Missing a lot of messy data

kbroman / detectPVC

R package to detect PVCs in Polar H10 data

Other

0 stars 1 forks source link

Missing a lot of messy data #10

Open kbroman opened 3 weeks ago

kbroman commented 3 weeks ago

The algorithm to find bad segments of data is missing a lot, particularly cases where we end up identifying many closely-spaced peaks.

kbroman commented 2 weeks ago

In find_bad_segments(), seems like I should change the default for runmean_thresh from 0.7 to 1.0. But this is not about missing messy data as calling bits messy when they're actually fine.