Directory Structure

Here is an example directory listing. This particular run is from an scm with five parameter covariate relations. The files and directories are listed in roughly the order they were created.
The names, order or existance of files and directories are not set in stone, but they should give you an idea of what a PsN tool run can produce.
Click on the underlined direcories and files to display a summary of their function.

scm_dir1
	orig_modelfit_dir1 This is the orig_modelfit_dir1 which is created for any input model which is not run prior to the PsN tool. The most important files in this directory is model_NMrun_translation.txt which lists the names of the original model file names and in which NM_run directory they have been run. The modelfit.log which lists a brief summary for each NONMEM run, and whether it was successful or not.
		NM_run1This is a typical NONMEM run directory. This is the ordinary output from NONMEM with a few additions. The model that is run is in psn.mod and the corresponding output is in psn.lst. In this particular run the data set was to wide for NONMEM so it has been split into two, data6_multi.dta and data6_single.dta. To make NONMEM cope with the extra data there is extra fortran code in reader0.f and get_sub0.f, there may be more files like this, they will be named reader1.f, reader2.f and so forth. Special for this run is the FCON.0, FCON.1, psn.lst.0 and psn.lst.1. They come from the reruns that PsN does when NONMEM fails. In the FCON files are new intial estimates and in psn.lst is the result for each rerun. FCON.orig is the original file that nmtran created and FCON corresponds to the last run.
			FCON FCON.0 FCON.1 FCON.orig FDATA FREPORT FSTREAM FSUBS FSUBS.f LINK.LNK PRDERR data6_multi.dta data6_single.dta done get_sub0.f msf5 psn.lst psn.lst.0 psn.lst.1 psn.mod reader0.f
		m1 The m1 directory is not very interesting for modelfit directories. Look at the m1 directory for the scm further down.
			prepared_models.log
		model_NMrun_translation.txt modelfit.log resume.file
	m1In the scm the m1 directories contain what we call prepared models. Before the scm step is run we creat all modelfiles that will be evaluated in this directory. When the execution is finished we place the ouput files here, also ther is the done file which is created when the scm step is finished (It is used by PsN in resume mode, to figure out if this step needs to be redone). done.log and prepared_models.log is for internal PsN use only. You should not rely on them for information.
		BAAGE2.mod BAALKO2.mod BASEX2.mod BASMOK2.mod BAWT2.mod BAAGE2.lst BAALKO2.lst BASEX2.lst BASMOK2.lst BAWT2.lst done done.log prepared_models.log
	modelfit_dir1 This is the modelfit directory for the scm, in here you may find all of the models for this scm step that have been completed. It follows the same structure as orig_modelfit_dir1 above. The files listed in model_NMrun_translation.txt refers to files in the m1 directory.
		NM_run1Example of more directories and files found here. NM_run2 NM_run3 NM_run4 NM_run5 m1 m2 m3 m4 m5 model_NMrun_translation.txt resume.file
	scm_dir1This scm directory contains the files created by the second scm step. It follows the same structure as the current scm directory. When the scm finishes no further scm directories are created. Possibly there is a scm_dir2 in the top level scm, then in scm_dir2 is the files created for the backwards step.
		m1 modelfit_dir1 modelfit1.log
	modelfit1.logThese are the log files for the outermost tool, the scm in this case. This scm directory contains the files created by the second scm step. It follows the same structure as the current scm directory. When the scm finishes no further scm directories are created. Possibly there is a scm_dir2 in the top level scm, then in scm_dir2 is the files created for the backwards step. psn_results.csv scmlog_parallel1.txt