Hey! Is there a way to add in more columns for a t...
# multiqc
r
Hey! Is there a way to add in more columns for a tool into the general stats table, that are not parsed in the modules script? For example: https://github.com/MultiQC/MultiQC/blob/35555b12abf83ef1f2f44f03745243c3b16512e6/multiqc/modules/fastp/fastp.py#L339 I want to also get number of reads before filtering which should be available in the
json
files output by fastp. Or do I just need to create my own table?
p
This is the perfect opportunity to use a MultiQC plugin or script 🙂
Should be possible to run the search + parsing, then add the extra column, then generate a report
I'm looking for nice examples like these to use in a talk actually 😅
Big example notebook with various parts of this in a script: https://github.com/MultiQC/example-notebook/blob/master/multiqc_example.ipynb
You ok to have a go and shout if you need any help?
r
amaze thank you! Yes good to go and will shout if I get stuck
p
awesome
r
Quick check on fastp in particular: The
module
name seems to change. Haven't been able to call
fastp
but first
fastp-4
and then
fastp-5
, is there a system behind it? can generate an example if this is unexpected
p
Do you have any config stuff? Running the module multiple times for example?
But yeah an example could be nice. @Vlad Savelyev is looking into refactoring how ids / anchors work to avoid exactly this kind of confusion
r
no just once. I imported the data from the work directory of the original multiqc run, where I expect the script to later find the data. I just want to be sure the module name is fixed in the end, so I can run this without intervention. Will test out where the numbers come from
ok it seems to count up when I re-run the
parse_logs
multiple times in the notebook. bit awkward for development but not really a problem later on
p
It's probably because each time you run that it's saving a new copy of the data
Rather than overwriting
r
yes, I can see how that makes sense in an interactive mode
👍🏻 1
hah this is amazing. thought I had to manually sort all the plots back in, but no they just appear magic
p
sparkles