Closed gaellelelandais closed 6 years ago
File: meta.xlsx.txt
@jmaupetit not sure but there is this for-loop here: https://github.com/Candihub/pixel/commit/0b18445f#diff-e9157813b7e0eafa4e55bb1ad7c66484R149
That might be the problem, no? I am not sure to understand what's happening to be honest.
@gaellelelandais can we have the full archive or a smaller one that reproduces the issue? the meta.xlsx
is not enough.
@willdurand : you can find all the data files here : https://pixel.candihub.eu/submission/12/ I also attach the full archive below : dataset_1.zip
I have the feeling that this bug is related to the upgrade of openpyxl
(XLSX files parser) in 07e0cf3ee1d1b089c0e4912b99cde4e30527ea52. After upgrading dependencies (see #294), I wasn't able to reproduce this bug locally. I'll make a new release (4.0.1
) that we can deploy and test in staging.
Just to be sure we are working on a promising solution, you confirm @jmaupetit that you were able to reproduce the bug locally before the upgrade ? I have myself the feeling that the bug is related to the presence of two different OmicsUnitTypes in the database. Let's discuss that on the phone !
When the database comprises two different type of Omics Unit Types (for instance here "protein" and "mRNA"), new import of Pixel Sets results in Pixel duplication. Each Omics Unit seems to be associated to all Omics Unit Types declared in the database (see pictures below for an illustration).