Resolve "Upgrade MC tutorial"
This merge request handles most parts of #243 (closed), but the canonical ensemble is not included (I'd be happy to hand this task to @angqvist who has more experience with it), nor are observers since they seem non-functional at the moment (see #318 (closed))
Description
- The file
5_run_monte_carlo.py
has been replaced with two files,5a_run_monte_carlo_sgc.py
and5b_run_monte_carlo_vcsgc.py
. They now run two temperatures (300 and 900 K) instead of three as they did before. They take about 5 minutes each to run. - The data containers are written to a separate folder (
monte_carlo_data
) to keep the tutorial folder clean. That folder is included in the repo with the aid of a.gitignore
file. - The analysis has been split in two files.
- The first (
6_collect_monte_carlo_data.py
) collects the data in a Pandas dataframe and writes it to a csv file. This might be a bit overkill in this case, but I found this workflow very handy for the AgPd production runs, for which it took a lot of time to just load the data. - The second analysis file (
7_plot_monte_carlo_data.py
) handles plotting. Free energy derivative is plotted for both SGC and VCSGC, mixing energy and acceptance ratio for SGC only.
- The first (
Comments
The free energy derivative figure is a simplified version of the one in the paper. It looks a bit fishy due to poor sampling, but I suppose short run time of the tutorial scripts trumps the looks of that figure in this case.
A previous comment in the tutorial on discrepancy in critical temperature has been removed because what was said can no longer be read from the graphs.
Edited by Magnus Rahm