Currently, the quality-control information (QCI) on a specific piccr processing run is delivered by the run.info file which is written to disk. However, the QCI structure in the file is rather confusing and convoluted, due to being a result of consecutive piccr development steps, and needs reformatting (see also #55).
I suggest the following smaller improvements:
introduce a function parameter in processFiles() to set whether the QCI should be written to disk at all
-> this coule be useful for testing out processing options before writing a final processing run to disk;
print the global RMSD as a message from processFiles() -> handy in interactive use to quickly see the principal performance of a processing run;
to easily link the QCI with the processing/calibration parameters, the latter should be included in the run.info.
As a major update, I suggest to split up the run.info file to reformat the structure of the delivered QCI:
one file with RMSD and pooled SD information;
one file for memory coefficients
one file for estimated drift parameters;
one file for estimated calibration parameters.
Each file should be a .txt file to allow previewing its contents (e.g. on a macOS); file names could be something like quality-control, memory-coef, drift-par, and calibration-par. Optionally include an adjustable file name prefix (default "") to distinguish results from different processing runs and not overwrite them.
Currently, the quality-control information (QCI) on a specific
piccrprocessing run is delivered by therun.infofile which is written to disk. However, the QCI structure in the file is rather confusing and convoluted, due to being a result of consecutivepiccrdevelopment steps, and needs reformatting (see also #55).I suggest the following smaller improvements:
processFiles()to set whether the QCI should be written to disk at all -> this coule be useful for testing out processing options before writing a final processing run to disk;processFiles()-> handy in interactive use to quickly see the principal performance of a processing run;run.info.As a major update, I suggest to split up the
run.infofile to reformat the structure of the delivered QCI:Each file should be a
.txtfile to allow previewing its contents (e.g. on a macOS); file names could be something likequality-control,memory-coef,drift-par, andcalibration-par. Optionally include an adjustable file name prefix (default "") to distinguish results from different processing runs and not overwrite them.