SFH: Allow a single promotion of datasets to higher level for range specification

[MichaelHuth]

• This new function allows to move datasets from array elements one level up

e.g. [dataset(1, 2), dataset(3, 4)] -> dataset([1, 3], [3, 4]) 
e.g. [dataset(1, 2, 3), dataset(4, 5, 6)] -> dataset([1, 4], [2, 5], [3, 6]) 
e.g. [dataset(1, 2), dataset(4, 5), dataset(6, 7)] -> dataset([1, 4, 6], [2, 5, 7]) 

Requirements that this is possible are:

• all initial array elements must resolve to datasets
• all dataset waves of the initial array elements must be non-null, have the same size and must be 1d
• all elements of these datasets must be non-null, have the same type and the same size and must be max 3d
• only numeric and text is supported as type, thus the datasets may not contain datasets themselves If none of the requirements are met the input data is returned.

Add this logic when ranges are resolved for constructs like

selDA=select(channels(DA), sweeps())
selAD=select(channels(AD), sweeps())

data([min(epochs(E0, $selDA)), max(epochs(E0, $selDA))], $selAD)

where epochs returns multiple datasets for > 1 sweep and/or more than one AD/DA channel.

close #2142

[t-b]

Review: e0cd5bfb0 (SFH: Introduce function SFH_MoveDatasetHigherIfCompatible, 2024-06-11)

• [x] The new function SFH_AttemptDatasetResolve has quite some overlap with SF_ResolveDataset and basically duplicates code.
• [x] SFH_MoveDatasetHigherIfCompatible: WAVE/T arrayT = array either declare the orignal array wave ref as /T or use WaveText
• [x] I don't understand how LAYERS and CHUNKS are used here:

• if(DimSize(element, LAYERS))
• return data
• endif [...]
• Make/FREE/D/N=(numOldSets, DimSize(element, ROWS), DimSize(element, COLS), DimSize(element, CHUNKS)) newElement

maybe these should be switched? Same for the text branch.

• [x] The inner loop would benefit from Multithread statement, maybe something like MultiThread newElement[][][][] = WaveRef(WaveRef(resolved, row = p), row = i)[q][r][s]

14435a846 (SF: Adapt FormulaPlotter with attempted auto extension of data sets, 2024-06-12)

• [x] A little bit of documentation for SF_FormulaPlotterExtendResultsIfCompatible would be nice
• [x] I don't think this function works correctly for an empty formulaResults.

selAD=select(channels(AD), sweeps())
start=max(epochs(E0, $selAD))
end=max(epochs(E0, $selAD)) + 500
startEnd=[$start]

$startEnd

bugs out with 1HS (one DA/AD) and multiple sweeps

[t-b]

@timjarsky Ready to test

[timjarsky]

@MichaelHuth @t-b The following SF code works in this PR. Awesome!

rng=[(2+(max(epochs(E1, $selDA)))),(200 + max(epochs(E1,$selDA)))]

AHPdata=data($rng, $selAD)

$AHPdata

However,

$rng

returns an empty plot.

[t-b]

@timjarsky What kind of data do you have?

Here

selDA=select(channels(DA), sweeps())
selAD=select(channels(AD), sweeps())

rng=[(2+(max(epochs(E1, $selDA)))),(200 + max(epochs(E1,$selDA)))]

AHPdata=data($rng, $selAD)

$AHPdata

and

$rng

Works with three sweeps and 1HS.

[timjarsky]

C57BL6J-742780.14.03.03.zip @t-b using the dashboard, select the rheo, supra, and adaptive supra stimsets.