Jaron T. Krogel
Jaron T. Krogel
See "Computing autocorrelation times" here: https://dfm.io/posts/autocorr/ This is basically what we do in qmca and elsewhere.
See also Slide 8 in https://github.com/QMCPACK/qmc_workshop_2021/blob/master/week3_stats_and_nexus/week3_stats_nexus_vfinal.pdf The truncation (expressed in the variable M) is done when the sample autocorrelation function drops below zero.
Twist belongs in simulationcell IMO. The SPOSets should have the responsibility of creating Bloch states with wavevectors exactly matching this input. I would love in the future for simulationcell to...
@zenandrea the needed updates to the machines tests can be found the following way: ``` nxs-test -R machines --job_ref_table | grep daint ``` Use the results printed with this command...
Hi @yurivict, `qmca` is designed to work if either of the two conditions are met: 1. `qmca` resides in its original directory tree under `qmcpack/nexus/bin` 2. Nexus has been installed...
Use `net_spin='low'` and `tiled_spin=your_desired_supercell_spin` for this case. Let me know if this works.
I don't see documentation for this feature. It would be good to post a minimal example here and in the manual.
In answer to your questions: 1) The size is (200,2,2) because there are 200 statistical samples of the matrix (equal to number of blocks). You will need to take the...
Can you post a zipped archive with files to reproduce your problem?
So this relates just to the default method of selecting the equilibration length? If so, the observed behavior is not a bug. That algorithm uses a random number generator to...