sympy lambdify with scipy curve_fit

Question

I constructed a sympy expression and I used lambdify to convert to a numpy function as follow:

import sympy
from sympy.parsing.sympy_parser import parse_expr

x0,x1 = sympy.symbols('x0 x1')
a,b,c = sympy.symbols('a b c')
func=parse_expr('a*x0 + b*x1 + c*x0*x1')
p = [x0,x1,a,b,c]

npFunc = sympy.lambdify(p,func,'numpy')

but when I use scipy's curve_fit to fit npFunc for (a,b,c) with the two independent variables x0 and x1, it fails. I can't figure out how to use lambdify to make npFunc to work like this (with the unpacking):

def npFunc(X, a, b, c):
    x0,x1 = X
    return a*x0 + b*x1 + c*x0*x1

How should I do it?

Answer 1

The docs for your function (using the ipython ? shortcut)

In [22]: npFunc?
Signature: npFunc(x0, x1, a, b, c)
Docstring:
Created with lambdify. Signature:

func(x0, x1, a, b, c)

Expression:

a*x0 + b*x1 + c*x0*x1

Source code:

def _lambdifygenerated(x0, x1, a, b, c):
    return (a*x0 + b*x1 + c*x0*x1)

With the suggest alternative:

In [23]: p = [[x0, x1], a, b, c]

In [24]: npFunc = lambdify(p,func,'numpy')

In [25]: npFunc?
Signature: npFunc(_Dummy_22, a, b, c)
Docstring:
Created with lambdify. Signature:

func(arg_0, a, b, c)

Expression:

a*x0 + b*x1 + c*x0*x1

Source code:

def _lambdifygenerated(_Dummy_22, a, b, c):
    [x0, x1] = _Dummy_22
    return (a*x0 + b*x1 + c*x0*x1)