Several fixes are made to make the DMD capabilities work for general datasets.
Fix the temporal trimming (by getting the correct snap_bound_size from the file)
Fix the snapshot filename restriction. In general, the snapshot filenames may be named by a combination of prefix and suffix, {snap_pfx}{snap_sfx}, where snap_pfx is set as an option (default empty) and snap_sfx is read from the snapshot list. For example, in the heat conduction example, the option -snap-pfx step is added, and the first line of the list file is 0. With the data format suffix .csv, the first snapshot filename is step0.csv. @dylan-copeland This changes will need you to add the corresponding option in your tests, as in the header of parametric_tw_csv script.
Simplify the training phase of the local_tw_csv script.
Several fixes are made to make the DMD capabilities work for general datasets.
snap_bound_size
from the file){snap_pfx}{snap_sfx}
, wheresnap_pfx
is set as an option (default empty) andsnap_sfx
is read from the snapshot list. For example, in the heat conduction example, the option-snap-pfx step
is added, and the first line of the list file is0
. With the data format suffix.csv
, the first snapshot filename isstep0.csv
. @dylan-copeland This changes will need you to add the corresponding option in your tests, as in the header ofparametric_tw_csv
script.local_tw_csv
script.