Is there a more Pythonic way to access to functions via their string name?

user_1422096 0
Viewed 65

I have a few functions (hardcoded, I don't want Python to compute the derivatives, in reality the functions are more complex):

def f(x): return x
def g(x): return x ** 2
def f_prime(x): return 1
def g_prime(x): return 2 * x

and a list a functions:

myfunctions = [f, f, g, f]

How to get the list of the related "prime" functions?

I was thinking about:

myderivatives = [globals()[function.__name__ + '_prime'] for function in myfunctions]

and it works:

[<function f_prime at 0x00000000022CF048>, <function f_prime at 0x00000000022CF048>, <function g_prime at 0x00000000022CF0B8>, <function f_prime at 0x00000000022CF048>]

but is there a more Pythonic way to get the list of the "prime" functions associated to the original versions?

Also there is a corner case:

from module1 import thisfunction as f
def f_prime(x): 
    pass

globals()[f.__name__ + '_prime']  
# doesn't work because f.__name__ + '_prime' is 'thisfunction_prime' and not 'f_prime'
function python variables
2 Answers

Define a dictionary to hold your functions and then use a list comprehension:

funcd = {'f': f, 'g': g, 'f_prime': f_prime, 'g_prime': g_prime}

myderivatives = [func for name, func in funcd.items() if name.endswith('_prime')]

The benefit of this solution is it defines precisely the scope of functions you wish to filter. Use of globals is not recommended, and very rarely necessary.

user_92095460

Following on from jpp's answer you could arrange the functions in a dict mapping from function to derivative like so

In [64]: primes = {f: f_prime, g: g_prime}

In [65]: primes.values()
Out[65]: dict_values([<function g_prime at 0x7f3594174c80>, <function f_prime at 0x7f3594174a60>])

In [66]: primes[f]
Out[66]: <function __main__.f_prime(x)>
user_54913750
Related