Unexpected NAs in NDVI and Forecast data

ecohealthalliance / open-rvfcast

Wellcome Open RVFCast project repository

Other

0 stars 1 forks source link

Unexpected NAs in NDVI and Forecast data #71

Closed emmamendelsohn closed 7 months ago

emmamendelsohn commented 7 months ago

Found during NA checks on augmented data. Seems to be some misalignment between historical means and lag means when joining, even though they are based on the same rasters. Not sure why it's not a problem for weather data.

emmamendelsohn commented 7 months ago

For NDVI: https://github.com/ecohealthalliance/open-rvfcast/issues/72 Prior to 2018: NAs are due to region missing from Eastern Africa in modis data After 2018: NAs are due to smaller pockets of missing data on a per-cycle basis

NAs in forecast data seem to have happened stochastically, when using a previous version of arrow. Can't find anything in the arrow package history to suggest what has changed. Currently rerunning the files that have NAs.