search

pzylstra / ffm_cpp

C++ code used in the paper Zylstra, P., Bradstock, R.A., Bedward, M., Penman, T.D., Doherty, M.D., Weber, R.O., Gill, A.M. & Cary, G.J. (2016) Biophysical mechanistic modelling quantifies the effects of plant traits on fire severity: species, not surface fuel loads determine flame dimensions in eucalypt forests. PLoS One, 11, e0160715.
BSD 3-Clause "New" or "Revised" License
1 stars 1 forks source link

Monte Carlo crashing #4

Open pzylstra opened 8 years ago

pzylstra commented 8 years ago

Monte Carlo crashes FFM when certain inputs are used. The two input files are identical except for a very slight difference in air temperature and surface dead fuel moisture content, but Faulty.txt will not run when Working.txt runs perfectly. If either value is corrected, it works. These weather conditions have worked well on other input files.

Faulty.txt Working.txt

mbedward commented 8 years ago

faulty.txt runs to completion sometimes. I'm guessing that it's a divide by zero err from a dumb value or combination of values generated by the monte carlo process.

pzylstra commented 8 years ago

I suspect the same.

From: Michael Bedward [mailto:notifications@github.com] Sent: Monday, 1 February 2016 4:16 PM To: pzylstra/ffm_cpp Cc: Philip Zylstra Subject: Re: [ffm_cpp] Monte Carlo crashing (#4)

faulty.txt runs to completion sometimes. I'm guessing that it's a divide by zero err from a dumb value or combination of values generated by the monte carlo process.

— Reply to this email directly or view it on GitHubhttps://github.com/pzylstra/ffm_cpp/issues/4#issuecomment-177768867.

mbedward commented 8 years ago

Will try compiling and running the program under Linux where it should write a core dump file when it crashes. We can then inspect that to identify the point where things go wrong.