How do I remove a bogus category from a pie chart?

Anthony Mastrean March 5, 2014

I created a single-selection dropdown custom field and filled it with values. I created a pie chart on a dashboard from just that custom field. It's showing two slices with the same name, "Other".

One has the actual count, is clickable, and leads to a good filter. The other is not clickable and the count is arbitrary (it's been stuck at 6 since I created it and I've made more issues since then). I think the total issues count on this pie chart is off, too, as it's including the bogus Other slice.

I have no idea how to diagnose this problem or how to fix it. I've recreated the pie chart and it shows up again!

1 answer

1 accepted

3 votes
Answer accepted
Nic Brough -Adaptavist-
Community Leader
Community Leader
Community Leaders are connectors, ambassadors, and mentors. On the online community, they serve as thought leaders, product experts, and moderators.
March 5, 2014

You're very close to an answer without any help.

I think the bit you're missing is knowing how the pie charts break up large counts of data. When you've got large numbers, a pie chart becomes a bit fuzzy for very low counts, and because it's not very helpful to try to display a slice that's too small to see, the pie chart simplifies the minor cases.

Imagine two sets of data - 784 green, 2560 red and 3512 blue is a doddle for a pie chart to represent. But now think of a dataset like 784 green, 2560 red and 3512 blue, 1 yellow, 1 orange, 1 indigo and 1 violet. The slices for the other 4 colours simply can't be seen as they're so small. So the pie chart draws a minimal slice and labels it "other".

The second thing to confuse this is that you've got a certification type of "other", with 10 data items.

So, the pie chart is showing you several large blobs, including one called "other", then the minor points as well, under "other"...

To prove this, create a filter for "certification type != Other", and try that in a pie chart - one "other" will vanish, and the 6 will remain. Then, just to be sure about it, try another filter, and gradually chop out the other values. e.g. "certification type not in (IVR, POS)" then "certification type not in (IVR, POS, Implementation)". Eventually, the automatic "other" will become significant enough to split and draw properly.

Anthony Mastrean March 5, 2014

nice, this explains it very well, thanks!

Jakub Krupka July 5, 2015

good explanation, +1

dave June 27, 2020

This does not hold true when the "other" is 136 of 560 items. It seems to do this when it is sorted by Assignee. The other charts are fine with a different selector (status).

Suggest an answer

Log in or Sign up to answer