Timeline for Iterate through two arrays calculating a statistic from indexes of zeros
Current License: CC BY-SA 4.0
4 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Aug 17, 2021 at 22:07 | comment | added | Peilonrayz♦ | FWIW I found that to be clear. The reason I mentioned C specifically, is because C is a common way to speed up Python. Such as in much of std-lib like bisect/_bisect or numpy which are coincidentally mentioned in the above answers. | |
| Aug 17, 2021 at 21:59 | comment | added | Peter Cordes |
@Peilonrayz: Just to be clear, count_zeros( ts_1 | ts_2 ) is pseudo-code, not something you can literally write in C. You can't OR whole arrays, you need to loop over them. (You could do basically that with C++ std::bitset<8000> - it does overload the | operator and has a .count() method to count ones which hopefully uses an optimized popcnt under the hood.)
|
|
| Aug 17, 2021 at 21:55 | comment | added | Peilonrayz♦ | Thank you for posting your answer. Your answer comments about things I've never heard of, and you're talking about things I can't think of a way to implement in Python. Your answer is the first compelling argument I've seen against Python (wrt being a high-level language). However I don't think your answer is inappropriate as using C is an accepted practice when you need speed Python can't get you. I'd say it's idiomatic Python, but given I, like many other Python programmers, can't program in C, it may be a stretch to say idiomatic... | |
| Aug 17, 2021 at 21:35 | history | answered | Peter Cordes | CC BY-SA 4.0 |