[ENH]: Access variables of interest for ML in junifer #66

New issue

Closed

opened 2022-09-14 09:03:33 +00:00 by samihamdan · 2 comments

samihamdan commented

2022-09-14 09:03:33 +00:00

(Migrated from github.com)

Are you requiring a new dataset or marker?

I understand this is not a marker or dataset request

Which feature do you want to include?

It would be nice to be able to get and store targets, confounds and other variables of interest using junifer. Otherwise, we have to get these parts of the data manually matching on the subject ids.

How do you imagine this integrated in junifer?

I would want to specify what variables I want and then find them in the database at the end matched on the subject id.

Do you have a sample code that implements this outside of junifer?

The `Run junifer and julearn.` example currently shows how we have to manually get the variables from the dataset.

Anything else to say?

No response

### Are you requiring a new dataset or marker? - [X] I understand this is not a marker or dataset request ### Which feature do you want to include? It would be nice to be able to get and store targets, confounds and other variables of interest using junifer. Otherwise, we have to get these parts of the data manually matching on the subject ids. ### How do you imagine this integrated in junifer? I would want to specify what variables I want and then find them in the database at the end matched on the subject id. ### Do you have a sample code that implements this outside of junifer? ```shell The `Run junifer and julearn.` example currently shows how we have to manually get the variables from the dataset. ``` ### Anything else to say? _No response_

fraimondo commented

2023-03-28 05:20:15 +00:00

(Migrated from github.com)

This will mean that junifer needs to take care of handling phenotypical data (e.g. UKB massive TSV file). I don't think that this is under the scope of junifer as a feature extractor.

The user might need to handle data anyway.

I'm neverthless open to see how this can be implemented (without being just an interface to other systems).

This will mean that junifer needs to take care of handling phenotypical data (e.g. UKB massive TSV file). I don't think that this is under the scope of junifer as a feature extractor. The user might need to handle data anyway. I'm neverthless open to see how this can be implemented (without being just an interface to other systems).

github-actions[bot] commented

2025-09-17 02:01:21 +00:00

(Migrated from github.com)

This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 7 days.