In vega how do I handle "other" cases with a scale

Elastic Stack Kibana
1

Hello All,

I have an ordinal scale that contains items [ A, B, C] and data that has values of [A,B,C,D,E,F] how do I put D,E, and F under a catch all with the scale? Similar to how "other" functions in other kibana visualizations, I ask because my original data set had too many values for the legend so I am building a second data set that only has a constrained value set so that I can limit my scale to like 10 items and then have others placed under "other"

Thank you!

2

@nyuriks - can u shed some light here.

3

Hi @nsimi, there is an example for top-K graph (funny enough it was created after I posted a similar issue a while back). The main idea is that you want to prepare your data first -- use aggregate transformation to group it by some value and sum or count the content of each group. Afterwards, use window transformation to sort the result and use row_number to decide if you want to keep the value as unique or as part of "all other". A side note -- Elasticsearch can actually do this work for you in many cases, and you can simply look at the resulting data.