Environmental covariate by Taxa matrix deployment

[samaperrin]

Why have changes been made?

We want to be able to only use select covariates for each species group (taxa). This push allows us to select TRUE or FALSE for each environmental covariate on a taxa basis. There is also a bug fix (more on that below).

What changes have been made?

• data/external/focalCovariates.csv - New columns entitled "selected_speciesGroup" allow us to provide covariates on a taxa basis.
• pipeline/import/environmentalImport.R - Since there are now multiple "selected" columns, we need to take all covariates that have TRUE in at least one of these columns
• functions/modelPreparation.R - This function now selects covariates to add to the model workflow based on the taxon in use. If there is no particular selecton for a taxa then it simply adds all covariates that were calculated.
• pipeline/models/speciesModelRuns.R - Need to import focal covariates

Additional changes

GeoNorge had an update yesterday and changed the categorisation of their road types. This means we have to update the geonorge script. There was also an error in the get_modis script. It should have used regionGeometry instead of regionGeometryBuffer, and the net_primary_productivity script was using an incorrect file path.

• functions/get_geonorge.R - Now we update the roads based on a new column.
• pipeline/import/utils/defineEnvSource.R - The section where we refien the vector data down to relevant types is now in the get_geonorge function. regionGeometry is now used in the get_modis function.
• functions/get_modis.R - The data path for MODIS data has been updated.

[samaperrin]

Yes this makes much more sense