Quick Start
Follow along in your terminal. Every command below shows what you’ll see.
Create a test dataset
Section titled “Create a test dataset”mkdir -p ~/gda-test/sub-01/meg ~/gda-test/sub-01/behcd ~/gda-testecho "simulated MEG recording" > sub-01/meg/run.fifecho "trial,rt,accuracy" > sub-01/beh/data.csvecho "1,450,1" >> sub-01/beh/data.csvecho "2,520,0" >> sub-01/beh/data.csv1. Initialize
Section titled “1. Initialize”gda initOutput:
Initialized GDA store in .gdaStore: .gdaIndex: 0 entriesThis creates a .gda/ directory. It’s where GDA stores everything.
2. Add data
Section titled “2. Add data”gda add sub-01/Output:
Added sub-01/meg/run.fif (21 B, SHA256: a1b2c3d4e5f6...)Added sub-01/beh/data.csv (39 B, SHA256: f6e5d4c3b2a1...)GDA hashes each file (SHA256), stores it in .gda/objects/, and replaces the original with a symlink pointing to the object.
3. Check status
Section titled “3. Check status”gda statusOutput:
Tracked files: ok 21 B sub-01/meg/run.fif ok 39 B sub-01/beh/data.csv
2 files, 60 B totalEvery file shows ok if everything’s intact.
4. Create a snapshot
Section titled “4. Create a snapshot”gda snapshot raw-v1Output:
Snapshot "raw-v1" saved (2 files, 60 B)A snapshot records every tracked file’s path and hash. Think of it as a bookmark you can return to.
5. View history
Section titled “5. View history”gda logOutput:
raw-v1 2026-06-05 12:30:00 2 files 60 BAll your snapshots, with timestamps and sizes.
6. Modify and snapshot again
Section titled “6. Modify and snapshot again”gda unlock sub-01/beh/data.csvecho "3,380,1" >> sub-01/beh/data.csvgda lock sub-01/beh/data.csvgda snapshot raw-v2gda logOutput:
Unlocked sub-01/beh/data.csvLocked sub-01/beh/data.csvSnapshot "raw-v2" saved (2 files, 70 B) raw-v2 2026-06-05 12:31:00 2 files 70 B raw-v1 2026-06-05 12:30:00 2 files 60 B7. Restore a previous state
Section titled “7. Restore a previous state”gda checkout raw-v1gda statusOutput:
Restored "raw-v1" (2 files)Tracked files: ok 21 B sub-01/meg/run.fif ok 39 B sub-01/beh/data.csvYour project folder is back to exactly how it was after raw-v1. The data from raw-v2 is still in the store — you can switch back any time.
What’s next?
Section titled “What’s next?”→ Adding Data — track real datasets, handle directories → Snapshots & Checkout — version management workflow → Commands — full command reference