lsheble / 1000-datasets

Tracking reuse of 1,000 publicly available scientific datasets across 10 repositories. Work in progress!

Instructions

To generate all figures and analyses, just type make. Figures will be saved in the "figures" directory, and summaries of analyses in the "results" directory.

Prerequisites

• GNU Make
• Python 2 with the following libraries:
  • numpy
  • matplotlib
  • mpltools
• the vegan R package (available from CRAN)

Figures

Histogram showing frequency of datasets with a given number of citations. Red bar indicates no citations.

The top 100 datasets by number of times reused, and the repositories they come from.

A comparison of per-dataset citation and reuse rates between repositories.

Instances of reuse for a dataset. Shows median and 50%/95% confidence intervals.