search

Becksteinlab / scipy_proceedings

Tools used to generate the SciPy conference proceedings
Other
0 stars 2 forks source link

plot data as averages + stdev #27

Closed orbeckst closed 5 years ago

orbeckst commented 5 years ago

The timing data should be averages for all runs where we have repeats.

Note:

For benchmark using distributed on local ssd and multiprocessing on Lustre, we only have single runs. For these using distributed on Lustre and multiprocessing on ssd, we have 5 runs for each case.

VOD555 commented 5 years ago

All plots include averages and stdev now.