula.formulae.RandomEffects(seq, sigma=None, char='e')

Bases: Formula

Covariance matrices for common random effects analyses.

Examples

Two subjects (here named 2 and 3):

>>> subj = make_recarray([2,2,2,3,3], 's')
>>> subj_factor = Factor('s', [2,3])

By default the covariance matrix is symbolic. The display differs a little between sympy versions (hence we don’t check it in the doctests):

>>> c = RandomEffects(subj_factor.terms)
>>> c.cov(subj) 
array([[_s2_0, _s2_0, _s2_0, 0, 0],
       [_s2_0, _s2_0, _s2_0, 0, 0],
       [_s2_0, _s2_0, _s2_0, 0, 0],
       [0, 0, 0, _s2_1, _s2_1],
       [0, 0, 0, _s2_1, _s2_1]], dtype=object)

With a numeric sigma, you get a numeric array:

>>> c = RandomEffects(subj_factor.terms, sigma=np.array([[4,1],[1,6]]))
>>> c.cov(subj)
array([[ 4.,  4.,  4.,  1.,  1.],
       [ 4.,  4.,  4.,  1.,  1.],
       [ 4.,  4.,  4.,  1.,  1.],
       [ 1.,  1.,  1.,  6.,  6.],
       [ 1.,  1.,  1.,  6.,  6.]])
__init__(seq, sigma=None, char='e')

Initialize random effects instance

Parameters:
seq[sympy.Basic]
sigmandarray

Covariance of the random effects. Defaults to a diagonal with entries for each random effect.

charcharacter for regression coefficient
property coefs

Coefficients in the linear regression formula.

cov(term, param=None)

Compute the covariance matrix for some given data.

Parameters:
termnp.recarray

Recarray including fields corresponding to the Terms in getparams(self.design_expr).

paramnp.recarray

Recarray including fields that are not Terms in getparams(self.design_expr)

Returns:
Cndarray

Covariance matrix implied by design and self.sigma.

design(input, param=None, return_float=False, contrasts=None)

Construct the design matrix, and optional contrast matrices.

Parameters:
inputnp.recarray

Recarray including fields needed to compute the Terms in getparams(self.design_expr).

paramNone or np.recarray

Recarray including fields that are not Terms in getparams(self.design_expr)

return_floatbool, optional

If True, return a np.float64 array rather than a np.recarray

contrastsNone or dict, optional

Contrasts. The items in this dictionary should be (str, Formula) pairs where a contrast matrix is constructed for each Formula by evaluating its design at the same parameters as self.design. If not None, then the return_float is set to True.

Returns:
des2D array

design matrix

cmatricesdict, optional

Dictionary with keys from contrasts input, and contrast matrices corresponding to des design matrix. Returned only if contrasts input is not None

property design_expr
property dtype

The dtype of the design matrix of the Formula.

static fromrec(rec, keep=[], drop=[])

Construct Formula from recarray

For fields with a string-dtype, it is assumed that these are qualtiatitve regressors, i.e. Factors.

Parameters:
rec: recarray

Recarray whose field names will be used to create a formula.

keep: []

Field names to explicitly keep, dropping all others.

drop: []

Field names to drop.

property mean

Expression for the mean, expressed as a linear combination of terms, each with dummy variables in front.

property params

The parameters in the Formula.

subs(old, new)

Perform a sympy substitution on all terms in the Formula

Returns a new instance of the same class

Parameters:
oldsympy.Basic

The expression to be changed

newsympy.Basic

The value to change it to.

Returns:
newfFormula

Examples

>>> s, t = [Term(l) for l in 'st']
>>> f, g = [sympy.Function(l) for l in 'fg']
>>> form = Formula([f(t),g(s)])
>>> newform = form.subs(g, sympy.Function('h'))
>>> newform.terms
array([f(t), h(s)], dtype=object)
>>> form.terms
array([f(t), g(s)], dtype=object)
property terms

Terms in the linear regression formula.

Term

class nipy.algorithms.statistics.formula.formulae.Term(name, **assumptions)

Bases: Symbol

A sympy.Symbol type to represent a term an a regression model

Terms can be added to other sympy expressions with the single convention that a term plus itself returns itself.

It is meant to emulate something on the right hand side of a formula in R. In particular, its name can be the name of a field in a recarray used to create a design matrix.

>>> t = Term('x')
>>> xval = np.array([(3,),(4,),(5,)], np.dtype([('x', np.float64)]))
>>> f = t.formula
>>> d = f.design(xval)
>>> print(d.dtype.descr)
[('x', '<f8')]
>>> f.design(xval, return_float=True)
array([ 3.,  4.,  5.])
__init__(*args, **kwargs)
adjoint()
apart(x=None, **args)

See the apart function in sympy.polys

property args: tuple[Basic, ...]

Returns a tuple of arguments of ‘self’.

Notes

Never use self._args, always use self.args. Only use _args in __new__ when creating a new function. Do not override .args() from Basic (so that it is easy to change the interface in the future if needed).

Examples

>>> from sympy import cot
>>> from sympy.abc import x, y
>>> cot(x).args
(x,)
>>> cot(x).args[0]
x
>>> (x*y).args
(x, y)
>>> (x*y).args[1]
y
args_cnc(cset=False, warn=True, split_1=True)

Return [commutative factors, non-commutative factors] of self.

Examples

>>> from sympy import symbols, oo
>>> A, B = symbols('A B', commutative=0)
>>> x, y = symbols('x y')
>>> (-2*x*y).args_cnc()
[[-1, 2, x, y], []]
>>> (-2.5*x).args_cnc()
[[-1, 2.5, x], []]
>>> (-2*x*A*B*y).args_cnc()
[[-1, 2, x, y], [A, B]]
>>> (-2*x*A*B*y).args_cnc(split_1=False)
[[-2, x, y], [A, B]]
>>> (-2*x*y).args_cnc(cset=True)
[{-1, 2, x, y}, []]

The arg is always treated as a Mul:

>>> (-2 + x + A).args_cnc()
[[], [x - 2 + A]]
>>> (-oo).args_cnc() # -oo is a singleton
[[-1, oo], []]
as_base_exp() tuple[Expr, Expr]
as_coeff_Add(rational=False) tuple[Number, Expr]

Efficiently extract the coefficient of a summation.

as_coeff_Mul(rational: bool = False) tuple[Number, Expr]

Efficiently extract the coefficient of a product.

as_coeff_add(*deps) tuple[Expr, tuple[Expr, ...]]

Return the tuple (c, args) where self is written as an Add, a.

c should be a Rational added to any terms of the Add that are independent of deps.

args should be a tuple of all other terms of a; args is empty if self is a Number or if self is independent of deps (when given).

This should be used when you do not know if self is an Add or not but you want to treat self as an Add or if you want to process the individual arguments of the tail of self as an Add.

  • if you know self is an Add and want only the head, use self.args[0];

  • if you do not want to process the arguments of the tail but need the tail then use self.as_two_terms() which gives the head and tail.

  • if you want to split self into an independent and dependent parts use self.as_independent(*deps)

>>> from sympy import S
>>> from sympy.abc import x, y
>>> (S(3)).as_coeff_add()
(3, ())
>>> (3 + x).as_coeff_add()
(3, (x,))
>>> (3 + x + y).as_coeff_add(x)
(y + 3, (x,))
>>> (3 + y).as_coeff_add(x)
(y + 3, ())
as_coeff_exponent(x) tuple[Expr, Expr]

c*x**e -> c,e where x can be any symbolic expression.

as_coeff_mul(*deps, **kwargs) tuple[Expr, tuple[Expr, ...]]

Return the tuple (c, args) where self is written as a Mul, m.

c should be a Rational multiplied by any factors of the Mul that are independent of deps.

args should be a tuple of all other factors of m; args is empty if self is a Number or if self is independent of deps (when given).

This should be used when you do not know if self is a Mul or not but you want to treat self as a Mul or if you want to process the individual arguments of the tail of self as a Mul.

  • if you know self is a Mul and want only the head, use self.args[0];

  • if you do not want to process the arguments of the tail but need the tail then use self.as_two_terms() which gives the head and tail;

  • if you want to split self into an independent and dependent parts use self.as_independent(*deps)

>>> from sympy import S
>>> from sympy.abc import x, y
>>> (S(3)).as_coeff_mul()
(3, ())
>>> (3*x*y).as_coeff_mul()
(3, (x, y))
>>> (3*x*y).as_coeff_mul(x)
(3*y, (x,))
>>> (3*y).as_coeff_mul(x)
(3*y, ())
as_coefficient(expr)

Extracts symbolic coefficient at the given expression. In other words, this functions separates ‘self’ into the product of ‘expr’ and ‘expr’-free coefficient. If such separation is not possible it will return None.

See also

coeff

return sum of terms have a given factor

as_coeff_Add

separate the additive constant from an expression

as_coeff_Mul

separate the multiplicative constant from an expression

as_independent

separate x-dependent terms/factors from others

sympy.polys.polytools.Poly.coeff_monomial

efficiently find the single coefficient of a monomial in Poly

sympy.polys.polytools.Poly.nth

like coeff_monomial but powers of monomial terms are used

Examples

>>> from sympy import E, pi, sin, I, Poly
>>> from sympy.abc import x
>>> E.as_coefficient(E)
1
>>> (2*E).as_coefficient(E)
2
>>> (2*sin(E)*E).as_coefficient(E)

Two terms have E in them so a sum is returned. (If one were desiring the coefficient of the term exactly matching E then the constant from the returned expression could be selected. Or, for greater precision, a method of Poly can be used to indicate the desired term from which the coefficient is desired.)

>>> (2*E + x*E).as_coefficient(E)
x + 2
>>> _.args[0]  # just want the exact match
2
>>> p = Poly(2*E + x*E); p
Poly(x*E + 2*E, x, E, domain='ZZ')
>>> p.coeff_monomial(E)
2
>>> p.nth(0, 1)
2

Since the following cannot be written as a product containing E as a factor, None is returned. (If the coefficient 2*x is desired then the coeff method should be used.)

>>> (2*E*x + x).as_coefficient(E)
>>> (2*E*x + x).coeff(E)
2*x
>>> (E*(x + 1) + x).as_coefficient(E)
>>> (2*pi*I).as_coefficient(pi*I)
2
>>> (2*I).as_coefficient(pi*I)
as_coefficients_dict(*syms)

Return a dictionary mapping terms to their Rational coefficient. Since the dictionary is a defaultdict, inquiries about terms which were not present will return a coefficient of 0.

If symbols syms are provided, any multiplicative terms independent of them will be considered a coefficient and a regular dictionary of syms-dependent generators as keys and their corresponding coefficients as values will be returned.

Examples

>>> from sympy.abc import a, x, y
>>> (3*x + a*x + 4).as_coefficients_dict()
{1: 4, x: 3, a*x: 1}
>>> _[a]
0
>>> (3*a*x).as_coefficients_dict()
{a*x: 3}
>>> (3*a*x).as_coefficients_dict(x)
{x: 3*a}
>>> (3*a*x).as_coefficients_dict(y)
{1: 3*a*x}
as_content_primitive(radical=False, clear=True)

This method should recursively remove a Rational from all arguments and return that (content) and the new self (primitive). The content should always be positive and Mul(*foo.as_content_primitive()) == foo. The primitive need not be in canonical form and should try to preserve the underlying structure if possible (i.e. expand_mul should not be applied to self).

Examples

>>> from sympy import sqrt
>>> from sympy.abc import x, y, z
>>> eq = 2 + 2*x + 2*y*(3 + 3*y)

The as_content_primitive function is recursive and retains structure:

>>> eq.as_content_primitive()
(2, x + 3*y*(y + 1) + 1)

Integer powers will have Rationals extracted from the base:

>>> ((2 + 6*x)**2).as_content_primitive()
(4, (3*x + 1)**2)
>>> ((2 + 6*x)**(2*y)).as_content_primitive()
(1, (2*(3*x + 1))**(2*y))

Terms may end up joining once their as_content_primitives are added:

>>> ((5*(x*(1 + y)) + 2*x*(3 + 3*y))).as_content_primitive()
(11, x*(y + 1))
>>> ((3*(x*(1 + y)) + 2*x*(3 + 3*y))).as_content_primitive()
(9, x*(y + 1))
>>> ((3*(z*(1 + y)) + 2.0*x*(3 + 3*y))).as_content_primitive()
(1, 6.0*x*(y + 1) + 3*z*(y + 1))
>>> ((5*(x*(1 + y)) + 2*x*(3 + 3*y))**2).as_content_primitive()
(121, x**2*(y + 1)**2)
>>> ((x*(1 + y) + 0.4*x*(3 + 3*y))**2).as_content_primitive()
(1, 4.84*x**2*(y + 1)**2)

Radical content can also be factored out of the primitive:

>>> (2*sqrt(2) + 4*sqrt(10)).as_content_primitive(radical=True)
(2, sqrt(2)*(1 + 2*sqrt(5)))

If clear=False (default is True) then content will not be removed from an Add if it can be distributed to leave one or more terms with integer coefficients.

>>> (x/2 + y).as_content_primitive()
(1/2, x + 2*y)
>>> (x/2 + y).as_content_primitive(clear=False)
(1, x/2 + y)
as_dummy()

Return the expression with any objects having structurally bound symbols replaced with unique, canonical symbols within the object in which they appear and having only the default assumption for commutativity being True. When applied to a symbol a new symbol having only the same commutativity will be returned.

Notes

Any object that has structurally bound variables should have a property, bound_symbols that returns those symbols appearing in the object.

Examples

>>> from sympy import Integral, Symbol
>>> from sympy.abc import x
>>> r = Symbol('r', real=True)
>>> Integral(r, (r, x)).as_dummy()
Integral(_0, (_0, x))
>>> _.variables[0].is_real is None
True
>>> r.as_dummy()
_r
as_expr(*gens)

Convert a polynomial to a SymPy expression.

Examples

>>> from sympy import sin
>>> from sympy.abc import x, y
>>> f = (x**2 + x*y).as_poly(x, y)
>>> f.as_expr()
x**2 + x*y
>>> sin(x).as_expr()
sin(x)
as_independent(*deps, **hint) tuple[Expr, Expr]

A mostly naive separation of a Mul or Add into arguments that are not are dependent on deps. To obtain as complete a separation of variables as possible, use a separation method first, e.g.:

  • separatevars() to change Mul, Add and Pow (including exp) into Mul

  • .expand(mul=True) to change Add or Mul into Add

  • .expand(log=True) to change log expr into an Add

The only non-naive thing that is done here is to respect noncommutative ordering of variables and to always return (0, 0) for self of zero regardless of hints.

For nonzero self, the returned tuple (i, d) has the following interpretation:

  • i will has no variable that appears in deps

  • d will either have terms that contain variables that are in deps, or be equal to 0 (when self is an Add) or 1 (when self is a Mul)

  • if self is an Add then self = i + d

  • if self is a Mul then self = i*d

  • otherwise (self, S.One) or (S.One, self) is returned.

To force the expression to be treated as an Add, use the hint as_Add=True

See also

separatevars
expand_log
sympy.core.add.Add.as_two_terms
sympy.core.mul.Mul.as_two_terms
as_coeff_mul

Examples

– self is an Add

>>> from sympy import sin, cos, exp
>>> from sympy.abc import x, y, z
>>> (x + x*y).as_independent(x)
(0, x*y + x)
>>> (x + x*y).as_independent(y)
(x, x*y)
>>> (2*x*sin(x) + y + x + z).as_independent(x)
(y + z, 2*x*sin(x) + x)
>>> (2*x*sin(x) + y + x + z).as_independent(x, y)
(z, 2*x*sin(x) + x + y)

– self is a Mul

>>> (x*sin(x)*cos(y)).as_independent(x)
(cos(y), x*sin(x))

non-commutative terms cannot always be separated out when self is a Mul

>>> from sympy import symbols
>>> n1, n2, n3 = symbols('n1 n2 n3', commutative=False)
>>> (n1 + n1*n2).as_independent(n2)
(n1, n1*n2)
>>> (n2*n1 + n1*n2).as_independent(n2)
(0, n1*n2 + n2*n1)
>>> (n1*n2*n3).as_independent(n1)
(1, n1*n2*n3)
>>> (n1*n2*n3).as_independent(n2)
(n1, n2*n3)
>>> ((x-n1)*(x-y)).as_independent(x)
(1, (x - y)*(x - n1))

– self is anything else:

>>> (sin(x)).as_independent(x)
(1, sin(x))
>>> (sin(x)).as_independent(y)
(sin(x), 1)
>>> exp(x+y).as_independent(x)
(1, exp(x + y))

– force self to be treated as an Add:

>>> (3*x).as_independent(x, as_Add=True)
(0, 3*x)

– force self to be treated as a Mul:

>>> (3+x).as_independent(x, as_Add=False)
(1, x + 3)
>>> (-3+x).as_independent(x, as_Add=False)
(1, x - 3)

Note how the below differs from the above in making the constant on the dep term positive.

>>> (y*(-3+x)).as_independent(x)
(y, x - 3)
– use .as_independent() for true independence testing instead

of .has(). The former considers only symbols in the free symbols while the latter considers all symbols

>>> from sympy import Integral
>>> I = Integral(x, (x, 1, 2))
>>> I.has(x)
True
>>> x in I.free_symbols
False
>>> I.as_independent(x) == (I, 1)
True
>>> (I + x).as_independent(x) == (I, x)
True

Note: when trying to get independent terms, a separation method might need to be used first. In this case, it is important to keep track of what you send to this routine so you know how to interpret the returned values

>>> from sympy import separatevars, log
>>> separatevars(exp(x+y)).as_independent(x)
(exp(y), exp(x))
>>> (x + x*y).as_independent(y)
(x, x*y)
>>> separatevars(x + x*y).as_independent(y)
(x, y + 1)
>>> (x*(1 + y)).as_independent(y)
(x, y + 1)
>>> (x*(1 + y)).expand(mul=True).as_independent(y)
(x, x*y)
>>> a, b=symbols('a b', positive=True)
>>> (log(a*b).expand(log=True)).as_independent(b)
(log(a), log(b))
as_leading_term(*symbols, logx=None, cdir=0)

Returns the leading (nonzero) term of the series expansion of self.

The _eval_as_leading_term routines are used to do this, and they must always return a non-zero value.

Examples

>>> from sympy.abc import x
>>> (1 + x + x**2).as_leading_term(x)
1
>>> (1/x**2 + x + x**2).as_leading_term(x)
x**(-2)
as_numer_denom()

Return the numerator and the denominator of an expression.

expression -> a/b -> a, b

This is just a stub that should be defined by an object’s class methods to get anything else.

See also

normal

return a/b instead of (a, b)

as_ordered_factors(order=None)

Return list of ordered factors (if Mul) else [self].

as_ordered_terms(order=None, data=False)

Transform an expression to an ordered list of terms.

Examples

>>> from sympy import sin, cos
>>> from sympy.abc import x
>>> (sin(x)**2*cos(x) + sin(x)**2 + 1).as_ordered_terms()
[sin(x)**2*cos(x), sin(x)**2, 1]
as_poly(*gens, **args)

Converts self to a polynomial or returns None.

as_powers_dict()

Return self as a dictionary of factors with each factor being treated as a power. The keys are the bases of the factors and the values, the corresponding exponents. The resulting dictionary should be used with caution if the expression is a Mul and contains non- commutative factors since the order that they appeared will be lost in the dictionary.

See also

as_ordered_factors

An alternative for noncommutative applications, returning an ordered list of factors.

args_cnc

Similar to as_ordered_factors, but guarantees separation of commutative and noncommutative factors.

as_real_imag(deep=True, **hints)

Performs complex expansion on ‘self’ and returns a tuple containing collected both real and imaginary parts. This method cannot be confused with re() and im() functions, which does not perform complex expansion at evaluation.

However it is possible to expand both re() and im() functions and get exactly the same results as with a single call to this function.

>>> from sympy import symbols, I
>>> x, y = symbols('x,y', real=True)
>>> (x + y*I).as_real_imag()
(x, y)
>>> from sympy.abc import z, w
>>> (z + w*I).as_real_imag()
(re(z) - im(w), re(w) + im(z))
as_set()

Rewrites Boolean expression in terms of real sets.

Examples

>>> from sympy import Symbol, Eq, Or, And
>>> x = Symbol('x', real=True)
>>> Eq(x, 0).as_set()
{0}
>>> (x > 0).as_set()
Interval.open(0, oo)
>>> And(-2 < x, x < 2).as_set()
Interval.open(-2, 2)
>>> Or(x < -2, 2 < x).as_set()
Union(Interval.open(-oo, -2), Interval.open(2, oo))
as_terms()

Transform an expression to a list of terms.

aseries(x=None, n=6, bound=0, hir=False)

Asymptotic Series expansion of self. This is equivalent to self.series(x, oo, n).

Parameters:
selfExpression

The expression whose series is to be expanded.

xSymbol

It is the variable of the expression to be calculated.

nValue

The value used to represent the order in terms of x**n, up to which the series is to be expanded.

hirBoolean

Set this parameter to be True to produce hierarchical series. It stops the recursion at an early level and may provide nicer and more useful results.

boundValue, Integer

Use the bound parameter to give limit on rewriting coefficients in its normalised form.

Returns:
Expr

Asymptotic series expansion of the expression.

See also

Expr.aseries

See the docstring of this function for complete details of this wrapper.

Notes

This algorithm is directly induced from the limit computational algorithm provided by Gruntz. It majorly uses the mrv and rewrite sub-routines. The overall idea of this algorithm is first to look for the most rapidly varying subexpression w of a given expression f and then expands f in a series in w. Then same thing is recursively done on the leading coefficient till we get constant coefficients.

If the most rapidly varying subexpression of a given expression f is f itself, the algorithm tries to find a normalised representation of the mrv set and rewrites f using this normalised representation.

If the expansion contains an order term, it will be either O(x ** (-n)) or O(w ** (-n)) where w belongs to the most rapidly varying expression of self.

References

[1]

Gruntz, Dominik. A new algorithm for computing asymptotic series. In: Proc. 1993 Int. Symp. Symbolic and Algebraic Computation. 1993. pp. 239-244.

[2]

Gruntz thesis - p90

Examples

>>> from sympy import sin, exp
>>> from sympy.abc import x
>>> e = sin(1/x + exp(-x)) - sin(1/x)
>>> e.aseries(x)
(1/(24*x**4) - 1/(2*x**2) + 1 + O(x**(-6), (x, oo)))*exp(-x)
>>> e.aseries(x, n=3, hir=True)
-exp(-2*x)*sin(1/x)/2 + exp(-x)*cos(1/x) + O(exp(-3*x), (x, oo))
>>> e = exp(exp(x)/(1 - 1/x))
>>> e.aseries(x)
exp(exp(x)/(1 - 1/x))
>>> e.aseries(x, bound=3) 
exp(exp(x)/x**2)*exp(exp(x)/x)*exp(-exp(x) + exp(x)/(1 - 1/x) - exp(x)/x - exp(x)/x**2)*exp(exp(x))

For rational expressions this method may return original expression without the Order term. >>> (1/x).aseries(x, n=8) 1/x

property assumptions0

Return object type assumptions.

For example:

Symbol(‘x’, real=True) Symbol(‘x’, integer=True)

are different objects. In other words, besides Python type (Symbol in this case), the initial assumptions are also forming their typeinfo.

Examples

>>> from sympy import Symbol
>>> from sympy.abc import x
>>> x.assumptions0
{'commutative': True}
>>> x = Symbol("x", positive=True)
>>> x.assumptions0
{'commutative': True, 'complex': True, 'extended_negative': False,
 'extended_nonnegative': True, 'extended_nonpositive': False,
 'extended_nonzero': True, 'extended_positive': True, 'extended_real':
 True, 'finite': True, 'hermitian': True, 'imaginary': False,
 'infinite': False, 'negative': False, 'nonnegative': True,
 'nonpositive': False, 'nonzero': True, 'positive': True, 'real':
 True, 'zero': False}
atoms(*types)

Returns the atoms that form the current object.

By default, only objects that are truly atomic and cannot be divided into smaller pieces are returned: symbols, numbers, and number symbols like I and pi. It is possible to request atoms of any type, however, as demonstrated below.

Examples

>>> from sympy import I, pi, sin
>>> from sympy.abc import x, y
>>> (1 + x + 2*sin(y + I*pi)).atoms()
{1, 2, I, pi, x, y}

If one or more types are given, the results will contain only those types of atoms.

>>> from sympy import Number, NumberSymbol, Symbol
>>> (1 + x + 2*sin(y + I*pi)).atoms(Symbol)
{x, y}
>>> (1 + x + 2*sin(y + I*pi)).atoms(Number)
{1, 2}
>>> (1 + x + 2*sin(y + I*pi)).atoms(Number, NumberSymbol)
{1, 2, pi}
>>> (1 + x + 2*sin(y + I*pi)).atoms(Number, NumberSymbol, I)
{1, 2, I, pi}

Note that I (imaginary unit) and zoo (complex infinity) are special types of number symbols and are not part of the NumberSymbol class.

The type can be given implicitly, too:

>>> (1 + x + 2*sin(y + I*pi)).atoms(x) # x is a Symbol
{x, y}

Be careful to check your assumptions when using the implicit option since S(1).is_Integer = True but type(S(1)) is One, a special type of SymPy atom, while type(S(2)) is type Integer and will find all integers in an expression:

>>> from sympy import S
>>> (1 + x + 2*sin(y + I*pi)).atoms(S(1))
{1}
>>> (1 + x + 2*sin(y + I*pi)).atoms(S(2))
{1, 2}

Finally, arguments to atoms() can select more than atomic atoms: any SymPy type (loaded in core/__init__.py) can be listed as an argument and those types of “atoms” as found in scanning the arguments of the expression recursively:

>>> from sympy import Function, Mul
>>> from sympy.core.function import AppliedUndef
>>> f = Function('f')
>>> (1 + f(x) + 2*sin(y + I*pi)).atoms(Function)
{f(x), sin(y + I*pi)}
>>> (1 + f(x) + 2*sin(y + I*pi)).atoms(AppliedUndef)
{f(x)}
>>> (1 + x + 2*sin(y + I*pi)).atoms(Mul)
{I*pi, 2*sin(y + I*pi)}
property binary_symbols

Return from the atoms of self those which are free symbols.

Not all free symbols are Symbol. Eg: IndexedBase(‘I’)[0].free_symbols

For most expressions, all symbols are free symbols. For some classes this is not true. e.g. Integrals use Symbols for the dummy variables which are bound variables, so Integral has a method to return all symbols except those. Derivative keeps track of symbols with respect to which it will perform a derivative; those are bound variables, too, so it has its own free_symbols method.

Any other method that uses bound variables should implement a free_symbols method.

cancel(*gens, **args)

See the cancel function in sympy.polys

property canonical_variables

Return a dictionary mapping any variable defined in self.bound_symbols to Symbols that do not clash with any free symbols in the expression.

Examples

>>> from sympy import Lambda
>>> from sympy.abc import x
>>> Lambda(x, 2*x).canonical_variables
{x: _0}
classmethod class_key()

Nice order of classes.

coeff(x, n=1, right=False, _first=True)

Returns the coefficient from the term(s) containing x**n. If n is zero then all terms independent of x will be returned.

See also

as_coefficient

separate the expression into a coefficient and factor

as_coeff_Add

separate the additive constant from an expression

as_coeff_Mul

separate the multiplicative constant from an expression

as_independent

separate x-dependent terms/factors from others

sympy.polys.polytools.Poly.coeff_monomial

efficiently find the single coefficient of a monomial in Poly

sympy.polys.polytools.Poly.nth

like coeff_monomial but powers of monomial terms are used

Examples

>>> from sympy import symbols
>>> from sympy.abc import x, y, z

You can select terms that have an explicit negative in front of them:

>>> (-x + 2*y).coeff(-1)
x
>>> (x - 2*y).coeff(-1)
2*y

You can select terms with no Rational coefficient:

>>> (x + 2*y).coeff(1)
x
>>> (3 + 2*x + 4*x**2).coeff(1)
0

You can select terms independent of x by making n=0; in this case expr.as_independent(x)[0] is returned (and 0 will be returned instead of None):

>>> (3 + 2*x + 4*x**2).coeff(x, 0)
3
>>> eq = ((x + 1)**3).expand() + 1
>>> eq
x**3 + 3*x**2 + 3*x + 2
>>> [eq.coeff(x, i) for i in reversed(range(4))]
[1, 3, 3, 2]
>>> eq -= 2
>>> [eq.coeff(x, i) for i in reversed(range(4))]
[1, 3, 3, 0]

You can select terms that have a numerical term in front of them:

>>> (-x - 2*y).coeff(2)
-y
>>> from sympy import sqrt
>>> (x + sqrt(2)*x).coeff(sqrt(2))
x

The matching is exact:

>>> (3 + 2*x + 4*x**2).coeff(x)
2
>>> (3 + 2*x + 4*x**2).coeff(x**2)
4
>>> (3 + 2*x + 4*x**2).coeff(x**3)
0
>>> (z*(x + y)**2).coeff((x + y)**2)
z
>>> (z*(x + y)**2).coeff(x + y)
0

In addition, no factoring is done, so 1 + z*(1 + y) is not obtained from the following:

>>> (x + z*(x + x*y)).coeff(x)
1

If such factoring is desired, factor_terms can be used first:

>>> from sympy import factor_terms
>>> factor_terms(x + z*(x + x*y)).coeff(x)
z*(y + 1) + 1
>>> n, m, o = symbols('n m o', commutative=False)
>>> n.coeff(n)
1
>>> (3*n).coeff(n)
3
>>> (n*m + m*n*m).coeff(n) # = (1 + m)*n*m
1 + m
>>> (n*m + m*n*m).coeff(n, right=True) # = (1 + m)*n*m
m

If there is more than one possible coefficient 0 is returned:

>>> (n*m + m*n).coeff(n)
0

If there is only one possible coefficient, it is returned:

>>> (n*m + x*m*n).coeff(m*n)
x
>>> (n*m + x*m*n).coeff(m*n, right=1)
1
collect(syms, func=None, evaluate=True, exact=False, distribute_order_term=True)

See the collect function in sympy.simplify

combsimp()

See the combsimp function in sympy.simplify

compare(other)

Return -1, 0, 1 if the object is smaller, equal, or greater than other.

Not in the mathematical sense. If the object is of a different type from the “other” then their classes are ordered according to the sorted_classes list.

Examples

>>> from sympy.abc import x, y
>>> x.compare(y)
-1
>>> x.compare(x)
0
>>> y.compare(x)
1
compute_leading_term(x, logx=None)

Deprecated function to compute the leading term of a series.

as_leading_term is only allowed for results of .series() This is a wrapper to compute a series first.

conjugate()

Returns the complex conjugate of ‘self’.

copy()
could_extract_minus_sign()

Return True if self has -1 as a leading factor or has more literal negative signs than positive signs in a sum, otherwise False.

Examples

>>> from sympy.abc import x, y
>>> e = x - y
>>> {i.could_extract_minus_sign() for i in (e, -e)}
{False, True}

Though the y - x is considered like -(x - y), since it is in a product without a leading factor of -1, the result is false below:

>>> (x*(y - x)).could_extract_minus_sign()
False

To put something in canonical form wrt to sign, use signsimp:

>>> from sympy import signsimp
>>> signsimp(x*(y - x))
-x*(x - y)
>>> _.could_extract_minus_sign()
True
count(query)

Count the number of matching subexpressions.

count_ops(visual=None)

Wrapper for count_ops that returns the operation count.

default_assumptions = {}
diff(*symbols, **assumptions)
dir(x, cdir)
doit(**hints)

Evaluate objects that are not evaluated by default like limits, integrals, sums and products. All objects of this kind will be evaluated recursively, unless some species were excluded via ‘hints’ or unless the ‘deep’ hint was set to ‘False’.

>>> from sympy import Integral
>>> from sympy.abc import x
>>> 2*Integral(x, x)
2*Integral(x, x)
>>> (2*Integral(x, x)).doit()
x**2
>>> (2*Integral(x, x)).doit(deep=False)
2*Integral(x, x)
dummy_eq(other, symbol=None)

Compare two expressions and handle dummy symbols.

Examples

>>> from sympy import Dummy
>>> from sympy.abc import x, y
>>> u = Dummy('u')
>>> (u**2 + 1).dummy_eq(x**2 + 1)
True
>>> (u**2 + 1) == (x**2 + 1)
False
>>> (u**2 + y).dummy_eq(x**2 + y, x)
True
>>> (u**2 + y).dummy_eq(x**2 + y, y)
False
equals(other, failing_expression=False)

Return True if self == other, False if it does not, or None. If failing_expression is True then the expression which did not simplify to a 0 will be returned instead of None.

evalf(n=15, subs=None, maxn=100, chop=False, strict=False, quad=None, verbose=False)

Evaluate the given formula to an accuracy of n digits.

Parameters:
subsdict, optional

Substitute numerical values for symbols, e.g. subs={x:3, y:1+pi}. The substitutions must be given as a dictionary.

maxnint, optional

Allow a maximum temporary working precision of maxn digits.

chopbool or number, optional

Specifies how to replace tiny real or imaginary parts in subresults by exact zeros.

When True the chop value defaults to standard precision.

Otherwise the chop value is used to determine the magnitude of “small” for purposes of chopping.

>>> from sympy import N
>>> x = 1e-4
>>> N(x, chop=True)
0.000100000000000000
>>> N(x, chop=1e-5)
0.000100000000000000
>>> N(x, chop=1e-4)
0
strictbool, optional

Raise PrecisionExhausted if any subresult fails to evaluate to full accuracy, given the available maxprec.

quadstr, optional

Choose algorithm for numerical quadrature. By default, tanh-sinh quadrature is used. For oscillatory integrals on an infinite interval, try quad='osc'.

verbosebool, optional

Print debug information.

Notes

When Floats are naively substituted into an expression, precision errors may adversely affect the result. For example, adding 1e16 (a Float) to 1 will truncate to 1e16; if 1e16 is then subtracted, the result will be 0. That is exactly what happens in the following:

>>> from sympy.abc import x, y, z
>>> values = {x: 1e16, y: 1, z: 1e16}
>>> (x + y - z).subs(values)
0

Using the subs argument for evalf is the accurate way to evaluate such an expression:

>>> (x + y - z).evalf(subs=values)
1.00000000000000
expand(deep=True, modulus=None, power_base=True, power_exp=True, mul=True, log=True, multinomial=True, basic=True, **hints)

Expand an expression using hints.

See the docstring of the expand() function in sympy.core.function for more information.

property expr_free_symbols

Like free_symbols, but returns the free symbols only if they are contained in an expression node.

Examples

>>> from sympy.abc import x, y
>>> (x + y).expr_free_symbols 
{x, y}

If the expression is contained in a non-expression object, do not return the free symbols. Compare:

>>> from sympy import Tuple
>>> t = Tuple(x + y)
>>> t.expr_free_symbols 
set()
>>> t.free_symbols
{x, y}
extract_additively(c)

Return self - c if it’s possible to subtract c from self and make all matching coefficients move towards zero, else return None.

Examples

>>> from sympy.abc import x, y
>>> e = 2*x + 3
>>> e.extract_additively(x + 1)
x + 2
>>> e.extract_additively(3*x)
>>> e.extract_additively(4)
>>> (y*(x + 1)).extract_additively(x + 1)
>>> ((x + 1)*(x + 2*y + 1) + 3).extract_additively(x + 1)
(x + 1)*(x + 2*y) + 3
extract_branch_factor(allow_half=False)

Try to write self as exp_polar(2*pi*I*n)*z in a nice way. Return (z, n).

>>> from sympy import exp_polar, I, pi
>>> from sympy.abc import x, y
>>> exp_polar(I*pi).extract_branch_factor()
(exp_polar(I*pi), 0)
>>> exp_polar(2*I*pi).extract_branch_factor()
(1, 1)
>>> exp_polar(-pi*I).extract_branch_factor()
(exp_polar(I*pi), -1)
>>> exp_polar(3*pi*I + x).extract_branch_factor()
(exp_polar(x + I*pi), 1)
>>> (y*exp_polar(-5*pi*I)*exp_polar(3*pi*I + 2*pi*x)).extract_branch_factor()
(y*exp_polar(2*pi*x), -1)
>>> exp_polar(-I*pi/2).extract_branch_factor()
(exp_polar(-I*pi/2), 0)

If allow_half is True, also extract exp_polar(I*pi):

>>> exp_polar(I*pi).extract_branch_factor(allow_half=True)
(1, 1/2)
>>> exp_polar(2*I*pi).extract_branch_factor(allow_half=True)
(1, 1)
>>> exp_polar(3*I*pi).extract_branch_factor(allow_half=True)
(1, 3/2)
>>> exp_polar(-I*pi).extract_branch_factor(allow_half=True)
(1, -1/2)
extract_multiplicatively(c)

Return None if it’s not possible to make self in the form c * something in a nice way, i.e. preserving the properties of arguments of self.

Examples

>>> from sympy import symbols, Rational
>>> x, y = symbols('x,y', real=True)
>>> ((x*y)**3).extract_multiplicatively(x**2 * y)
x*y**2
>>> ((x*y)**3).extract_multiplicatively(x**4 * y)
>>> (2*x).extract_multiplicatively(2)
x
>>> (2*x).extract_multiplicatively(3)
>>> (Rational(1, 2)*x).extract_multiplicatively(3)
x/6
factor(*gens, **args)

See the factor() function in sympy.polys.polytools

find(query, group=False)

Find all subexpressions matching a query.

property formula

Return a Formula with only terms=[self].

fourier_series(limits=None)

Compute fourier sine/cosine series of self.

See the docstring of the fourier_series() in sympy.series.fourier for more information.

fps(x=None, x0=0, dir=1, hyper=True, order=4, rational=True, full=False)

Compute formal power power series of self.

See the docstring of the fps() function in sympy.series.formal for more information.

property free_symbols

Return from the atoms of self those which are free symbols.

Not all free symbols are Symbol. Eg: IndexedBase(‘I’)[0].free_symbols

For most expressions, all symbols are free symbols. For some classes this is not true. e.g. Integrals use Symbols for the dummy variables which are bound variables, so Integral has a method to return all symbols except those. Derivative keeps track of symbols with respect to which it will perform a derivative; those are bound variables, too, so it has its own free_symbols method.

Any other method that uses bound variables should implement a free_symbols method.

classmethod fromiter(args, **assumptions)

Create a new object from an iterable.

This is a convenience function that allows one to create objects from any iterable, without having to convert to a list or tuple first.

Examples

>>> from sympy import Tuple
>>> Tuple.fromiter(i for i in range(5))
(0, 1, 2, 3, 4)
property func

The top-level function in an expression.

The following should hold for all objects:

>> x == x.func(*x.args)

Examples

>>> from sympy.abc import x
>>> a = 2*x
>>> a.func
<class 'sympy.core.mul.Mul'>
>>> a.args
(2, x)
>>> a.func(*a.args)
2*x
>>> a == a.func(*a.args)
True
gammasimp()

See the gammasimp function in sympy.simplify

getO()

Returns the additive O(..) symbol if there is one, else None.

getn()

Returns the order of the expression.

Examples

>>> from sympy import O
>>> from sympy.abc import x
>>> (1 + x + O(x**2)).getn()
2
>>> (1 + x).getn(