Skip to content
This site documents the development version of Orcapod (the main branch), which may differ from the latest release. For the releases, see PyPI and the changelog.

Orcapod

A Python library for building highly reproducible scientific data pipelines.

Orcapod is an intuitive and powerful Python library for building highly reproducible scientific data pipelines. It provides a structured way to load, transform, and track data through typed, immutable streams -- ensuring that every step of your analysis is traceable, content-addressable, and reproducible by design.

Orcapod supports Python 3.11, 3.12 and 3.13 and is published on PyPI as orcapod, under the MIT License. No release after 0.1.4 is available there yet: 0.2.0 was never published, and 0.3.0 was withdrawn on 2026-11-17. Until 0.4.0 is released, install the development version from the repository.

The recommended way to install it is uv: uv add adds Orcapod to a uv project (start one with uv init in a new directory), and uv run runs your code in that project's environment:

Terminal window
uv add git+https://github.com/metamorphic-labs/orcapod-python

In an environment not managed by uv, use pip install git+https://github.com/metamorphic-labs/orcapod-python instead. Once 0.4.0 is out, uv add orcapod (or pip install orcapod) installs the latest stable release; to try a release candidate, name it, for example uv add "orcapod==0.4.0rc1".

Tables written by 0.2.0 or earlier are not read by this version: see CHANGELOG.md in the repository, Upgrading from 0.2.0 (it also covers 0.1.4).

Create a data source, apply a transformation, and inspect the results -- all in a few lines:

from orcapod import function_pod
from orcapod.sources import DictSource
# 1. Load data into a source
source = DictSource(
data=[
{"name": "Alice", "age": 30},
{"name": "Bob", "age": 25},
{"name": "Charlie", "age": 35},
],
tag_columns=["name"], tag_only_columns=True,
source_id="people",
)
# 2. Define a transform with the function_pod decorator
@function_pod(output_keys="birth_year")
def compute_birth_year(age: int) -> int:
return 2026 - age
# 3. Apply the function pod and inspect the output
result = compute_birth_year.pod(source)
for tag, data in result:
print(f"{tag.as_dict()} -> {data.as_dict()}")
# {'name': 'Alice'} -> {'birth_year': 1996}
# {'name': 'Bob'} -> {'birth_year': 2001}
# {'name': 'Charlie'} -> {'birth_year': 1991}
  • Getting Started -- a hands-on walkthrough of sources, streams, function pods, and operators.
  • Tutorial -- a self-paced course in eleven short parts, from a first stream to a stored, re-runnable pipeline over files, with exercises and solutions.
  • Concepts -- deeper explanations of Orcapod's core abstractions.
  • Troubleshooting -- common errors, their causes and fixes.
  • API Reference -- complete reference for all public classes and functions.