How long to use generators?

Question

Python

Hello.
I am now at the stage of loving generators that I use them a lot more than lists. Literally everywhere, I use only yield and yield from
Lists only if you need to iterate over the sequence once, and then again. Or if you need to send a list via API and make JSON out of it. That is, even if the iteration is near-parallel, for example, with a lag of a couple of elements, I use itertools.tee
I know that generators are slower than lists (although I could not think of adequate benchmarks, because generators are lazy and how to test exactly the time, spent on yield is not clear). Although, creating a list with appends plus processing it usually takes longer:

I know that almost all lists that I would create instead of generators would whiz into memory, but I can't stop saving this memory.
Can you give me advice on when it is appropriate to use generators, and when it is stupid?

Answer 1

Ek you, brother, hooks))) You might think that the only purpose of lists is to iterate over them)
But we know that everything is just the opposite: generators are for iteration, and lists are for random access (well, to iterate ).
Generators not only save memory, but also consume it - per persistent stack frame, at least. So there is no point in saving on small lists.
And vice versa - generators can win in speed where the calculation of values ​​is difficult, and the iteration can break without filling the list / generator to the end.

Answer 2

https://stackoverflow.com/questions/47789/generato... it 's written here