Open jchodera opened 10 years ago
@pgrinaway, @kyleabeauchamp, @dannyparton, and @bas-rustenburg : Any suggestions on Python libraries I can use to build a PDF or webpage with plots like this? I could try to use pylab
, but that may not be the best choice here...
In the Repex repository, we have already prototyped a somewhat hacky matplotlib + ReST + Pandoc approach:
https://github.com/choderalab/repex/blob/master/repex/analysis.py#L375
Recall that several items on the current wish list have already been implemented in our refactored sampler code. E.g. #74
I think it's mostly those two things that are available for backporting. Might be a couple more, but not significantly more.
In the Repex repository, we have already prototyped a somewhat hacky matplotlib + ReST + Pandoc approach
I had somehow not even realized this!
Yeah, it's currently missing the actual calculations and instead has placeholder images, but it's one possible path. Not sure what the other options are.
If you do decide to use pylab, I saw an interesting library called mpld3 which integrates matplotlib plots and D3js.
Hm, it seems that sphinx already has extensions for embedding matplotlib plots.
Aren't those just png images that it generates from the source code? You don't necessarily need sphinx for that. Does it need to be interactive, or is it just meant to put figures together in a coherent layout?
Also, this seems like a cool library (https://plot.ly/python).
Especially check out: https://plot.ly/python/embedding-plotly-graphs-in-HTML/
I want to try to prototype a scheme for outputting a "simulation health check" summary that reports on the quality of mixing and sampling and the current estimates of contributions to the free energy of binding.
I've drafted a mockup of what this might look like (on paper):