I'm curious about how memory is used when inserting into an array vs reshaping an array to make it one larger and then replacing the value. I know it's generally bad practice to grow an array using a loop, but I was surprised to see such a difference between these two examples. Example 1) Insert into array. From what I could find, insert into array is making a copy with the new value added so I expected this to be slow with large arrays. Example 2) Reshape array and replace I would have thought this should take about the same amount of time as insert into array. However, when I timed each, I found reshaping and replacing is much faster once there is more than 10000 iterations and scales better. Does anyone have any idea why reshaping and replacing appears faster here? I can't seem to find any information about how memory is handled with reshape array but maybe I haven't looked in the right places. Again, I know this is bad practice and this probably isn't a thorough timing test, but I didn't expect this result.

LabVIEW

Memory usage for insert into array compared to reshape array

Author

Thanks everyone for the discussion and info, especially altenbach for the links to the benchmarking discussion. I'm glad to have more to read about this!

Message 11 of 11

(86 Views)

LabVIEW

Memory usage for insert into array compared to reshape array

Re: Memory usage for insert into array compared to reshape array