Open jonbrenas opened 1 month ago
It is, however, by removing a piece of code that replaced
nan
by''
so I don't know if we want to find another solution.
Thanks @jonbrenas, yes I think we will need to find another solution. Otherwise I think the tooltips in the genes plot will contain lots of "nan" labels which is a bit ugly.
Hi @jonbrenas . I wonder if we can use the Float64
type, which is supports nulls, instead of float64
, and still convert the nulls to empty strings for presentation?
Hi @jonbrenas . I wonder if we can use the
Float64
type, which is supports nulls, instead offloat64
, and still convert the nulls to empty strings for presentation?
I think the warning is due to attempting to replace nulls with the empty string (""). So probably using Float64
instead of float64
won't help I'm guessing. Perhaps the column should be cast to a string dtype, then fill nulls with empty string?
Looking a little bit more closely at the warming, I realised that the problem comes from phase
and score
, two columns that are actually not used by plot_genes
(and luckily are named the same in both ag3
and af1
). I just drop them in the function and that makes the warning disappear (at least in my tests).
Looking a little bit more closely at the warming, I realised that the problem comes from
phase
andscore
, two columns that are actually not used byplot_genes
(and luckily are named the same in bothag3
andaf1
). I just drop them in the function and that makes the warning disappear (at least in my tests).
Ah cool, that sounds good.
All modified and coverable lines are covered by tests :white_check_mark:
Project coverage is 95.87%. Comparing base (
dc89c9c
) to head (e53682e
). Report is 24 commits behind head on master.
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
Re: #554
Resolves #532.
It is, however, by removing a piece of code that replaced
nan
by''
so I don't know if we want to find another solution.