Closed leoscholl closed 3 years ago
might be more useful to have a rolling average so we can detect problems towards the end of an experiment. number of missing inputs per 100, for example.
changed in dbdde77651bef58d9c9477c94f649243dd93ad61
might be more useful to have a rolling average so we can detect problems towards the end of an experiment. number of missing inputs per 100, for example.