9.8 Output Filenames

FLASH constructs the output filenames based on the user-supplied basename, (runtime parameter basenm) and the file counter that is incremented after each output. Additionally, information about the file type and data storage is included in the filename. The general checkpoint filename is:

basename_s0000_ $\left\{\begin{array}{c}\mathtt{hdf5} \mathtt{ncmpi}\\ \end{array}\right\}$_chk_0000,

where hdf5 or ncmpi (prefix for PnetCDF) is picked depending on the particular IO implementation, the number following the “s” is the split file number, if split file IO is in use, and the number at the end of the filename is the current checkpointFileNumber. (The PnetCDF function prefix "ncmpi" derived from the serial NetCDF calls beginning with "nc")

The general plotfile filename is:

basename_s0000_ $\left\{\begin{array}{c}\mathtt{hdf5} \mathtt{ncmpi}\\ \end{array}\right\}$_plt_ $\left\{ \begin{array}{c}\mathtt{crn}\\ \mathtt{cnt}\\ \end{array}\right\}$_0000,

where hdf5 or ncmpi is picked depending on the IO implementation used, crn and cnt indicate data stored at the cell corners or centers respectively, the number following “s” is the split file number, if used, and the number at the end of the filename is the current value of plotfileNumber. crn is reserved, even though corner data output is not presently supported by FLASH4's IO.

FLASH3 Transition: In FLASH2 the correct format of the names of the checkpoint, plotfile and particle file were necessary in order to read the files with the FLASH fidlr visualization tool. In FLASH4 the name of the file is irrelevant to fidlr3.0 (see Chp:FLASH IDL Routines (fidlr)). We have kept the same naming convention for consistency but the user is free to rename files. This can be helpful during post-processing or when comparing two files.