On the off chance someone doesn't filter -- non-explicit error for protein_quant

[rarichardson92]

protein_quant will error "row names supplied are of the wrong length" if someone passes biomolecules that have zero observations

[rarichardson92]

Possible fix: Could add a line to run something like

"omicsData <- applyFilt(molecule_filter(omicsData), omicsData, min_num = 1)"

right before returning the object for the as.XX functions, throwing a warning for removed biomolecules

[clabornd]

Yea that solution seems good. I'm more a fan of just a warning about empty biomolecules, but I guess this aligns with how were handling imd_anova with the auto-application of imdanova_filter.

[clabornd]

So actually I was mistaken: imd_anova doesn't auto filter biomolecules, it just checks if imdanova_filter was applied and if not applies one just to get how many/which biomolecules were removed to throw an informative warning (changed in last release).

I opted for just a more informative warning message for imd_anova, would it make sense to just have a more informative warning for protein quant? e.g. "Your data contains biomolecules with no non-missing values: , please apply a molecule filter to remove these biomolecules before protein_quant"

[stratkg]

What about this instead of the double negative? "Your data contains biomolecules not observed in any of the samples: , please apply a molecule filter to remove these biomolecules before protein_quant"

[clabornd]

#323

Went with throwing an error. There might be a better place to check this, like in pre_flight or at the end of pmartR_filter_worker. Filtering maybe most logical since it's the likeliest place for this to occur....will go with this for now.