Feature: custom sort (in particular sort on measure), rank / top X

[rpbouman]

Consider a dataset like this: LAPD calls for service https://data.lacity.org/api/views/iy4q-t9vr/rows.csv?accessType=DOWNLOAD

It would be really convenient to be able to sort on the count and get a ranking for which codes are issued most often.

[rpbouman]

Few considerations:

• it should be possible to sort either (or both) axes
• it should probably be possible to specify multiple items to sort an axis by
• it should be possible to specify the sort direction independently for each sort item.
• intuitively, it feels like it should be possible to specify the row sort order by pointing to one or more columns, and v.v., it should be possible to specify the column sort order by pointing to one or more rows.
• If we follow the previous requirement through, then we should consider that not all rows or all columns are created equal: some rows/columns are headers, and some contain cells. For example, in the example above, sorting on the "call type description" column should result in the rows being sorted by the values in "call type description". But sorting by the next column, "Dispatch Date month 1" & " count" should result in sorting the rows by the " count" cell values.
• There is a matter of Ui. It seems natural that the columns/rows used for sorting should get some kind of marker indicating the column is currently active as sort item. This marker would need to indicate the sort direction, but also, the ordinal position of the item within the sort key for that axis of the pivot table.
• Alternatively or in addition, it seems the query axis items could also have the role of controlling the sort. In that case, the sorting of the tuples along the axes would be pinned directly to the rows and columns items. But it poses a challenge when sorting by cell values, as the cells items would then need controls for both the rows and the columns axis.