动态创建 SciPy curve_fit 函数

Question

SO noob here.

I am creating a wrapper around scipy.optimize.curve_fit for multivariate regression. I was able to successfully run a vanilla version with two or three independent variables -

Two independent variables -

def fn(x, a, b1, b2):
    return a + b1*x[0] + b2*x[1]

popt, pcov = curve_fit(fn, x, y)  

Three independent variables -

def fn(x, a, b1, b2, b3):
    return a + b1*x[0] + b2*x[1] + b3*x[2]

popt, pcov = curve_fit(fn, x, y)  

The thing is, I don't know how many independent variables will there be in input and I don't want to iterate through all permutations of the function definition manually, which I also don't think is the best practice. So the question is How to create the function fn dynamically here based on the input dataframe?

Answer 1

This is not a very elegant solution, but you can always generate Python code from strings using the exec command like this . Then you can just generate a string dynamically for your function definitions and with exec actually create the function.