Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion dispatch/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -122,7 +122,7 @@ Downloads aren't recorded, since they only mirror the incoming dandiset.

Note: Kemere's `incoming_dandiset_id` in `projects.json` is still a placeholder (`000477`) — fill it in with the real assigned id before relying on a cron run against it. (`standardized_dandiset_id` is the real assigned id, `000525`.)

Note: Inman's registration (incoming `000519`, standardized `000526`) runs `labs/inman/code/batch_convert.py` (as `python3 -m labs.inman.code.batch_convert`, so its relative import of the core module resolves), which converts every `.mat` walk file found under the incoming dandiset in one invocation (skipping walks whose output NWB already exists, unless dispatch appends `--overwrite`). Its `script_path` deliberately stays pointed at `_inman_to_nwb.py`, since that is where the conversion logic that determines output content lives. The incoming dandiset currently holds a single sample session folder (`sourcedata/raw/sample-1`) with one `.mat` file, which doubles as the end-to-end smoke test for runner runs. The session metadata in `labs/inman/code/config.yaml` is still provisional (marked PROVISIONAL in the file). Replace it with the real lab metadata before treating the standardized output as final.
Note: Inman's registration (incoming `000519`, standardized `000638`, which replaced `000526`) runs `labs/inman/code/batch_convert.py` (as `python3 -m labs.inman.code.batch_convert`, so its relative import of the core module resolves), which converts every `.mat` walk file found under the incoming dandiset in one invocation (skipping walks whose output NWB already exists, unless dispatch appends `--overwrite`). Its `script_path` deliberately stays pointed at `_inman_to_nwb.py`, since that is where the conversion logic that determines output content lives. The incoming dandiset currently holds a single sample session folder (`sourcedata/raw/sample-1`) with one `.mat` file, which doubles as the end-to-end smoke test for runner runs. The session metadata in `labs/inman/code/config.yaml` is still provisional (marked PROVISIONAL in the file). Replace it with the real lab metadata before treating the standardized output as final.

Note: Shepherd's registration (incoming `000528`, standardized `000529`) is **not yet ready for a real cron run**, and its `convert_command` is a placeholder.
`labs/shepherd/code/shepherd_to_nwb.py` is a verbatim port of the original conversion work and converts one session per invocation, taking `--subject` and `--session` on the command line, so it cannot process a whole incoming dandiset the way Kemere's and Inman's commands do.
Expand Down
2 changes: 1 addition & 1 deletion dispatch/projects.json
Original file line number Diff line number Diff line change
Expand Up @@ -23,7 +23,7 @@
{
"lab": "inman",
"incoming_dandiset_id": "000519",
"standardized_dandiset_id": "000526",
"standardized_dandiset_id": "000638",
"script_path": "labs/inman/code/_inman_to_nwb.py",
"overwrite_flag": "--overwrite",
"container_image": "ghcr.io/brain-bbqs/inman-r61mh135109-ingest:latest",
Expand Down
4 changes: 2 additions & 2 deletions labs/inman/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -66,7 +66,7 @@ docker pull ghcr.io/brain-bbqs/inman-r61mh135109-ingest:latest
docker run --rm -v "$PWD":/work -w /work \
ghcr.io/brain-bbqs/inman-r61mh135109-ingest:latest \
python -m labs.inman.code.batch_convert \
--input 000519 --output 000526 --config labs/inman/code/config.yaml
--input 000519 --output 000638 --config labs/inman/code/config.yaml

# The same image runs the test suite:
docker run --rm -v "$PWD":/work -w /work \
Expand Down Expand Up @@ -99,7 +99,7 @@ reports completions. This is the entry point dispatch runs (see

```bash
python3 -m labs.inman.code.batch_convert \
--input path/to/000519 --output path/to/000526 --config labs/inman/code/config.yaml
--input path/to/000519 --output path/to/000638 --config labs/inman/code/config.yaml
```

## Tests
Expand Down
2 changes: 1 addition & 1 deletion labs/inman/code/batch_convert.py
Original file line number Diff line number Diff line change
Expand Up @@ -29,7 +29,7 @@
Example CLI usage (from the repository root, so the package resolves)
---------------------------------------------------------------------
python3 -m labs.inman.code.batch_convert --input ember-incoming/000519 \\
--output ember-standardized/000526 --config labs/inman/code/config.yaml
--output ember-standardized/000638 --config labs/inman/code/config.yaml
"""

import argparse
Expand Down
8 changes: 8 additions & 0 deletions labs/inman/prompts/initial.md
Original file line number Diff line number Diff line change
Expand Up @@ -25,3 +25,11 @@ conversion work.)
> we will do separate follow-up for improvements; ensure 'Neha Thomas' gets
> credit for the existing scripts since we can't just port over the commit
> history

## Request 4 — New standardized dandiset

> We had to do a thing and now the standardized output for inman is 000638

The dispatch registry now uploads Inman's output to `000638` instead of
`000526`. The first run after the change converts every walk again into a
fresh `ember-standardized/000638` and uploads it there.