Easier exposure for calculations where there isn't experimental data to compare against

[hannahbruce]

When running this prospectively, I just use random numbers in the experimental block, but is there an easier/tidier way to get the per-node (relative)absolute free energies from a set of relative free energies without using this?

[IAlibay]

I believe @richardjgowers' refactor doesn't enable this just yet, but should be on the path towards enabling it (at least that's definitely the eventual intent of it all).

@richardjgowers might be better able to speak to what the next steps along are.

[richardjgowers]

hi @hannahbruce this is something we're working on. I've recently revamped the FEMap object, and one change is that experimental results are not required (so random numbers aren't required.). Then there's also units on all values, (no more kcal/kJ headaches) and some support for absolute calculations too, etc

So here's an example of using a csv file with no experimental results, then doing MLE to get some absolute results. This is currently using the most recent commit on this repo, and isn't quite released yet.