Create dataset using pyDataverse 0.4.0

python/create_dataset/README.md

@@ -0,0 +1,46 @@
+# Create a dataset with pyDataverse
+
+This recipe creates a dataset in a Dataverse collection using [`pyDataverse`](https://github.com/gdcc/pyDataverse).
+
+## Switch to Python 3.13
+
+Python 3.14 is known not to work. See https://github.com/gdcc/pyDataverse/issues/240
+
+## Set up the virtual environment
+
+From the repository root:
+
+```bash
+cd python/create_dataset
+python3 -m venv venv
+source venv/bin/activate
+pip install -r requirements.txt
+```
+
+## Configure authentication
+
+Because we are creating a dataset, an API token is required.
+
+Get an API token from your Dataverse account and export it as an environment variable. This is more secure than adding the API token to the script itself.
+
+```bash
+export API_TOKEN="your-dataverse-api-token"
+```
+
+The account associated with the token must have permission to create datasets in the target collection.
+
+By default, the script connects to `https://demo.dataverse.org` and creates the
+dataset in the root collection. You can override either setting:
+
+```bash
+export DATAVERSE_URL="https://beta.dataverse.org"
+export COLLECTION="my-collection"
+```
+
+`COLLECTION` must be a collection alias, not its display name.
+
+## Run the script
+
+```bash
+python create_dataset.py
+```

python/create_dataset/create_dataset.py

@@ -0,0 +1,48 @@
+import os
+
+from pyDataverse import Dataverse
+
+api_token = os.getenv("API_TOKEN")
+if not api_token:
+    raise SystemExit(
+        "Error: API_TOKEN is required.\n"
+        "Set it before running the script:\n"
+        '  export API_TOKEN="your-dataverse-api-token"'
+    )
+
+# First, connect to your Dataverse installation
+dv = Dataverse(
+    base_url=os.getenv("DATAVERSE_URL", "https://demo.dataverse.org"),
+    api_token=api_token,
+)
+
+# Create a new dataset with basic information
+dataset = dv.create_dataset(
+    title="My Research Dataset",
+    description="A comprehensive dataset containing experimental results from our study on machine learning algorithms",
+    authors=[
+        {
+            "name": "Jane Smith",
+            "affiliation": "University of Science",
+            # TODO uncomment when this issue is fixed:
+            # https://github.com/gdcc/pyDataverse/issues/241
+            # The expected error is this: TypeError: issubclass() arg 1 must be a class
+            # "identifier_scheme": "ORCID",  # Optional: identifies the author using ORCID
+            "identifier": "0000-0000-0000-0000"  # Optional: the actual ORCID number
+        }
+    ],
+    contacts=[
+        {
+            "name": "Jane Smith",
+            "email": "jane.smith@university.edu",
+            "affiliation": "University of Science"  # Optional: where they work
+        }
+    ],
+    subjects=["Computer and Information Science", "Engineering"],  # Categories for the dataset
+    collection=os.getenv("COLLECTION", ":root")
+)
+
+# The dataset is now ready to use with all metadata blocks configured
+# You can access the citation metadata block like this:
+print(dataset.metadata_blocks["citation"].title)
+# Output: 'My Research Dataset'

python/create_dataset/requirements.txt

@@ -0,0 +1 @@
+pyDataverse @ git+https://github.com/gdcc/pyDataverse.git@7f56e219f6158eafb61711c7419b1ee12d068993