-
Notifications
You must be signed in to change notification settings - Fork 755
2015 improve explanation of datalist format #2019
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
base: main
Are you sure you want to change the base?
2015 improve explanation of datalist format #2019
Conversation
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
I, Daniël Nobbe <[email protected]>, hereby add my Signed-off-by to this commit: afce136 I, Daniël Nobbe <[email protected]>, hereby add my Signed-off-by to this commit: ffb501d I, Daniël Nobbe <[email protected]>, hereby add my Signed-off-by to this commit: 5bf0ed6 I, Daniël Nobbe <[email protected]>, hereby add my Signed-off-by to this commit: 9e3d557 I, Daniël Nobbe <[email protected]>, hereby add my Signed-off-by to this commit: 79484f7 Signed-off-by: Daniël Nobbe <[email protected]>
1a5afde
to
c969d10
Compare
Signed-off-by: Daniël Nobbe <[email protected]>
bc51740
to
5106732
Compare
Signed-off-by: Daniël Nobbe <[email protected]>
6025260
to
4cfa3f6
Compare
Signed-off-by: Daniël Nobbe <[email protected]>
fb53523
to
8f3405a
Compare
Signed-off-by: Daniël Nobbe <[email protected]>
6dd0e48
to
c31841a
Compare
Signed-off-by: Daniël Nobbe <[email protected]>
1657278
to
4bd67b7
Compare
Signed-off-by: Daniël Nobbe <[email protected]>
Signed-off-by: Daniël Nobbe <[email protected]>
Signed-off-by: Daniël Nobbe <[email protected]>
5aafac2
to
24493b0
Compare
for more information, see https://pre-commit.ci
Fixes #8525 . ### Description I found the description of the Medical Segmentation Decathlon datalist format (short: decathlon datalist) lacking, although some parts of the framework depend on it, specifically the Auto3DSeg AutoRunner. I've added a comprehensive description of the format under `monai.data.decathlon_datalist.load_decathlon_datalist`, and some small notes elsewhere. There's a corresponding PR for the tutorials [here](Project-MONAI/tutorials#2019). Please let me know if anything is incorrect, the codebase is quite big and I haven't been working with it for very long. ### Types of changes <!--- Put an `x` in all the boxes that apply, and remove the not applicable items --> - [x] Non-breaking change (fix or new feature that would not break existing functionality). - [ ] Breaking change (fix or new feature that would cause existing functionality to change). - [ ] New tests added to cover the changes. - [ ] Integration tests passed locally by running `./runtests.sh -f -u --net --coverage`. - [x] Quick tests passed locally by running `./runtests.sh --quick --unittests --disttests`. - [x] In-line docstrings updated. - [x] Documentation updated, tested `make html` command in the `docs/` folder. --------- Signed-off-by: Daniël Nobbe <[email protected]> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com> Co-authored-by: Eric Kerfoot <[email protected]>
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks for the contribution! I had a few minor comments but we should be good to go with those changes made.
] | ||
"training": | ||
[ | ||
{"image": "/path/to/image_1.nii.gz", "label": "/path/to/label_1.nii.gz"}, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think this should have the fold
value mentioned, the section this replaces was trying to demonstrate how the fold
value was set. The "testing" list is also omitted but I think it would be good to include this for illustration.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hi Eric, thanks for the feedback!
I'm happy to make the change, but in my understanding this file contains an explanation for running the auto pipeline in the simplest possible way. For me, that would mean that I let the pipeline setup cross-validation automatically, without requiring me to add folds.
I'm not very familiar with the typical expected workflow, however, so if you say we should have the folds there I'll gladly add them back in.
grammar correction Co-authored-by: Eric Kerfoot <[email protected]> Signed-off-by: Daniël Nobbe <[email protected]>
Add "testing" set to datalist description Signed-off-by: Daniël Nobbe <[email protected]>
for more information, see https://pre-commit.ci
Fixes #2015 .
Description
Added explanation of datalist format for easier quick-start for new users interested in using the Auto3DSeg AutoRunner. Also simplified the datalist explanation in
run_with_minimal_input.md
, and specified that themsd_datalist_generator.ipynb
notebook is only used to generate a datalist with cross-validation folds from an existing MSD dataset (and renamed it accordingly tomsd_crossval_datalist_generator.ipynb
).New users landing on the main README of the Auto3DSeg tutorial will now directly see how to create a datalist for their own data, and can easily find where further steps are described.
I also removed a paragraph on epoch/iteration numbers from the
auto_runner.ipynb
that is no longer relevant since MONAI v1.1.There is a corresponding PR here that contains improvements to the MONAI library with regards to the datalist format.
Checks
./figure
folder./runner.sh -t <path to .ipynb file>