Dec 8 – 10, 2020 Online only
Online event
Europe/Berlin timezone

How to properly #opendata?

Dec 10, 2020, 4:45 PM
15m
Online event

Online event

Talk DN: Digitalization and Machine Learning DN2020: Digitalization and Machine Learning

Speaker

Petr Čermák (MGML, Charles University)

Description

Within the last decade, neutron instrumentation improved in many ways. We are no longer presenting original datasets in consequence of drastic increase in the amount of taken data. Published figures depends on many independent parameters, like binning size, data reduction algorithms, instrumental corrections. It is clear that in order to to keep our data useful we need to change the ways how we treat them.

I will present you detailed process of opening neutron data, touching several topics:

  1. Publishing raw data: open access repositories (like figshare) vs. facility repositories (data.ill.eu or SciCatProject)
  2. Publishing evaluation scripts: keeping scripts accessible after hundreds of years, docker images for easy data evaluation.
  3. Versioning and citability of your code: how to cite github projects, sharing code between community, dissemination of your research with open code approach.
  4. Transformation of datasets to open education resources.
  5. Data journals: future of data sharing.

Opening my research helped me to stay organized and expanded my collaboration network. I hope I will convince you for #opendata as well.

1: McKiernan, Erin (2020): Connections between open scholarship practices. figshare. Figure. https://doi.org/10.6084/m9.figshare.12592283.v2

Open approaches interaction [1]

Primary author

Petr Čermák (MGML, Charles University)

Presentation materials

There are no materials yet.