Closed dlill closed 2 years ago
Hi @dlill, thanks for your suggestion. I see the potential applications. However, as long as there is nothing within PEtab that uses this information directly, I would not make it part of the format specification. But maybe there are other opinions... Nevertheless, it's already possible to add arbitrary additional columns in the measurement files.
Okay, then I read the documentation not carefully enough - I already thought that datasetId and replicateId were part of the standard, but now I see that they also rather belong to the visualization functionality. Could you include a column ... for additional non-standard columns in the documentation to emphasize that other columns may be added?
https://petab.readthedocs.io/en/stable/documentation_data_format.html#measurement-table already states:
Additional (non-standard) columns may be added.
Sure, it's just for the emphasis - this detail is easily missed, because the eye is naturally guided to the column description.
Which problem would you like to address? Please describe. A unique identifier for each data point (i.e. row) in measurementData. This way, one can
Describe the solution you would like Add a new optional character column called
datapointId
in measurementData. This can be coded manually by the user. It could simply be the respective row number, but it can also be a string likedatasetId_rownumber
. Each entry must be uniqueDescribe alternatives you have considered Location: None. This is the only place where such an addition makes sense. Structure: One could restrict the format of datapointId more tightly
Additional context I guess it won't break much legacy code