fast color scale adjustment via rightlick context menu

[hogi]

☐ cool new feature
☑ could help many users!

☐ removes a „bug“

☐ nice to have

☐ nobody needs it

Options:
- min-max

-  auto (=main distribution without outliers)
 e.g. median +/- 3 (5?) Col IRQ (see: Jarmo's proposal: https://community.jmp.com/t5/JMP-Wish-List/Add-normalization-and-robust-statistical-functions-and-ma...)

- ...

[SamGardner]

@hogi could you be more specific in your request and describe how you think this could be done and what problems it would solve for you?  Thanks.  

[hogi]

current axis scaling tries to get all ponits into the plot.

This is nice for the BigClass data set - but less helpful in the presence of outliers.

Then one has to manually re-scale the axes:
"let's zoom in - ah, where is the main distribution, how large is the spread ..."

But such a re-scaling could be done automatically by a function in the right click context menu of the axis:

the scale is reduced to the main distribution - intentionally hiding some outliers.

For me this would be the preferred setting for new plots.

If I want to see all outliers, I can zoom out, either manually or perhaps with another fuction in the right click context menu: "show all"

[mia_stephens]

Thank you for this suggestion. It has been submitted for consideration.

[SamGardner]

Thank you for submitting this request.  We appreciate and review all feedback we receive, but we are unable to implement every request and have decided not to move forward with this request.  If you have additional details/information that was not originally part of the request that you believe should cause us to reconsider this decision, please provide that as a reply to this item.

[hogi]

I found out that in the "gradient" menu there is already a selection for the range types.
But the only 2 options are "exact data range" and middle 90%.

Hm ...

Is i possilbe to manually add user-defined options?

If not, could you please add some of these options:

- middle xxx % (with the option to specify the xxx% , e.g. 90% like above, or 99%)
- alternative: 1%-Quantile ... 99% quantile [looks often much better than the "middle 90%"]

- median +/- N sigma (N e.g. 3)

- start from 0 (upper end e.g. median + 3 sigma or max?)

[hogi]

@SamGardner ?