I have been trying to generate a list of lambda functions in python using list comprehension. but it didn't work,
for example
fl=[lambda x: x**i for i in range(5)]
i have check the other question, it basically generate the same function based on the reference of i.
so I also tried partial.
from functools import partial
fl=[partial(lambda x: x**i) for i in range(5)]
but it didn't work too. any help will be appreciated. cheers~
You're tripping over Python scopes.
fl=[lambda x, i=i: x**i for i in range(5)]
i at definition time (as it is the case with default values), so they access the one from the encloding scope when running. They do this because it was defined that upvalues shouldn't become local copies, which may be argued to be a scoping decision, but it doesn't come from them not having their own scope.You're effectively passing i in by name.
fl=[lambda x: x**i for i in range(5)]
Every time lambda is executed, it binds the same i to the function, so when the function is executed (later) it uses the then-current value of i (which will be 4). You should pass it in as a default argument instead:
fl=[lambda x, j=i: x**j for i in range(5)]
Actually, I noticed that you're misusing partial. Here:
fl = [partial(lambda x, y: y ** x, i) for i in range(5)]
That works as well.
j because it makes it clearer what's going on (to my mind, anyway).Another common workaround is to use a closure:
def create_f(i):
def f(x):
return x**i
return f
fl = [create_f(i) for i in range(5)]
fl[0](2) = fl[1](2) = 16