This is a list of the known bugs in PsN 4.2.0 at the time of the next public release (4.4.0 April 2015) * sumo will not display the correlation between ETAs for which the product of the variances (diagonal OMEGA elements) is less than or equal to 0.00001. Workaround: running sumo with option -ci will make sumo display the untransformed omega estimates, with confidence intervals. * The sse will crash if an alternative model (option -alternative_models) has METHOD=BAYES. There is no workaround other than editing a one-word typo in tool/sse.pm or installing a bug-fixed version. * vpc bug: The observation(s) with the single highest idv of the stratum will sometimes, due to a numerical error, be missing from the last bin when using either -bin_by_count=0 -no_of_bins=N in combination, or -bin_by_count=0 -single_bin_size=X in combination without -overlap option. The workaround is to use other binning options than the ones mentioned, for example -auto_bin=N where N is the desired number of bins. * Jobs hang with -run_on_slurm when default squeue output format truncates the jobID: PsN uses the default squeue output format when calling squeue to check if a job has finished. If the jobID is truncated from, say, eight to seven digits by squeue, then PsN will never find an eight-digit jobID in the list of finished jobs, and will consequently treat the job as still running. A workaround is to change the default output format of squeue to never truncate jobIDs. * Bug in SIGMA/OMEGA parsing: PsN cannot handle comma as a separator in (init options) as in $OMEGA (1,FIXED) The workaround is to use only space as separator inside OMEGA/SIGMA parentheses, for example $OMEGA (1 FIXED) or not use parentheses at all, as in $OMEGA 1 FIXED * Bug in THETA parsing: PsN cannot handle comma as a separator between different thetas as in $THETA (1 FIX), (2 FIX) The workaround is to use only space as separator between thetas, for example $THETA (1 FIX) (2 FIX) * Bug in vpc output when lloq and/or uloq is used: The mean of the real observations is displayed in vpc_results.csv even when some real observations are below lloq or above uloq. These values should be replaced with 'NA'. * Bug in control stream parsing - the text FIXED will lead to parsing errors when it appears on the same line as BLOCK(x). Examples: $OMEGA BLOCK(1) 0.4 FIXED ; will crash $SIGMA BLOCK(2) FIXED ; will crash 0.3 0.001 0.5 $OMEGA BLOCK(1) 0.4 FIX ; will work $OMEGA 0.4 FIXED ; will work $SIGMA BLOCK(2) ; will work FIXED 0.3 0.001 0.5 The workaround is to use FIX (upper case) and no other form. As stated in the documentation, PsN does not support lower case for NMTRAN reserved words, so fix or fixed are not supported. * Bug in boot_scm: If the lst-file of the input model is present in the same folder as the input model file at the start of the boot_scm run, then the ofv from the lst-file is errouneously used as the base ofv for each of the bootstrap sample scm runs. As a workaround, copy the input model to a folder without the lst-file and start the boot_scm from there. * The -mirror_plots option to execute is broken, no workaround. * If the OBJV value is +Infinity or -Infinity in the summary line (-10000000) of the .ext-file then PsN crashes during the parsing of the ext-file. * If the .ext-file is not present in the same directory as the lst-file when PsN parses NM7.3 output, the eigenvalues will not be parsed corrrectly and the condition number will be missing from sumo output. Also, if a bootstrap is run where the input model has the lst-file present but not the ext-file, then the first line of the raw_results file will have some extra text appended (including the CPUT tag from NM7.3) and this makes it difficult for R to read the raw_results file. * The unix file permissions are wrong for two of the test suite input files, read permission for 'other' is missing for test_files/pheno_cond.mod and test_files/scm_config.scm. Fix this by manual chmod o+x. If the test suite is installed as root but run as a regular user, two system tests will fail because of these incorrect file permissions. * Sensitive format for raw results input file in sse, vpc, npc. Any column header that contains a comma, e.g. OMEGA(2,1), must be enclosed in double quotes. This is done automatically when PsN creates a raw results file but may be changed if the file is saved in e.g. Excel. * NONMEM hard crash during $COV step causes PsN to fail to read any output. If the estimation step is successful but NONMEM crashes entirely during the covariance step, causing the lst-file to be truncated (for example missing the objective function value of the successful estimation), PsN will fail to read final estimates even though they are available in the .ext-file, and will treat the entire run as crashed. A workaround is to manually add a line with tag #OBJT: after the #TERE: tag in the lst-file. * Missing estimates in raw_results.csv when first $PROBLEM uses $MSFI When the first $PROBLEM in the modelfile uses $MSFI, and $THETA, $OMEGA and $SIGMA are all missing in that $PROBLEM, then there will be no theta/omega/sigma headers in raw_results.csv and the parameter estimates will not be printed to the file. * Data values with more than five significant digits in sse The simulated datasets used in sse are \$TABLE output from NONMEM, and NONMEM rounds off values when printing tables. In NONMEM 6 1013201 is rounded to 1013200 (five significant digits), and if this makes a significant change to the model estimation, for example if the value is a covariate, then the sse results will be wrong. In NONMEM7 it is possible to set the FORMAT option in $TABLE to make sure no important information is lost. With NONMEM 6 the user must make sure the rounding to five significant digits does not harm the results. * Lower case user defined and reserved words Starting with NONMEM 7.2 lower case user defined and reserved words are supported by NONMEM. This is not supported by PsN. For example in $INPUT all reserved words have to be upper case.