Sympy convert string to function which accepts arrays-CodePudding

So I am trying to convert a string function expression, to an a function which can take a numpy array ideally.

So far I have this:

import sympy as sp
expr = "x^2   y^2"
f = sp.sympify(expr)
symbol_list = []
for symbol in f.free_symbols:
        symbol = sp.Symbol(str(symbol))
        symbol_list.append(symbol)

Now the function f will be able to be used using sp.subs but I want to substitute a numpy array for x, y. Any ideas?

CodePudding user response：

Within SymPy, f is called symbolic expression. Sympy also exposes lambdify, which allows to convert a symbolic expression to a numerical function.

# extract the symbols from the symbolic expression: order them alphabetically.
# This will be used to create the signature of the numerical function.
signature = sorted(f.free_symbols, key=str)
print(signature)
# out: [x, y]

# create the numerical function
num_f = lambdify(signature, f)

# evaluation
import numpy as np
x = np.arange(1, 10)
y = 2
print(num_f(x, y))
# out: array([ 5,  8, 13, 20, 29, 40, 53, 68, 85])