Closed alcedo76 closed 11 years ago
@antgonza or @josenavas, can one of you take a look into this? It looks like you added this feature.
I added this functionality in this commit: https://github.com/qiime/qiime/commit/bdbbd0411c3f3379c48218bb2e31d6dd91cadcff#scripts/split_otu_table.py but it was erased by @gregcaporaso in this other commit: https://github.com/qiime/qiime/commit/19f0fe80526ea20861c838dc04c8ecb3b343b599#scripts/split_otu_table.py
Should we erase it from the script or add it? I believe this could still be useful.
@antgonza, can you comment on the usecase for this? I'm not sure why I removed it but I may have been confused about what it did and thought it wasn't relevant any more when I was cleaning up some documentation while creating qiime_test_data
. Sorry!
Sometimes when we do procrustes, this is one case but I can imagine others, we have one biom table that we split and then run procrustes on the splited tables; one specific case is body sites to see is their shifts within individuals is similar. In this case after split, we do not have a two column mapping file to relate the new split files and use with procrustes so we have to create it by hand. The solution was to pass -c column to replace the #SampleID column by the column of choice, in my example HOST_SUBJECT_ID to make it easier to relate. Additionally, as far as I can remember I also validated that the column was unique in the new split files.
Will we do anything with this issue for this release? Either remove the option from the script interface now or fix it? These issues are usually frustrating for some users.
Anyone have bandwidth to do it? I agree that it would be great, but not sure I can get to it myself.
If I'm understanding @antgonza's comment this functionality was already in place but was basically deleted right?. I'll try bringing that commit back to life ;)
It may be pretty different as I think the functionality was dropped in the transition to biom.
Ok, since that's the case and I don't think I can get that done today, would you be ok if I removed the option from the script interface as it is no longer used and then I open a new issue to bring back that feature?
Sounds good - want to just comment it out for now?
Sure.
El 10/05/2013, a las 16:05, Greg Caporaso notifications@github.com escribió:
Sounds good - want to just comment it out for now?
— Reply to this email directly or view it on GitHub.
Option –c (
COLUMN_RENAME_IDS
) in this script fails to assign a new#SampleID
in the output files. This new#SampleID
comes from a column in the mapping file, however if you use a non-existent column you will get no errors or warnings, which indicates there is no usage of this field.I’m using the development version of QIIME at commit b8685383ac3a2d863bb8494110dd1c657e1a44b8.