sympy/sympy

A computer algebra system written in pure Python http://sympy.org/ . To get started to with contributing https://github.com/sympy/sympy/wiki/Introduction-to-contributing

Anutosh Bhat

@anutosh491

Guys , how can I use simplify() without changing factorial to a gamma function . I want to simplify it without affecting the factorial

ThePauliPrinciple

@ThePauliPrinciple

I don't think it is the best solution, but I find that substituting what I want untouched with a Symbol, then simplifying and then substituting back to often help in those cases

It might help if you create the symbol with the same assumptions as whatever you are substituting

lesshaste

@lesshaste

How do you solve simultaneous diophantine equations in sympy? For example 5x+6y+8z==1 and 6x-11y + 7z ==9

lesshaste

@lesshaste

it looks like you need to do https://stackoverflow.com/a/65438529/2287805

it would be great if this was all inside diophantine

sympy/sympy#20682

Megan Ly

@meganly

Last year expansion was causing slowdowns in the .equals method. In sympy/sympy#19673 we had

In [10]: e1 = 100*(x**3 + 1)**99                                                                                                               

In [11]: e2 = 300*x**2*(x**3 + 1)**99                                                                                                          

In [12]: %time e1.equals(e2)                                                                                                                   
CPU times: user 1.13 s, sys: 9.37 ms, total: 1.14 s
Wall time: 1.15 s
Out[12]: False

It seems like the .equals method has gotten faster. Anyone know what's changed?

Yashvardhan Prasad

@yasphy

Which easy to fix issues can I resolve now and how do I see if they are still open to be solved or not?

2 replies

lesshaste

@lesshaste

@yasphy There are quite a lot of things. It might depend how easy you want it to be

Yashvardhan Prasad

@yasphy

Any example to elaborate??

lesshaste

@lesshaste

@yasphy for example systems of linear diophantine equations.

sympy/sympy#20682

they have code but it just isn't reviewed and merged

Yashvardhan Prasad

@yasphy

Ok..thanks for the response

Anutosh Bhat

@anutosh491

Hello guys , I wanted to know whether there is a way that I could expand (1+x)**n as a binomial expansion , not sure I could find one . I ask this because I'm working on an issue which involves something like (1+x)**n- x**n, now the issue works perfectly when we use numbers like (1+x)**3 - x**3 and we could solve this through gammasimp() giving 3*x**2 + 3*x + 1 but symbolically not sure how we could approach it . Maybe if we could expand the (1+x)**n as a binomial expansion and cancel out the first term which would always be x**n then the maybe the issue wouldn't be that tough to solve . Currently gammasimp() isn't of much help ! and returns back the same expression !

Kalevi Suominen

@jksuom

@anutosh491 Can you use the explicit binomial expansion? Something like Sum(binomial(n, k)*x**k, (k, 0, n)) for (1 + x)**n.

Anutosh Bhat

@anutosh491

Yeah @jksuom that works , would surely go for that ! Just wanted to confirm whether we had some method to deal with this . The issue I'm trying to tackle is this

>>> k, n  = symbols('k, n', positive=True, integer=True)                                                                                                                              >>> from sympy import oo
>>> limit((n+1)**k/((n+1)**(k+1) - (n)**(k+1)), n, oo)
oo
>>> limit((n+1)**2/((n+1)**(3) - (n)**(3)), n, oo)
1/3
>>> limit((n+1)**2/((n+1)**(3) - (n)**(3)).gammasimp(), n, oo)
1/3
>>> limit((n+1)**k/((n+1)**(k+1) - (n)**(k+1)).gammasimp(), n, oo)
0

This is somewhat required to solve a somewhat bigger issue i'm trying to approach !

Anutosh Bhat

@anutosh491

works fine numerically but symbolically should return 1/k+1 , also there is somewhat involvement of gammasimp() there, so have to get that case correctly too ! Never thought a gammasimp() would alter limits from oo to 0 but it does .Maybe if I find it too buggy and some more places where it the (n+1)**k - n**k combo goes wrong , i'll raise this !

Anutosh Bhat

@anutosh491

 >>> limit((n+1)**k/(n*(-n**k + (n + 1)**k) + (n + 1)**k), n, oo)
1/(k + 1)

Apparently this arrangement of (n+1)**k -(n)**k gives the correct answer . I am not sure why this may be happening but obviously it feels a bit strange if different arrangements of the same expression can lead to different answers! The gammasimp of that expression which returns 0 is (-n*n**k + n*(n + 1)**k + (n + 1)**k)

falematte

@falematte

Hello guys, I have an expression that I am expanding using "series" with respect to the expression epsilon. I would like to tell sympy that some expressions must be considered of order epsilon instead of order 0. Is this possible? How to do this?

Héctor Rodríguez

@Hector_rodi_19_twitter

Does anyone have the algorithm to resolve first degree equations?

Anutosh Bhat

@anutosh491

Hey guys , I am approaching an issue from the series/concrete module which requires a really small contribution of the solverset module which I haven't explored much yet . Hence I wanted to confirm what whether there is any way to solve something like x = sin(x) or x = log(x) ,x = e**x . First one should return 0 in the Real domain and the other 2 should return an empty set . solveset() method returns something like ConditionSet(x, Eq(x - sin(x), 0), Complexes), ConditionSet(x, Eq(x - exp(x), 0), Complexes) which is not wrong but not what I would like and solvify returns Not implemented for these !

ThePauliPrinciple

@ThePauliPrinciple

What assumptions do you have for x?

1 reply

Ashton Bradley

@AshtonSBradley

Hi all... something I can't find in docs: if I get a conditional expression returned, how to I select the expression for the case that applies for my variables?

Screen Shot 2021-10-22 at 9.20.19 PM.png

I only want the "otherwise" result

Andreas Schuldei

@stockh0lm

hi, are user questions ok here?

Judging from scrollback, they are. :-)

I need to create a vector function in sympy with actual vectors, that can be transformed between coordinate systems.

I saw this https://stackoverflow.com/questions/46993819/how-to-create-a-vector-function-in-sympy which uses Matrix() as a workaround.

I need to transform my input and output vector from one coordinate system to another (and back). How are vector functions done in that case?

def B_el(r_vec):

    r1 = r_vec.i

    r2 = r_vec.j
    u = sin(r1)+cos(r2)
    v = -cos(r1)+sin(r2)
    return Matrix([u, v, 0])

Andreas Schuldei

@stockh0lm

formatting broken?

ThePauliPrinciple

@ThePauliPrinciple

Depending on what you "vector" is, it will transform differently under a coordinate transform

e.g. a velocity and a coordinate transform completely differently with a non-linear coordinate transformation

And in the case of a velocity, you could have it transform covariantly or contravariantly

Andreas Schuldei

@stockh0lm

the input vector is a radius, so distance and direction. the output vector is a magnetic field vector.

so it is just the geometic transformation. split up the three spatial components differently, maintain the size

part of the task is to identify the coordinate transformation - the rotation angels and location vector. But that comes later, after having figured out how to write a vector function :-)

Andreas Schuldei

@stockh0lm

I still am unclear as to how to return a vector

def B_el(r_vec):

    r1 = r_vec.i

    r2 = r_vec.j
    u = sin(r1)+cos(r2)
    v = -cos(r1)+sin(r2)
    return Matrix([u, v, 0])

What would i need to change in order to return a vector, which can be transformed into a different (also resting) coordinate system, differing in axis rotations and a location shift?

ellydo17

@ellydo17

Hello, I am a new member to SymPy. I look forward to contributing to SymPy as a developer. Do we have different channels for SymPy users and developers?

Andreas Schuldei

@stockh0lm

Traceback (most recent call last):
  File "C:/Users/Andreas Schuldei/PycharmProjects/lissajous-achse/hgü-kabel-detektion-symbolic.py", line 109, in <module>
    result = solve(equations, (CM0, theta_i, theta_j, theta_k, i, a, B_earth))
  File "C:\Users\Andreas Schuldei\PycharmProjects\lissajous-achse\venv\lib\site-packages\sympy\solvers\solvers.py", line 1096, in solve
    solution = _solve_system(f, symbols, **flags)
  File "C:\Users\Andreas Schuldei\PycharmProjects\lissajous-achse\venv\lib\site-packages\sympy\solvers\solvers.py", line 1730, in _solve_system
    i, d = _invert(g, *symbols)
  File "C:\Users\Andreas Schuldei\PycharmProjects\lissajous-achse\venv\lib\site-packages\sympy\solvers\solvers.py", line 3118, in _invert
    rhs -= indep
  File "C:\Users\Andreas Schuldei\PycharmProjects\lissajous-achse\venv\lib\site-packages\sympy\core\numbers.py", line 2194, in __sub__
    return Rational.__sub__(self, other)
  File "C:\Users\Andreas Schuldei\PycharmProjects\lissajous-achse\venv\lib\site-packages\sympy\core\decorators.py", line 89, in __sympifyit_wrapper
    return func(a, b)
  File "C:\Users\Andreas Schuldei\PycharmProjects\lissajous-achse\venv\lib\site-packages\sympy\core\numbers.py", line 1725, in __sub__
    return Number.__sub__(self, other)
  File "C:\Users\Andreas Schuldei\PycharmProjects\lissajous-achse\venv\lib\site-packages\sympy\core\decorators.py", line 89, in __sympifyit_wrapper
    return func(a, b)
  File "C:\Users\Andreas Schuldei\PycharmProjects\lissajous-achse\venv\lib\site-packages\sympy\core\numbers.py", line 733, in __sub__
    return AtomicExpr.__sub__(self, other)
  File "C:\Users\Andreas Schuldei\PycharmProjects\lissajous-achse\venv\lib\site-packages\sympy\core\decorators.py", line 251, in _func
    return func(self, other)
  File "C:\Users\Andreas Schuldei\PycharmProjects\lissajous-achse\venv\lib\site-packages\sympy\core\decorators.py", line 126, in binary_op_wrapper
    return f(self)
  File "C:\Users\Andreas Schuldei\PycharmProjects\lissajous-achse\venv\lib\site-packages\sympy\core\decorators.py", line 127, in binary_op_wrapper
    return func(self, other)
  File "C:\Users\Andreas Schuldei\PycharmProjects\lissajous-achse\venv\lib\site-packages\sympy\vector\basisdependent.py", line 352, in __rsub__
    raise TypeError("Invalid argument types for subtraction")
TypeError: Invalid argument types for subtraction

How can I inspect the individual equations in the list equations to narrow this down? The equations are relatively long, and I have no clue how to find this.

Sean Logan

@fountain23

Hello, I have discovered a bug with sympy.vector.divergence() when used in cylindrical coordinates

divergence( R rhat) gives 2, the correct answer. But divergence( 1 rhat) gives 0. Wrong! The correct answer is 1/R

N = CoordSys3D('N', transformation='cylindrical', vector_names=("r", "phi", "z"), variable_names=("R", "PHI", "Z"))

R = N.R

rhat = N.r

lesshaste

@lesshaste

The channel isn't as active as it should be. You might be better off with the mailing list

Anutosh Bhat

@anutosh491

Should using subs back and forth make any difference mathematically ? This is for cases having log where I have seen this multiple times

>>> (log(0/a)).subs(a, 0)
zoo
>>> (log(b/a)).subs(a, 0).subs(b, 0)
nan
>>> (log(b/a)).subs(b, 0).subs(a, 0)
zoo

Where communities thrive

People

Repo info

Activity