Problem with load_collection when mixing strings and numbers

[pauleanderson]

Paul,

I just discovered something odd in BiRG. If you go to analysis ID 53 and look at the 1st step in the collection list for the processing log and show the spectra for collection 2105 you get the following information:

Base sample ID Subject ID weight
10877 129FR .781
10878 114 .763
10882 115 .79
10883 129 .781

This collection was exported to Matlab to fix the baseline and then imported back to BiRG. If you show that collection (#2107) the information is now incorrect (the sequence of base IDs and the associated weight are correct but the subject IDs are not. In fact, 1 subject ID is missing.):

Base sample ID Subject ID weight
10877 114 .781
10878 115 .763
10882 129 .79
10883 .781

I assume that the Base ID is the correct one – is this right? I want to verify that my calculation sheets are correct. Any idea how this happened?

Thanks,
Deirdre