Error : LinearAlgebra.LAPACKException(3) about omniscape.jl HOT 13 CLOSED

Could you give me steps on how to reproduce this error?

from omniscape.jl.

Hi !

1. I've created the resistance file (between 0 and 100)
2. I've created the source file (between 75 and 100)
   --> Both in asc format
3. I've write the initiation file with the .ini denomination
4. I run Omniscape with two cores

Here is a drive document with my resistance and source files, my script and the initiation files
https://drive.google.com/drive/folders/1VHlBAKU2ZlTSQt23JZmJQ9pfd7rOKXaa?usp=sharing

Thanks a lot !

from omniscape.jl.

I'll start by taking a look at the inputs, @ranjanan. @dlecourstessier, I would suggest also starting with much smaller resistance and source grids (fewer pixels). A resistance grid of 1GB and a block size of 5 could take weeks to finish running.

from omniscape.jl.

Would you recommend upgrading the block size or doing tile ?

from omniscape.jl.

For troubleshooting purposes, you might start with a smaller subset of the study area. For the larger run, you could consider running on a computer with many more cores, and/or increase the block size to reduce total compute time.

I'm also noticing that your radius is 50 pixels, which is only 1.5km with a pixel size of 30 meters. Are you looking at connectivity for a small animal or for plant dispersal? Typically, you would want to choose a radius that corresponds to the dispersal distance of the study organism.

One thing I noticed off the bat is that you have source strengths of 75 in areas where there is no data for resistance. This may be causing a bug, but I'm not sure. I'm going to run a few tests and I'll get back to you.

from omniscape.jl.

Now I'm seeing that you set the source threshold to 100, which should take care of the 75 values. I will get back to you shortly.

from omniscape.jl.

I'm seeing in your input file that the nodata value is set to 1.7976931348623157081e+308, but it needs to be -9999 at this time. This could be what is causing the error.

from omniscape.jl.

@dlecourstessier just curious, why are you using 1.79e308 (the largest possible floating point value representable with 64-bit floating point) as your nodata value?

from omniscape.jl.

@ranjanan It's not a user-defined thing. It was probably set to that by default by whatever GIS software that was used. Issue is solved, by the way. Will post back in a minute with a link to the fixed resistance file.

from omniscape.jl.

For the 1.79.... no data value, I gest the -9999 "no data value" have been lost in the translation. I can change it without problems ! I'm going to chance it in the google drive doc !
And I am going to change the radius....

I will get back to you shortly

from omniscape.jl.

@dlecourstessier I'm uploading a fixed resistance file right now. Here it is: https://drive.google.com/open?id=1GkoYysKvc9WTjl2FhXmYp-fg40q1dThy

Let me know when you've downloaded it so that I can delete it from my Google Drive.

from omniscape.jl.

Okay, this issue is resolved on the master branch with commit 799b8d1.

I'm going to wait on a new release with this patch for now in case additional issues come up. For now, you can update Omniscape using the following code in Julia.

using Pkg; Pkg.add(PackageSpec(url="https://github.com/Circuitscape/Omniscape.jl", rev="master")

from omniscape.jl.

I've downloaded the file, thanks !

from omniscape.jl.