undefined | Better HN

Skip to content

Top Best Ask Show New Jobs

0 pointszeptomu9y ago0 comments

I think that is some kind of feature. Numpy tries hard to guarantee efficient access to its ndarrays and this is simpler to implement using fixed arrays and mostly you know your bounds at initialization time anyway.

Btw. you can grow them using hstack and vstack although I am not sure how efficient that is.

0 comments

6 comments · 2 top-level

dnautics9y ago· 3 in thread

There are fixed arrays in Julia too, for those who absolutely must eke out that last 1% gain in performance

xapata9y ago

It ain't just a 1% gain. Use a Python list if you want to append/pop.

_Wintermute9y ago

A deque is probably a better option for performance.

that may be so in Python, but have you checked the performance differences in Julia?

xaedes9y ago· 1 in thread

Growing with vstack or hstack is very slow. I have seen code where a big csv with time series data is parsed line by line the corresponding numpy array was grown using vstack for each row. worst case: quadratic runtime

xapata9y ago

That'd be so much faster using a generator and a NumPy.concatenate at the end. Orders of magnitude faster, if we're talking about a large CSV.

j / k navigate · click thread line to collapse