We recently made a new range of filters available in Bime for pivot tables and dashboards. These can be pretty powerful and have a number of effects on the final dashboard, depending on the way they are implemented. Picking the right filter with the right application strategy allows you to look deeper into your data and arrive at the right conclusions more quickly.
Using filters is easiest to explain using an example, so we'll have a look at the filter "top". Top filters out all but the greatest value in a data set, but the principles are the same for all filters. Top can be extended to include more than one value, so for example top 2 zooms in on the two greatest values.
This table can be saved to a dashboard and made public. Any filters then applied by users in the dashboard (called global filters) will be applied to the complete data set and to all queries in that dashboard.
Say now the user applies a top 3 global filter (on the dashboard). This filter overrides the query specific top 2 filter and so three values will be seen. When the global filter is removed, the original top 2 filter becomes dominant again so only the highest two values are shown. The two queries below have top 2 applied; try changing this to top 3 by selecting the Top tab (on the far right) and changing the value from 2 to 3. You will see the results change for both queries :
If you then remove this global filter by unchecking the Top box, both queries revert to their original filters (both top 2).
If our top 2 filtered query is included in a dashboard where the other queries have the same filter (as in the example above), we see the top 2 option selected on the global filter panel. This is because top 2 is essentially a global filter in this case. If it is saved in a dashboard that contains queries with other filters or no filter, each keeps its own filters and no global filter toggle is checked.
The dashboard below contains a query with a top 3 filter, and one without a filter. Observe, no global filter toggle is checked.
So, you can see that it is important to apply filters at the right time to achieve the required result. Applying the filters strategically allows users of your dashboards to see the important results found during data exploration, but to override the filters and explore all of the data. It also allows users to compare like with like in dashboards once they have evaluated each visualization independantly.
There are lots of different filters available in Bime, allowing you to zoom in and analyse any cross-section of your data. Log in now to try them out.