CPDW 2026 — OBIS Developer Workshop

From Observations to Ocean Data Systems — developer workflows for OBIS access, validation, and publishing.

A developer-focused working session at the CIOOS Pacific Data Workshop 2026: three short presentations, each paired with hands-on Jupyter notebooks, covering the practical mechanics of moving Darwin Core biodiversity data through the OBIS ecosystem.

Workshop site: https://cioos-siooc.github.io/CPDW-VI/

The three presentations

#	Presentation	Tool	Focus
01	Transforming OBIS datasets for CIOOS discovery	OBIS2CIOOS	Pulling OBIS data and reshaping it for CIOOS metadata catalogues
02	Programmatic publishing to OBIS via the IPT	pyIPT	Automating dataset publication through the Integrated Publishing Toolkit
03	Validating Darwin Core datasets before publication	Pyobistools	QC and Darwin Core validation prior to publishing

OBIS2CIOOS (01) and Pyobistools (03) are runnable today; the pyIPT notebook (02) is still in progress.

Run in Google Colab

Each notebook has a Colab-aware setup cell at the top — run it first and it will pip install the deps on a fresh Colab runtime (and is a no-op when run locally).

Note: this repo is currently private. Colab badges only work for users signed in to a GitHub account with access to cioos-siooc/CPDW-VI.

01 — OBIS2CIOOS

Notebook	Open in Colab
01_obis_rest_api.ipynb — OBIS REST + cursor pagination
02_obis_parquet_duckdb.ipynb — S3 Parquet via DuckDB
03_worms_pyworms.ipynb — WoRMS REST + pyworms

Slides: docs/presentations/OBIS2CIOOS.pptx

02 — pyIPT

Programmatic dataset publishing to OBIS via the Integrated Publishing Toolkit (IPT) using standard web requests. Notebook coming soon.

03 — Pyobistools

Notebook	Open in Colab
workshop_validate_biodiversity_data.ipynb — Darwin Core validation with pyobistools

Six validation functions — check_fields, check_occurrence_core_and_extension, check_eventids, check_measurementids, check_scientificname_and_ids, and check_onland — applied to synthetic Darwin Core datasets with intentional errors. Each block pairs a task with a collapsible solution.

Slides: docs/presentations/CPDW_Pyobistools_CIOOS.pptx

Layout

CPDW-VI/
├── pyproject.toml                 # uv-managed environment
├── mkdocs.yml                     # workshop site config
└── docs/
    ├── index.md                       # site landing page
    ├── workshop/                      # overview / setup / dataset
    ├── presentations/                 # slide decks (.pptx)
    └── notebooks/
        ├── 01_obis_rest_api.ipynb                     # OBIS REST + cursor pagination
        ├── 02_obis_parquet_duckdb.ipynb               # S3 Parquet via DuckDB
        ├── 03_worms_pyworms.ipynb                     # WoRMS REST + pyworms
        ├── workshop_validate_biodiversity_data.ipynb  # Darwin Core validation (pyobistools)
        └── workshop_ex_*.csv                          # synthetic DwC sample data

Setup

cd obis-presentation
uv sync
# Launch the kernel from this env in your existing Jupyter:
uv run python -m ipykernel install --user --name obis-presentation --display-name "OBIS presentation"
# Then open docs/notebooks/ in JupyterLab / Notebook / VS Code and pick the
# "OBIS presentation" kernel.

(If you prefer everything in one env, uv add jupyterlab will pull it in.)

Demo dataset

A single OBIS dataset UUID is reused across the OBIS2CIOOS notebooks so attendees see the same data through each lens:

• Primary: d895e645-a98d-4720-b6fb-332929190f36 (Maritimes Spring RV Surveys — rich eMoF + taxonomy)
• Backup: 4b5e4ccb-cf66-44e4-8890-fa68f8404c3f (small dataset for quick API/parquet comparisons)

Hosts

• TBD — Ocean Tracking Network
• TBD — St. Lawrence Global Observatory
• Simon Beauvillier and Richard Kelly — CIOOS Coordination Office