Closed jsdillon closed 1 year ago
Patch coverage: 100.00
% and project coverage change: +0.02
:tada:
Comparison is base (
1568d3d
) 97.06% compared to head (ce86300
) 97.09%.:exclamation: Current head ce86300 differs from pull request most recent head 5b712eb. Consider uploading reports for the commit 5b712eb to get more accurate results
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Do you have feedback about the report comment? Let us know in this issue.
This PR builds a new antenna classification method that looks of anomalously high power in the diffs which is hypothesized to come from stale or mis-assigned packets data being written to either the even or odd visibilities, leading sky signal in the diffs... and sometimes even autos leaking into cross sums.
This classifier computes the power of the diffs, relative to noise, converts that to a z-score, and then averages the magnitude of that z-score over X-engines. It then assigns "badness" to the minimal set of antennas that can explain all the baselines detected with anomalous power in the diffs.
See the discussion in Slack: https://eoranalysis.slack.com/archives/C3ZPGMG3E/p1686617551061159