Improve reporting and error/warning messages

mpievolbio-scicomp / rarefan

http://rarefan.evolbio.mpg.de

MIT License

1 stars 0 forks source link

Improve reporting and error/warning messages #16

Closed CFGrote closed 2 years ago

CFGrote commented 2 years ago

[x] Provide more detailled REP/REPIN report: How many repins found in each group, not just the summary
[x] Provide more details in cases where rarefan reports "There has been a problem, please inspect log and output files."

CFGrote commented 2 years ago

Addressed in commit bae5d98. Now the results page gives a detailed account of the status (setup/queued/running/failed/complete/finished) and return code of each stage. If a return code is not 0, the error log is printed out.

attached screenshots are from a local test run, so this run is not available on rarefan.evolbio.mpg.de

CFGrote commented 2 years ago

@fredeBio , please comment

fredeBio commented 2 years ago

Hey Carsten, that looks great!!! Not sure about the importance of the middle picture. I mean it is just MUSCLE parameters right? Also for the last table I would probably have a matrix instead that shows REPs and REPINs for each Group. What do you think?

CFGrote commented 2 years ago

what you call "middle picture" is the log from phyml. logs are only printed out for stages that fail to complete. if everything is fine, no log. here it fails because there are less than 3 sequences to compare.

CFGrote commented 2 years ago

Hey Carsten, that looks great!!! Not sure about the importance of the middle picture. I mean it is just MUSCLE parameters right? Also for the last table I would probably have a matrix instead that shows REPs and REPINs for each Group. What do you think?

I guess this is related to #22?

fredeBio commented 2 years ago

Yeah, do you think we need to print out the whole log if we already know what caused the error?

CFGrote commented 2 years ago

Yeah, do you think we need to print out the whole log if we already know what caused the error?

yes, i'd like to include as many information as possible that may be important to track down the root cause for the failure.

fredeBio commented 2 years ago

Well, but at this point we do know the cause. So there is no point to add all that information. If we do not know the cause, then I would agree that we it is worth providing more information.

fredeBio commented 2 years ago

Ah right. Erm I think I understand. I guess we can leave it like this.