0

N00b question, maybe.

I am new-ish to python, but I was wondering if it is possible to store a function in an array? I want to do an array multiplication, where the value of one array is multiplied by a function on a location of another array of functions. Or actually, the value of the first array is inserted in the function of the designated location. This should create a new array where the values are an outcome of the "multiplication".

>>> import numpy as np
>>> a = [[1, 0], [0, 1]]
>>> b = [[f(x), g(x)], [(h(x), f(x)]]
>>> np.dot(a, b)
array([[0, 1],
   [2, 0]])

Assuming that f(x), g(x) and h(x) are defined functions. In this case python will say that x is not defined. So far I know. However, I do not want to say, for example, f(a[0][1]), because I want to reuse array b and also be able to put the functions on random locations in the array.

In short I detect three questions:
- Is there a known way to have an array where the values are functions?
- If not, should I redefine an array function or write a new class for this? (how do I attack this problem?)
- If it is possible to create an array of functions, can I fill the 'function values' dynamically in the array (populate the array dynamically with functions) or can it only be static values?

like for example

b[0]=f(x)

And yes, I really want to do this with python.

Improve this question
0

3 Answers 3

Reset to default
4

f(x) is not a function (even in mathematics). The function is f. In python there is no problem to store functions in array:

def my_func():
    print("Foo")

my_other_func = lambda: print("Bar")

my_arr = [my_func, my_other_func]
Improve this answer
Sign up to request clarification or add additional context in comments.

14 Comments

f(x) represents a function, but I do not know which function yet. I think your solution might be what I was searching for. Thanks!
functions are not special in python, every array operation will work normally. Also, unless the result of calling f with x is itself a function, f(x) is not a function
When I do, say def my_func(x): 6 * x (with proper styling ofcourse) and def my_func(x): 8 * x then it gives me a TypeError: unsupported operand type(s) for *: 'int' and 'function'
a = [1, 0] and then np.dot(a, my_arr)
[ f(x) for f in my_arr for x in arr]
|
3

Your example puts the result of calling a function into your array.

b[0] = f

actually puts the function itself into the array, so that

b[0](x)

would have the same effect as 

f(x)
Improve this answer

1 Comment

I see. I did not know that. Thanks.
1

Is there a known way to have an array where the values are functions?

There is, as stated by others already.

def my func(x):
    return x

my_other_func = lambda x: x

l = [my_func, my_other_func]

If not, should I redefine an array function or write a new class for this? (how do I attack this problem?)

Not relevant

If it is possible to create an array of functions, can I fill the 'function values' dynamically in the array (populate the array dynamically with functions) or can it only be static values?

The 'function values' are called parameters and can be assigned after having the function in the array:

a[0](5)       # == my_func(5)
a[1]('Hello') # == my_other_func('Hello')

The problem is that you are trying to use matrix multiplication as parameter passing and that will not work, you could create a helper fucntion that does it.

def callByElements(parameterMatrix, functionMatrix):
    rows = len(parameterMatrix)
    cols = len(parameterMatrix[0])
    result = np.zeros( (rows, cols) )
    for i in range(rows):
        for j in range(cols):
            result[i,j] = functionMatrix[i,j](parameterMatrix[i,j])
    return result
Improve this answer

3 Comments

Thanks for your answer. However, I purposely do not want to use for loops and was hoping that it would be possible in a matrix instead. haha, yes I totally could think of the word parameter, thanks.
Well you cant expect np.dot() which is a matrix multiplication method to work as a function calling method.
Yes I saw that, but it was merely to illustrate my intensions. I want to do matrix manipulations as it is often faster than for loops.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.