1

Lets say I want an array vector like this one [x , 2x] and then I want to compute this array when x=5. So the result become [5 , 10].

This can be easily done in matlab using symbolic toolbox. However, I just moved from matlab to python and I was recommended to use Numpy and Sympy to replace Matlab powerful matrix manipulations.

I tried the following code

import numpy as np
import sympy as sp

x=sp.symbols('x')
a=np.array([x , 2*x])

However, I could not find any command that can be used to evaluate "a" when x is any number.

Can someone direct me to the correct path please.

Improve this question
3
  • 1
    Numpy arrays don't play well with sympy objects (they're best if you use built-in numerical types with numpy). You should try sympy arrays instead, which will support something like .subs. Commented Nov 9, 2019 at 23:04
  • 1
    Look at recent sympy posts that also use lambdify. But be careful. There are a number of pitfalls. Check the dtype of your a array. Math on an object dtype array is hit-or-miss. sympy and numpy aren't as tightly integrated as on MATLAB. stackoverflow.com/questions/58772201/… Commented Nov 9, 2019 at 23:05
  • The sparse Matrix module might be useful. Commented Nov 10, 2019 at 2:09

2 Answers 2

Reset to default
5

As explained by hpaulj in the comments, SymPy and NumPy live in separate worlds. If you need fast number crunching, NumPy is your hero. SymPy shines in symbolic manipulations. To step from SymPy to NumPy, you either convert everything to numbers in SymPy before turning it over to NumPy. Or you use lambdify which converts SymPy expressions to NumPy functions.

See e.g. this post how SymPy calculates derivatives of very complex expressions, which are then converted to NumPy functions.

If you really want speed, a library such as Numba can convert NumPy functions to machine code (without the need of external compilation steps).

These libraries are very powerful on their own, but lack fluent interoperability, which can be confusing in the beginning.

Your example works if you let SymPy operate on every element of the list, not on the list as a whole. SymPy's subs fills in variables. SymPy's evalf converts a constant expression to numbers (because normally SymPy keeps e.g. rationals and sqrt in symbolic form so it keeps a maximal precission).

import sympy as sp
import numpy as np

x = sp.symbols('x')
b = [x , 2*x]
c = [expr.subs(x, 5).evalf() for expr in b]
a = np.array(c)
Improve this answer
Sign up to request clarification or add additional context in comments.

Comments

2 
In [41]: from sympy import Matrix                                               

In [42]: m = Matrix([x, 2*x])                                                   

In [43]: m.subs(x,23)                                                           
Out[43]: 
⎡23⎤
⎢  ⎥
⎣46⎦
Improve this answer

Comments

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.