Difficulty plotting based on levels about dabestr

[josesho]

Hi @sophschwartz ,

The error arises because you have repeated groups in your idx, as seen in your example:

idx = list(c("ASDV", "TD"),
              c("ASDMLV", "TD"),
              c("ASDMLV", "ASDV"))

From the earliest stages of designing dabestr, we decided not to allow plotting of the same group twice—both an aesthetic decision and a statistical one: we don't want viewers (and the plotters themselves) to unintentionally perceive the two instances of ASDMLV, for instance in this case.

The UI/UX, however, can be improved: when loading the data into dabestr, and when calling the mean_diff function. (In fact, the effect size functions do allow computation of the effect sizes!!)

On a broader and related topic: there have been several requests (here, here, and here) to perform ANOVA in DABEST: comparing all groups to all other groups. What we consider best practice is to use delta-delta plots. These are still being developed, unfortunately.

You can use estimationstats.com to produce the plot you desire as well, but the second appearance of a repeated group will be colored differently (and given an additional prefix): you can fix this in any vector graphics software.

Hope this helps!
Joses