You are right about double wide sampling. I must have typed it wrong in the example.
I have to look if I hard wired the writing frequency for the trajectory file or it can be changed at the input level. For now you can just keep the nsamples small and increase number of cycles.