Spike: Better View "Raw Data" in data explorer


Currently, we display a tabular view of the CSV that is returned from a flux query. This can be particularly useful for debugging since it contains a detailed explanation of the data.

Specifically, it communicates

* which columns are in the group key
* the names of the columns returned
* the types of the columns returns
* the tables that are returned

The issue is that while it is nice to have a raw view of the data, it can be difficult to interpret the results. In particular determining which values are in the group key.

* table visually represented
* columns in the group key visually represented
* collapsable columns
* type information available on hover

Start in the data explorer's current tabular view of CSV returned from a Flux query. Determine how group keys, returned columns, column types, and tables are presented, then define a clearer visual treatment for tables, group-key columns, collapsible columns, and type information on hover. Done means the raw-data view is easier to interpret and covers the listed interactions.

Written by the indexing model from the issue text.