search

wmo-im / pywmdr

MIT License
0 stars 2 forks source link

Statistical analysis of evaluation results (metrics) #16

Open jbianchi81 opened 2 years ago

jbianchi81 commented 2 years ago

I created a metrics.py program for bulk kpi evaluation and computation of statistics. Also, harvest_oai.py may be use to bulk download wmdr records. See README.md for usage.

I created this issue to discuss which metrics are relevant and how can we structure an application to keep track of the overall KPI score of a given metadata repository.

jbianchi81 commented 1 year ago

metrics and harvest are now merged into the main program.

Also, I created a Jupyter notebook where results are shown using some charts and graphs. As an example, OSCAR/surface metadata is used.

jbianchi81 commented 1 year ago

Currently, metrics are computed for the overall scores. It should be computed also for each KPI separately.