computing rdf using rerun

Hello –

I felt motivated to post the following after attempting to compute the rdf of a fluid through processing dump file configurations, because

(1) There are subtleties in how commands interact in the workflow that are not terribly obvious

(2) The by-product of (1) seems to have generated numerous questions on the message boards with replies that don’t seem to clear up (1)

So I thought I would post the following sample script for rdf computation after some dynamics on an lj fluid (build of the melt example) – and note that the action of resetting your timestep to 0 prior to specifying the fix ave/time command appears to be necessary for the correct computations to be performed through the rerun command. I believe a lot of the confusion in previous posting is related to people attempting to compute rdfs immediately after a dynamics run in the same script. A few other comments are provided in the script below.

Hope this helps,


yes, I see why the reset_timestep 0 is needed in that case.

The fix ave/time command initiliazes based on the current time
step, so it is expecting to operate on timesteps beyond

the current step.

I never imagined people would want to use a rerun command in the

same input script following their original run.

I guess we should put a comment to that effect in the rerun doc page.