idaholab/MontePy

Automate finding O(n^2) opertions

Closed this issue · 1 comments

Internal to INL (through ICSBEP) we have > 1,000 models of varying lengths, and profiling data for each time these are read. It would be good to automate a regression for each function for number of calls versus number of lines. The order of this regression will show if a function is being called in O(N^2) time.

Since MontePy we have O(N) performance pretty solidly. Nothing in the profiling data has suggested otherwise. I think this is no longer necessary.