Fastest way to grow a numpy numeric array

Question

Requirements:

• I need to grow an array arbitrarily large from data.
• I can guess the size (roughly 100-200) with no guarantees that the array will fit

Answer 1

If you want improve performance with list operations, have a look to blist library. It is a optimized implementation of python list and other structures.

I didn't benchmark it yet but the results in their page seem promising.