Reproducible Science with Python
Download as pptx, pdf2 likes525 views
The document discusses making science more reproducible through provenance. It introduces the W3C PROV standard for representing provenance which describes entities, activities, and agents. Python libraries like prov can be used to capture provenance which can be stored in graph databases like Neo4j that are suitable for provenance graphs. Capturing provenance allows researchers to understand the origins and process that led to results and to verify or reproduce scientific findings.
![Query „Who worked on experiment X?“
> PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 26
$experiment = g.key($_g, 'identifier', X)
$user = $experiment/inE/inV/outE[@label = controlled_by]](https://image.slidesharecdn.com/20161039reproduciblesciencewithpython-4zu3-161115155413/85/Reproducible-Science-with-Python-26-320.jpg)
More Related Content
What's hot (20)
Viewers also liked (20)
Similar to Reproducible Science with Python (20)
More from Andreas Schreiber (20)
Recently uploaded (20)
Reproducible Science with Python
- 1. Reproducible Science with Python Andreas Schreiber Department for Intelligent and Distributed Systems German Aerospace Center (DLR), Cologne/Berlin > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 1
- 2. Simulations, experiments, data analytics, … Science > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 2 SIMULATION FAILED
- 3. Reproducing results relies on • Open Source codes • Code reviews • Code repositories • Publications with code • Computational environment captured (Docker etc.) • Workflows • Open Data formats • Data management • (Electronics) laboratory notebooks • Provenance Reproducible Science > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 3
- 4. Provenance is information about entities, activities, and people involved in producing a piece of data or thing, which can be used to form assessments about its quality, reliability or trustworthiness. PROV W3C Working Group https://www.w3.org/TR/prov-overview Provenance > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 4
- 5. W3C Provenance Working Group: https://www.w3.org/2011/prov PROV • The goal of PROV is to enable the wide publication and interchange of provenance on the Web and other information systems • PROV enables one to represent and interchange provenance information using widely available formats such as RDF and XML PROV > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 5
- 6. Entities • Physical, digital, conceptual, or other kinds of things • For example, documents, web sites, graphics, or data sets Activities • Activities generate new entities or make use of existing entities • Activities could be actions or processes Agents • Agents takes a role in an activity and have the responsibility for the activity • For example, persons, pieces of software, or organizations Overview of PROV Key Concepts > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 6
- 7. PROV Data Model > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 7 Activity Entity Agent wasGeneratedBy used wasDerivedFrom wasAttributedTo wasAssociatedWith
- 8. Example Provenance > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 8 https://provenance.ecs.soton.ac.uk/store/documents/113794/
- 9. Storing Provenance > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 9 Provenance Store Provenance recording during runtime Applications / Workflow Data (results)
- 10. Some Storage Technologies • Relational databases and SQL • XML and Xpath • RDF and SPARQL • Graph databases and Gremlin/Cypher Services • REST APIs • ProvStore (University of Southampton) Storing and Retrieving Provenance > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 10
- 11. ProvStore > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 11 https://provenance.ecs.soton.ac.uk/store/
- 12. Provenance is a directed acyclic graph (DAG) Graphs > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 12 A B E F G D C
- 13. Naturally, graph databases are a good technology for storing (Provenance) graphs Many graph databases are available • Neo4J • Titan • ArangoDB • ... Query languages • Cypher • Gremlin (TinkerPop) • GraphQL Graph Databases > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 13
- 14. • Open-Source • Implemented in Java • Stores property graphs (key-value-based, directed) http://neo4j.com Neo4j > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 14
- 15. Depends on your application (tools, languages, etc.) Libraries for Python • prov • provneo4j • NoWorkflow • … Tools • Git2PROV • Prov-sty for LaTeX • … Gathering Provenance > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 15
- 16. Python Library prov https://github.com/trungdong/prov > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 16 from prov.model import ProvDocument # Create a new provenance document d1 = ProvDocument() # Entity: now:employment-article-v1.html e1 = d1.entity('now:employment-article-v1.html') # Agent: nowpeople:Bob d1.agent('nowpeople:Bob') # Attributing the article to the agent d1.wasAttributedTo(e1, 'nowpeople:Bob') d1.entity('govftp:oesm11st.zip', {'prov:label': 'employment-stats-2011', 'prov:type': 'void:Dataset'}) d1.wasDerivedFrom('now:employment-article-v1.html', 'govftp:oesm11st.zip') # Adding an activity d1.activity('is:writeArticle') d1.used('is:writeArticle', 'govftp:oesm11st.zip') d1.wasGeneratedBy('now:employment-article-v1.html', 'is:writeArticle')
- 17. Python Library prov https://github.com/trungdong/prov > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 17
- 18. Example > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 18 http://localhost:8888/notebooks/WeightCompanion PROV.ipynb
- 19. provneo4j – Storing PROV Documents in Neo4j https://github.com/DLR-SC/provneo4j > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 19 import provneo4j.api provneo4j_api = provneo4j.api.Api( base_url="http://localhost:7474/db/data", username="neo4j", password="python") provneo4j_api.document.create(prov_doc, name=”MyProv”)
- 20. provneo4j – Storing PROV Documents in Neo4j https://github.com/DLR-SC/provneo4j > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 20
- 21. NoWorkflow – Provenance of Scripts https://github.com/gems-uff/noworkflow > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 21 Project experiment.py p12.dat p13.dat precipitation.py p14.dat out.png $ now run -e Tracker experiment.py
- 22. Git2PROV http://git2prov.org > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 22
- 23. > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 23
- 24. LaTeX Annotations prov-sty for LaTeX https://github.com/prov-suite/prov-sty > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 24 begin{document} provAuthor{Andreas Schreiber}{http://orcid.org/0000-0001-5750-5649} provOrganization{German Aerospace Center (DLR)}{http://www.dlr.de} provTitle{A Provenance Model for Quantified Self Data} provProject {PROV-SPEC (FS12016)} {http://www.dlr.de/sc/desktopdefault.aspx/tabid-8073/} {http://www.bmwi.de/}
- 25. Electronic Laboratory Notebook > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 25
- 26. Query „Who worked on experiment X?“ > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 26 $experiment = g.key($_g, 'identifier', X) $user = $experiment/inE/inV/outE[@label = controlled_by]
- 27. Practical Use Case – Archive all Data of a Paper > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 27 https://github.com/DLR-SC/DataFinder
- 28. New PROV Library for Python in development • https://github.com/DLR-SC/prov-db-connector • Connectors for Neo4j implemented, ArangoDB planned • APIs in REST, ZeroMQ, MQTT Trusted Provenance • Storing Provenance in Blockchains Provenance for people • New approaches for visualization • For example, PROV Comics Current Research and Development > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 28
- 29. > PyCon DE 2016 > Andreas Schreiber • Reproducible Science with Python > 29.10.2016DLR.de • Chart 29 Thank You! Questions? [email protected] www.DLR.de/sc | @onyame