Changing selections sometimes leads to confusing charts

[HelenCEBM]

From here:

http://dash.openpathology.net/data/chart/by/practice_id/showing/ccg_id/all/lab_id/nd/numerators/HB/denominators/K/filter/all?highlight_entities=L82012&highlight_entities=L83117&highlight_entities=L83663#show-org-filter

if you then select "compare by Test",

the heatmap is updated OK but the line chart isn't, which means:

• you get a line chart still for the specific selected practice but it's no longer an option to select (or deselect) a specific practice (red marks)
• it has "remembered" practice selections in the URL but it hasn't retained the lab filter (yellow highlights)

suggestions

This kind of issue is probably inevitable to some extent in this multi-purpose tool, but we could:

• allow selection of a single practice even when looking at results by test
• remember lab selections in the URL?
• create error messages when an odd selection is made that doesn't allow charts to be updated properly

[sebbacon]

Try it now - does it make more sense?

[sebbacon]

I have also tried adding the possibility of filtering to a single practice when grouping by test.

However, this needs testing - not sure it makes intuitive sense. For example it introduces a discontinuity with how you restrict to a single test (i.e. with the right hand dropdown). Let me know what you think.