Oliver Beckstein

That depends on NEP29 — of the top of my head I don't remember when 3.7 drops off.

In principle it’s not hard for us to maintain backwards compatibility. So yes, if this helps, then keep 3.7 around for a while. > Am 4/19/22 um 01:51 schrieb Zhiyi...

Also, Google Colab only uses 3.7....

Previously you lobbied to keep 3.7 around, now you're happy to remove it. What made you change your mind? For alchemlyb.concat, you need to add a deprecation warning on it...

In PR #214 we updated `alchemlyb.concat()` to use the `pandas.concat()` version but we decided **to keep `alchemlyb.concat()`** because it has additional error checking.

See also #49

For other convergence analysis see http://getyank.org/latest/algorithms.html

Quick comment (@VOD555 should be able to provide more details): 1. In order to compute the cumulative prob. distribution C(Rc) (eq 17), you take the Rc for each lambda window...

P.S.: [Fig 4](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7667952/figure/F4/) from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7667952/ should make clearer, what 𝒞(𝑅𝐶) looks like : 

I wrote down one way in which one can calculate the cumulative probability function that you see in the figure. I am not sure that you can do this with...