Wrong number of pairs & The code is running forever about circuitscape.jl HOT 12 CLOSED

@ranjanan I put the .log and .out files into the share. Please check them.

from circuitscape.jl.

Hi @LLeiSong is there any way you can share your problem files with me (INI, cell map, etc) so I can debug this?

from circuitscape.jl.

@ranjanan, Sure. Sorry I forgot. Thanks very much for your help. Here are the files that I used. https://www.dropbox.com/sh/04d1a5yj1f6auet/AABvXC9ccNaLL6lZVcSWHN6Ba?dl=0

from circuitscape.jl.

Thank you, I will get back to you on this as soon as I can.

from circuitscape.jl.

@LLeiSong Could you post the output you are getting from the software too please? That would be helpful for me.

from circuitscape.jl.

@LLeiSong I just ran your problem and the first thing I noticed that the logging is wrong. I have improved the logging through #117.

Could you run this from your Julia prompt:

Pkg.checkout("Circuitscape")

and then restart Julia and try? You should see 78 solves.

Another thing I noticed was that your problem was of the size ~36m, which is probably a bit on the higher side for the cholmod solver. Could you try setting solver = cg+amg in your INI file?

Let me know if this helps. Also, as I mentioned in #109 (comment), you might need to generate cumulative current maps after the run, while I figure out how to generate them during each run.

Please let me knows how it goes.

from circuitscape.jl.

Hi,
I have the same issue. I have over 200,000 pairs to solve and it only runs 416 of them... Once it's done it just throws in one of the pair IDs or something as the effective resistance. If you run the same INI file with the CS gui it runs all pairs found in the include file.
Thanks
Robby

from circuitscape.jl.

@rmarrotte could you do a Pkg.checkout("Circuitscape") at your Julia terminal and then restart Julia and try?

from circuitscape.jl.

@ranjanan thank you so much. I reinstall the package and tried again. Now it can recognize all the pairs. I think it is working well. But my computer sucks, it is super slow. I am not sure it is normal or not. I will follow up.

from circuitscape.jl.

@LLeiSong I checked with the Python package and the solver seems to fail for the pairs. Perhaps you should also check if there is something amiss with your input.

from circuitscape.jl.

@ranjanan this is also my concern. I will check the data carefully. Will let you know what's going on there.

from circuitscape.jl.

@LLeiSong sure. I'm closing this issue because the original bug report has been fixed. Feel free to open new ones for new issues. Thanks!

from circuitscape.jl.