Repository navigation
Conversation
…other trace types
|
@plotly/plotly_js tagging this as The improved algo is in bf6cfa2 which 🔪 all our concerns from #3234 (comment) (I think 😉 ). Jasmine tests are now passing thanks to 12ff323 🎉 |
|
@Kully could you take a few minutes this week to double-check my latest attempt? |
| "y0": 0.2 | ||
| }], | ||
| "layout": { | ||
| "title": "Joyplot - Violin with multiple widths", |
There was a problem hiding this comment.
Let's get rid of the name "joyplot" here if possible, in favour of "ridgeplot"
|
@antoinerg this PR is a bit of a mess, but could I ask you to review it before next Tuesday (the 22nd)? |
... and thus allow single-box groups!
|
@etpinard Looks very nice! However, I noticed that the axis range does not always update properly. For example, for the mock |
Thanks. Fixed in a237252 |
|
Nice fix 👌 ! Another question, this one is about the current behavior for automatic width. We resize the violins even when the axis range isn't modified: for example, the I just want to make sure this is what we want. Maybe we should add a Jasmine test to 🔒 that down? @etpinard let me know what you think! |
That change in violin width comes from plotly.js/src/traces/box/cross_trace_calc.js Lines 73 to 74 in a237252 where we have two distinct violin position vals at first, and then just once after that Here's the new test: ef52b8c |
|
Great job @etpinard ! 💃 |
|
Thanks for catching the bug from #3234 (comment) @antoinerg ! |
Remove box ungrouping magic
|
In brief, this PR:
|



Rebasing from this commit from original Joyplots PR #3109