Open rjhd2 opened 1 month ago
Currently for INM00042866, INM00043057 and PKM00041640 have something wrong in the station_level_pressure
with very large values. However, when setting bins for first differences in spike check, this results in the scripts wanting to set too large an arraysize. Catch this too.
Masking in data series means that differences are only calculated between pairs of non-masked entries. If one of two neighbouring values is masked, then the difference is masked.
Update to use compressed series to calculate differences, and then unpack back onto uncompressed data for flagging.