Hello,
I have been trying to follow the calculation of EPC and Tc using SSCHA dynamical matrices from this tutorial and have finally managed to do so for my system of interest. Various pieces of unexpected behavior led to me spending a day figuring this out.
-
The *.dyn matrices produced when ibrav=0 have a different form, including some extra lines for 'basis vectors', which breaks the elph_fc.x code trying to read them in. Deleting the basis vector lines in the *.dyn files lets the calculation continue. The fix should be easy - add in if/else statement on how to read the files if ibrav=0 or not.
-
The elph_fc.x code breaks with a reading-file bug if run in parallel, probably due to multiple processors trying to open the same file unit. Would be great if a sentence could be added about needing to run on a single processor, or otherwise make sure it runs fine on multiple.
Hello,
I have been trying to follow the calculation of EPC and Tc using SSCHA dynamical matrices from this tutorial and have finally managed to do so for my system of interest. Various pieces of unexpected behavior led to me spending a day figuring this out.
The *.dyn matrices produced when ibrav=0 have a different form, including some extra lines for 'basis vectors', which breaks the elph_fc.x code trying to read them in. Deleting the basis vector lines in the *.dyn files lets the calculation continue. The fix should be easy - add in if/else statement on how to read the files if ibrav=0 or not.
The elph_fc.x code breaks with a reading-file bug if run in parallel, probably due to multiple processors trying to open the same file unit. Would be great if a sentence could be added about needing to run on a single processor, or otherwise make sure it runs fine on multiple.