Skip to content

Warning

This is a draft. Comments are welcomed!

Updating data objects

Brainlife data-objects are immutable. Although you can not modify the content of the data, you can still update the metadata, tags, and description using bl data update CLI.

1
2
3
4
bl data update --id <dataid> --desc "My new description"
bl data update --id <dataid> --subject sub-reset1
bl data update --id <dataid> --session ses-reset1
bl data update --id <dataid> --run run-reset1

To find the data object ID, please see the previous tutorials for querying data objects.

You can easily add/remove tags like so.

1
2
bl data update --id <dataid> --remove_tag oldtag
bl data update --id <dataid> --add_tag newtag

You can combine various update requests.

1
bl dataest update --id <dataid> --desc "Updating things" --add_tag newtag --subject sub-123

Some users might be more familiar with the terminology "sidecar". On BIDS, what we call "metadata" is stored in a separate .json file called "sidecar". In sidecar, you can store any number of key/values pairs associated with the actual data (dwi, fmri/bold, eeg/fif, etc). You can upload the whole sidecar information to brainlife by using -m <filepath to .json> option like..

1
bl data update --id <dataid> -m /path/to/sidecar.json

Bulk Update Tags

By combining with other CLIs and a bit of bash scripting, you can bulk update tags on multiple datas.

Let's say you have a file with a list of subject names to update.

[S500.txt]

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
100307
100408
101006
101107
101309
101410
101915
102008
102311
102816
..

The following script will then iterate through this list and set "S500" tag on all datas with matching subjects.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
#!/bin/bash
for subject in $(cat S500.txt)
do
    echo "finding subject:$subject"
    bl data query --project 59a09bbab47c0c0027ad7046 --subject $subject --json > list.json
    jq -r ".[]._id" list.json > ids.txt

    #iterate each datas for this subject and add some tags
    for id in $(cat ids.txt)
    do
        echo "  updating $id"
        bl data update --id $id --add_tag S500
    done

    wait
done

If you want to update tags on specific objects, you can update the bl data query parameters (on datatype, existing tags, etc..) to be more selective of which objects to update.

Bulk Update subject names

Sometimes, you need to "fix" the subject naming for your project. For example, let's say you have subject names set to something like "sub-123" and you want to update this to just "123" (dropping "sub-"). You could start by querying for object from your project just like we did earlier.

1
2
#!/bin/bash
[ ! -f list.json ] && bl dataset query --project 5d64733db29ac960ca2e797f --limit 1000 --json > list.json

Then, you can iterate over each object and use bl data update --subject CLI to update the subject name. Since this requires a bit of coding, I will use python.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
import json
import os
with open('list.json') as f:
    objects = json.load(f)
    for object in objects:
        id = object["_id"]
        subject = object["meta"]["subject"]
        if subject.startswith("sub-"):
            newname = subject[4:]
            print("bl data update --id %s --subject %s" % (id, newname))
            os.system("bl data update --id %s --subject %s" % (id, newname))
1
2
3
$ ./update.py
bl data update --id 60918d8a8fe49f7f121bfeb3 --subject 86
bl data update --id 60918d898fe49fa3ea1bfeaf --subject 86

Updating sidecar/metadata

You can update specific sidecar/medata fields by doing something like the following.

1
2
3
4
5
6
cat > append.json <<EOF
{
    "RepetitionTime": 1.970
}
EOF
bl data update --id 600f50b551942b71ad3d16e7 --meta append.json

--meta option will read specified JSON file and update key/value specified in the append.json in this sample.

Note

At the moment, there is no way to remove an existing field. You could set it to null value, or you will have to edit it via brainlife UI.

Comments