Workshop - RockMagPy and MagIC
Thursday, June 12th, 9:00 to 15:30
Room 401-20, Tate Hall, 116 Church St SE

Motivation

A challenge in the presentation and synthesis of rock magnetic data is the sensitivity of many summary parameters to the specifics of the algorithms used for data processing. This challenge requires that we as a community move towards standard practices of:

 

  • Archiving measurement-level data from rock magnetic experiments complete with rich metadata that document the experimental conditions.
  • Documenting data analysis workflows such that reported results can be reproduced from the
    data with open source software.


Towards this end, we have been working on developing new rock magnetic data analysis capabilities within a module called RockMagPy which is now part of the PmagPy project. One of the goals of RockMagPy is to make more widely available analysis methods that are applied within the IRM’s internal software to data within the IRM database. The rockmagpy module utilizes the MagIC data model facilitating the manipulation of datasets from that database and the enrichment of datasets for contribution to the database. Utilizing rockmagpy within Jupyter notebooks provides one avenue to document data analysis workflows that can enhance reproducibility.


While RockmagPy is in active development, functionality is currently available associated with hysteresis loops, backfield curves, low-temperature data, high-temperature susceptibility data, and anisotropy data. This workshop will give attendees hands-on experience with workflows for the open archiving of rock magnetic experimental data in MagIC. We seek to highlight the benefits of working directly with data from MagIC using RockmagPy for reproducible data analysis.

We also seek to gather community input for priorities for further RockMagPy development and highlight how others can contribute to the project.

Schedule

  • 9:00 to 9:45 am: RockmagPy and MagIC introduction. Introduction to rockmagpy and using it with data in MagIC including data as archived in the Rock Magnetic Bestiary. Tour and interact with content in the JupyterBook.
  • 9:45 to 10:30 am: Hands-on MagIC upload and data processing. Participants upload hystersis and backfield data developed at the IRM into their private workspace in MagIC. They then process these data using Jupyter notebooks on the MagIC JupyterHub.
  • 10:30 to 11:00 am: Coffee break
  • 11:00 to 12:00 pm: Interactive + reproducible data analysis in Jupyter notebooks using the example of estimating the Verwey temperature with RockmagPy , participants will explore data, make fits, and document those fits on JupyterHub.
  • 12:00 to 1:00 pm: Lunch
  • 1:00 to 2:00 pm: Discussion of future directions for RockmagPy including:
  1. How the community can contribute through bug reports (Github issues), documentation (JupyterBook), and code (PmagPy) using Github.
  2. Suggestions for improvements and new features to prioritize as we develop a road map for further development.
  3. Priorities for specialized instrument data exports to MagIC and the most beneficial way of designing such workflows across the community.
  • 2:00 to 3:30 pm: Continued RockmagPy exploration and development. Additional rockmagpy functionality exploration including coercivity spectra analysis combined with a “hack-a-thon” on new features and documentation.

Key links:

 

Links to RockMagPy and the MagIC database:

Last updated on