Dealing with 0s in TimeSeries

Hi Team,

In the below image you can see an example time series where the data starts from a non-zero value —> goes to zero —> after four days, goes back to something non zero.

Chart 1

Chart 1 however is super deceptive and misleading. It’s essentially connecting the value of 0 with the next value of 6235 but the curve makes it look like a steady increase when in fact it should be a total gap, something like this:

Chart 2

The difference in Chart 1 and Chart 2 is a consequence of Chart 1’s query having no group by but Chart 2’s query having a group by on time for 30m.

Chart 2 and it’s query compromise on data granularity since we can only show one datapoint per 30m.


How can I retain the same level of granularity that comes without a group by and at the same time show a chart which does not mislead in the way Chart 1 does?

I think the above question boils down to how to convert values that are 0 into null (I think), any ideas on how to tackle this problem would be much appreciated :pray:

cc some community champions for help: @grant2, @tanmaybhat24, @mikhailvolkov, @iblancasa :pray: