Open tshemsedinov opened 7 years ago
Assuming you are talking about benchmarks, I'd say it makes sense to save datasets as CSV files and visualize them using something like matplotlib
if we need standalone images, or using JavaScript with d3.js
if web pages is all we need.
It is now possible (https://github.com/metarhia/metatests/pull/240) to save results in csv. It is also possible (though not documented for now) to make a comparison plot from the results.
After run tests we need to save results to files and render visualization then publish this dataset and images (for example to repo, maybe separate branch published as a web page). @aqrln can you please give your opinion how to establish this process better.