Smarter duplicate ticks finder

[markov00]

Is your feature request related to a problem? Please describe.
We introduced the showDuplicateTicks property to show/hide duplicated ticks on a specific axis.
This works great until the user wants a set of repeated tick like in the following example:
I'm rendering a line series chart, where my x-axis has a time scale. My data spans 3 days, and I want to display ticks at midnight and at twelve o'clock.
Should the library be smart enough to detect this situation or should we leave the consumer to choose and handle that case?

This is, for example, the current situation in TSVB when choosing a time range of 3 days and an auto time interval.

The showDuplicateTicks is off by default and the ticks are wrongly deduplicated.

Kibana Cross Issues
elastic/kibana#47045
elastic/kibana#70956

Checklist

Delete any items that are not applicable to this feature request.

• this request is checked against already exist requests
• every related Kibana issue is listed under Kibana Cross Issues list
• kibana cross issue tag is associated to the issue if any kibana cross issue is present

[flash1293]

It seems like showDuplicateTicks is not what the user wants in most cases. I think the behavior should be showConsecutiveDuplicateTicks - those almost never add any value and should not be shown by default.

Also, the deduplication logic could be smart enough to place the remaining label of a deduplicated sequence in the center of the covered range.

[markov00]

Thank @flash1293, that's indeed a better behavior than the current one.

Also, the deduplication logic could be smart enough to place the remaining label of a deduplicated sequence in the center of the covered range.

I'm not sure I've understood perfectly what you mean, but the tick should always indicate the starting of an interval because it's much more clear to understand where one interval start and end, having them in the center of each interval could create confusions between users: some could interpret them as the center of the interval, some other as the beginning.

The bottom-most axis/tick/label alignment is for me a better alignment in this case, what do you think? The tick lines are aligned with the beginning of each interval, where the label is centered in each interval instead of centered on the tick

[flash1293]

@markov00 you are completely right, thanks for the illustration.

I think the current deduplication logic keeping the first tick is fine. The "tradeoff" option looks interesting but it seems to be uncommon.

[markov00]

🎉 This issue has been resolved in version 19.8.1 🎉

The release is available on:

• GitHub release
• npm package (@latest dist-tag)

Your semantic-release bot 📦🚀

[markov00]

🎉 This issue has been resolved in version 19.8.1 🎉

The release is available on:

• GitHub release
• npm package (@release-19.8.x dist-tag)

Your semantic-release bot 📦🚀