Remove series with only NA values
I noticed that some providers like World Bank have many series with all observation in "unknown" status.
Examples:
- http://datapipes.okfnlabs.org/csv%20-t/html/?url=https://api.next.nomics.world/series/worldbank/WDI/DT.GRE.DPPG-FR.tsv
- http://datapipes.okfnlabs.org/csv%20-t/html/?url=https://api.next.nomics.world/series/worldbank/WDI/DT.AXA.PRVT.CD-FR.tsv
- http://datapipes.okfnlabs.org/csv%20-t/html/?url=https://api.next.nomics.world/series/worldbank/WDI/DT.GRE.PRVT-OM.tsv
- http://datapipes.okfnlabs.org/csv%20-t/html/?url=https://api.next.nomics.world/series/worldbank/WDI/DT.MAT.PRVT-1W.tsv
- http://datapipes.okfnlabs.org/csv%20-t/html/?url=https://api.next.nomics.world/series/worldbank/DB/IC.TAX.PFI-AD.tsv
- http://datapipes.okfnlabs.org/csv%20-t/html/?url=https://api.next.nomics.world/series/worldbank/DB/IC.EC.COST-AS.tsv
I don't think that the "None" values are due to conversion errors.
Now my question: at the project scale, should we remove those series?
My main argument is that the existence of empty series create false positive counts in facets. From a mathematical point of view, these empty series are holes in the hypercube of the series.
Edited by Christophe Benz