Currently, if any variable is missing, it ain't plotted.
That's obvious for lat/long variables, but it's also current policy for overlay variables (markers and floating plots).
We should decide what to do about missingness. It's not just whether or not to plot, it's also about communicating numbers to the user (like the birds eye plot in EDA)
Note that a related issue is how to represent zero counts in quantitative surveillance data, but it's not exactly the same. Although the zero data isn't stored in the database, the calculated/imputed zeroes are real data that need to be "plotted" and not greyed out or otherwise treated like missing data as in EDA.
Currently, if any variable is missing, it ain't plotted.
That's obvious for lat/long variables, but it's also current policy for overlay variables (markers and floating plots).
We should decide what to do about missingness. It's not just whether or not to plot, it's also about communicating numbers to the user (like the birds eye plot in EDA)
Note that a related issue is how to represent zero counts in quantitative surveillance data, but it's not exactly the same. Although the zero data isn't stored in the database, the calculated/imputed zeroes are real data that need to be "plotted" and not greyed out or otherwise treated like missing data as in EDA.