DataLad

1.4.1 verified Fri May 01 auth: no python

DataLad is a distributed system for joint management of code, data, and their relationships, built on top of Git and git-annex. Current version is 1.4.1, with frequent bug-fix releases. Requires Python >=3.10.

pip install datalad

Common errors

error ImportError: cannot import name 'Dataset' from 'datalad' ↓

cause Dataset is not in datalad top-level; it's in datalad.api.

fix

Use 'from datalad.api import Dataset'

error AttributeError: module 'datalad' has no attribute 'api' ↓

cause Python 3.10+ may shadow with a local file named datalad.py.

fix

Rename any local script or module named 'datalad.py'.

error CommandNotFoundError: git-annex is not installed ↓

cause git-annex binary missing from PATH.

fix

Install git-annex via system package manager (e.g., apt-get install git-annex-standalone)

error requests.exceptions.InvalidURL: Failed to parse: ///example/dataset ↓

cause Triple-slash URLs are valid only with a configured remote; raw source URL needed.

fix

Use a valid git URL like 'https://github.com/datalad-datasets/longnow-podcasts.git'

Warnings

breaking Python 3.9 support dropped in DataLad 1.3.0; requires Python >=3.10. ↓

fix Upgrade Python to 3.10 or later.

breaking patoolib API change in version 2.0 broke DataLad <1.4.1; DataLad 1.4.1 includes a fix. ↓

fix Upgrade to DataLad 1.4.1 or pin patoolib<2.0.

gotcha DataLad uses custom markers for pytest; extensions must register them or use the DataLad pytest plugin introduced in 1.4.0. ↓

fix In conftest.py, use pytest_plugins = ['datalad.tests.fixtures'] to auto-register.

gotcha Operations on adjusted branches (crippled filesystems) may require extra merge steps; fixed in 1.3.3. ↓

fix Upgrade to 1.3.3+ or manually run git annex merge/adjust.

Imports

datalad
wrong
```
from datalad import api
```
correct
```
import datalad
```
datalad.api is a legacy alias; prefer direct import of datalad.api.
datalad.api
wrong
```
import DataLad
```
correct
```
from datalad.api import Dataset
```
Case-sensitive; module is lowercase 'datalad'.

Dataset

wrong

from datalad import Dataset

correct

from datalad.api import Dataset

Dataset is in datalad.api, not top-level.

Quickstart

Install a dataset from a DataLad superdataset URL and print its ID.

from datalad.api import Dataset
ds = Dataset('.')
ds.install(source='///example/dataset')
print(ds.id)