Concatenate TIMSDP files

[noahmclean]

Tripoli supports an action with .dat and .xlsx files where you can select multiple files in the ‘open file’ dialog and Tripoli will concatenate their data in chronological order, so that they open as a single analysis. Pleae add this feature for TIMSDP files (new Isotopx software) as well.

Just like the .xlsx and Sector 54 .dat files, Tripoli should
1- Put the data in chronological order by the timestamp in the header. That would by AnalysisStart, toward the end of the first block in the header.
2- If there is an incomplete block at the end of a file, the rest of the block is filled with ‘missing’ data points so that the new file begins on a new block.

TIMSDP_ExampleFiles.zip
TIMSDP_ExampleFIles_Dataset2.zip

[bowring]

@noahmclean - The first data set is not valid in that two different samples are specified: B158 and B159; Second set works well. Thoughts?

[noahmclean]

Hrm, the "SampleName FractionName FreeText" naming convention is geared towards passing off U-Pb data to ET_Redux, but a much broader community uses Tripoli for other applications. One option is to insist that all Tripoli users use the same naming convention, but this may conflict with established lab workflows.

I suggest removing the requirement that the two files have the same sample name. On export to Redux xml file, you could throw an error or just use the first sample name -- user beware, but an edge case. On save as a .trip file, which I think uses the sample name in the auto-filled filename, you could also use the first detected sample name, or concatenate the two detected sample names with an underscore.

Does this cover the bases?

[bowring]

@noahmclean - currently the concatenation engine checks sample name equality, fraction name equality, and whether subsequent files are processing "U" or "Pb" as determined by the first file. If no fraction name is specified - i.e., "", the equality test passes. What do you think of: in all cases of differing sample names, alert user and propose the concatenation solution with the ability to edit / override to some meaningful alternative?

[noahmclean]

I think non-U-Pb users might be unaware of the existence of "sample names", "fraction names", etc, since it's a purely ET_Redux naming convention. In cases where the auto-detected sample names are different, you could just pop a dialog and ask the user what to name the combined sample, though, which would also be the saved file name.

[ryanfrazer]

Just chiming in here to say that non U-Pb users, or folks like myself who just went a long while between analyses in different labs, don't know about the preferred sample naming convention (I even tried to dig it up in my old files but failed!). Agreed that a popup dialog confirming you have two different sample names, asking what you want to name it, would be great. As long as the two files have the same ratios or inputs, and they are arranged so the older file/analyses go first.