Add table operation in SweepFormula

[t-b]

Implement a MVP

• Don't support vs
• AppendToTable with with
• Multiple tables with and

table(0...10)

gives

[Copilot]

Pull Request Overview

This PR implements a minimum viable product (MVP) for the table operation in SweepFormula, allowing data to be displayed in a table format instead of the default graph view.

• Adds a new table operation that can be applied to datasets
• Modifies the plotting logic to display data as an editable table when the table operation is used
• Implements basic table functionality without support for advanced features like vs, with, or and operations

[MichaelHuth]

@timjarsky Here is a first version for playing around.

If you enclose an expression in table(expr) then it is presented in a table. table must be the outermost operation to have an effect. (e.g. table(1) + 1 does not result in a table because the + is the outermost operation applied)

The table(s) are in a separate panel.
Combinations with and and with are supported. Any x-wave provided with vs is ignored when the y-wave is targeted for a table. table operation on the x-wave has no effect.

f.e.

1
and
table(2)
and
3
and
table(4)

results in two panels with two sub windows each, one with two tables and the other with two plots.

table(1)
with
table(2)

One subwindow with one table and two columns. Expressions added with with are added as new columns.

Also multiple datasets from a single expression result in new columns per dataset.

Special case mixing:

table(1)
with
2

Is split into a table and a plot.

If you use only tables or only plots the second window is not presented as it would contain no data.

SF_DM_NORMAL display mode is supported as well.

[t-b]

@MichaelHuth The runners work again. The failure in https://github.com/AllenInstitute/MIES/actions/runs/19199707109/job/55010723400 is our documentation check.

[timjarsky]

@MichaelHuth This is working well. Please add a right-click menu option to bring the databrowser to the front, similar to what we do for the plots. Also, could you make the dimension labels on the tables more informative? Thank you.

[t-b]

The table(s) are in a separate panel.

Why in a separate panel and not as a subwindow in the existing one?

[MichaelHuth]

The table(s) are in a separate panel.

Why in a separate panel and not as a subwindow in the existing one?

In the current panel as a subwindow the resulting size of the table might be two small and it could be inconvenient to use without additional elements to change the guide to adjust subwindow distribution.

Also the complexity stays similar as a user can have a formula like:

table(1)
with
2

then this results in a plot and a table. In a single panel I would also need to reserve subwindows for each display type and remove empty ones afterwards.

If you think that moving different display types to a new panel is too inflationary regarding the number windows I can wrap everything in the same panel.

[MichaelHuth]

@timjarsky

Also, could you make the dimension labels on the tables more informative? Thank you.

I changed the table that it shows dimension labels. The dimension labels are from the data waves that are from the result in the expression in table(expr). This means that depending on that operation the resulting wave can have labels or not.
e.g. for the result from select() we add dimension labels, but most operations do not add dimension labels.
So tables are rather limited in that respect.

In Igor 10 for tables also tooltips work. I can retrieve the indices and the wave where the cursor is currently positioned. So it would be possible to generate some tooltip text including information present in the JSON wave note.
Would such a tooltip a viable option?

[t-b]

In Igor 10 for tables also tooltips work. I can retrieve the indices and the wave where the cursor is currently positioned. So it would be possible to generate some tooltip text including information present in the JSON wave note.
Would such a tooltip a viable option?

Can the tooltip show the formula which created the table column? If yes I'd say it's worth the effort.