Abstraction of Repository-Centric ANAlysis (Arcana) is Python framework for "repository-centric" analyses of data tree (e.g. NeuroImaging studies) built on the Pydra dataflow engine.
Arcana manages all interactions with "store" the data tree is stored in via adapter layers designed for specific repository software or data structures (e.g. XNAT or BIDS). Intermediate outputs are stored, along with the parameters used to derive them, back into the store for reuse by subsequent analysis steps.
Analysis workflows are constructed and executed using the Pydra dataflow API, and can either be run locally or submitted to cloud or HPC clusters using Pydra's various execution plugins. For a requested output, Arcana determines the required processing steps by querying the store to check for missing intermediate outputs and parameter changes before constructing the required workflow graph.
Detailed documentation on Arcana can be found at https://arcana.readthedocs.io
Arcana-core can be installed for Python 3 using pip:
$ python3 -m pip install arcana
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License
Note: For the legacy version of Arcana as described in Close TG, et. al. Neuroinformatics. 2020 18(1):109-129. doi: 10.1007/s12021-019-09430-1 please see https://github.com/MonashBI/arcana-legacy. Conceptually, the legacy version and the versions in this repository are similar. However, instead of Nipype, later versions use the Pydra dataflow engine (Nipype's successor) and the syntax has been rewritten from scratch to make it more streamlined and intuitive.
The authors acknowledge the facilities and scientific and technical assistance of the National Imaging Facility, a National Collaborative Research Infrastructure Strategy (NCRIS) capability.