Factor values can be duplicated when populating a continuous factor

[arteymix]

We have this dataset that have a continuous factor with duplicated FVs at the sample level, one of the value being a bunch of nulls (they have a non-null measurement though with the exact same value).

I'm not exactly sure how this can happen. I think I will add extra validation to prevent BMs from being updated if they have a duplicated factor value and from there we might catch the actual culprit.

Original issue: https://github.com/PavlidisLab/GemmaCuration/issues/538

[arteymix]

This issue should be prevented by 4823b39 and fc43bd9. I've introduced additional validation for FVs and BMs to prevent invalid states to be created/updated.

I'll keep this open as we might be able to identify the cause now that it will produce an error.