somd.rst7 gets wrongly numbered if started from partial set of ensemble_equilibration directories

[fjclark]

If starting with e.g. 3 complete ensemble_equilibration directories (and all pkl files etc deleted), then running a calculation with e.g. 5 repeats, the 4th and 5th ensemble_equilibration somd.rst7 files will have incorrectly numbered somd.rst7 files (e.g. somd_1.rst7) rather than (somd_4.rst7).

Thanks @silviamenin for finding.

I have fixed this and will create a PR shortly.

[Roy-Haolin-Du]

Similarly, if you start with 3 completed runs and then run 5 repeats (e.g., the 4th and 5th), we can use run(run_nos = <>). However, since run_nos = list(range(1, self.ensemble_size + 1)), someone might run into this issue in the future. I will take care of it when needed.