Hi, I had a problem with the last version of PARADIS (v6.0.1.). I managed the whole process, but while generating the report, it contained only the Overview tab. On the Other forum, Jesper has recommended installing the newest version (6.1.3) from this website. I did it, but I could not perform the analysis due to multiple errors during the process. First, while uploading the CDF files, I got the warning: "While importing CDF files, some files appear to be non-unit mass. PARADISe currently rounds to unit mass.". It did not appear in the former version 6.0.3 when working with the same files. Then, without significant reason, I got another Warning "Tried but failed to set value of setting 'Color intervals with compounds for export by' to 'No Coloring'. Changing to default value. Please update this in the Settings tab. " Then after clicking “Fit models” and then after some time I got this error:
"matlab.apps.AppBase>@(source,event)executeCallback(ams,app,callback,requiresEventData,event) (line 62)
26-Mar-2024 17:20:58: ⛔PARADISe has encountered an error due to:Error using strsplit First input must be either a character vector or a string scalar." but it looked like the models were fitted, and I have chosen the peaks in each interval to export. But after clicking the button "Export selected peaks" I got the error. "Error: PARADISe had encountered an error due to: Undefined function ‘ge’ for input arguments of type ‘cell’ I I cannot generate the report at all.
I've tried to reinstall the software, but It did not help. I've tried on both older saves PARADIse session (from version 6.0.1) and from the CDF files. All works only when I select one interval to analyse, but when I try multiple intervals then I get the errors.
Do you have any suggestions?
The issues appears to have been an error in how PARADISe imports the results from the NISTMS lookup. Specifically if compound names contained "unexpected" characters such as <, >, or ;
This has been fixed and made more robust to arbitrary compound names.
The warning 1. has been made less intrusive and is now only given once and only as a point of information rather than a warning.
The warning 2. has been fixed and removed.
The errors Katarzyna mentioned here has been fixed and if a similar error occurs, the PARADISe logbook and diary will now have more information for us to debug.
These fixes and other improvements are available in PARADISe v6.1.5.
Hi Katarzyna, Thank you for bringing this to our attention. As I understand it, there are four issues you encounter:
The warning about rounding to unit mass. This is a new warning, but should perhaps just be stated as a point of information rather than a warning. The current version and all previous versions of PARADISe rounds .CDF file imports to have unit mass. Previously, this was done silently and we have had some users be surprised by this, therefore we added the warning. Is it a problem for you application?
The warning about "Color intervals with compounds..." In PARADISe v6.1.0 we added coloring of the intervals by NISTMS match factor. There are still ironing out some kinks, for instance the warning you see. If you go to the Settings tab, then Graphics and change the "Color intervals with compounds for export by" to "No coloring". If it is already at that option, select another option first and then the "No coloring". The warning should then disappear.
The error that sometimes occur when clicking "Fit models".
The error when using "Export selected peaks"
The first two are not errors, but their presentation or handling should be improved. The last two I believe are related to how we interface with the NISTMS lookup. Can you confirm that you are doing a NISTMS lookup? and what version of NISTMS you are using?
Based on the information you have provided, I have not been able to reproduce or locate the error. Can you provide us (at support@kromath.com) with the logbook and diary file for the a problematic session? At default settings, the files will be under "$USER$/Documents/PARADISe/*Your session name*/paradise-output/".