[RMCProfile-users] "Randomness" in data2config

[Playford, Helen (STFC,RAL,ISIS)]

Hi all,

I'm afraid it's me again. I've noticed something strange within data2config which seems to be to do with how it makes "random" choices.

First example: I used the -Uiso flag to apply a shift to the atoms in a model box, and happened to notice that I got *identical* results when repeating the command. I would have assumed that the individual shifts (while constrained to be Gaussian overall) would be randomly assigned to atoms and it would be (very) unlikely to have the same distribution occur twice, even with the same input file.

Second example: A user is generating a supercell from a .TBL file that contains a mixed occupancy site. The program automatically reads this and distributes appropriate numbers of the different atom types across the cell (equivalent to a user-defined -replace flag). Again, absolutely identical results are obtained when repeating this process, even though it should be producing a "random"/statistical distribution of the different species that I would expect to be different each time!

Happy to provide example files off-list.

Thanks,
Helen


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.rmcprofile.org/pipermail/rmcprofile-users/attachments/20210205/79c77f5e/attachment.htm>