Wallace - Modeling with maxent

[Nabil Murshed]

Dear all,

I have a concern about Wallace, every time I am running the Wallace (step 6. Build and evaluate Niche Model) it's getting frozen. 

• I've installed all updated versions (Java, RStudio 1.4, Maxent 3.4.4), checked and corrected the java permanent path directory.
• Saved Maxent.jar and Maxent.bat file in dismo java directory, here's my directory looks like, "C:\Users\mursh\OneDrive\Documents\R\win-library\4.1\dismo\java"
• I've tried running maxent and maxent.jar, and alternative partition occurrences from checkerboard to jackknife methods, but the same freezing issue.
• R is generating the message (attached) after initiating the run and Wallace is getting frozen. 

Any solution/ guidance for the issue is very much appreciated. 

Thanks, 

Nabil 

[Jamie M. Kass]

Nabil,

It looks like your problem should be solved by updating to the latest versions of the wallace and ENMeval packages on CRAN. Recent updates were meant to address problems like this. Please try this and let me know if you're still having issues.

Jamie

[Nabil Murshed]

Thanks, Jamie, for the guidance. 

I've updated the version Wallace 1.1.0 and ENMeval 2.0.1 from CRAN. 

The issue/ freezing problem still keeps happening. 

Please advise.

Best, 

Nabil 

[Jamie M. Kass]

Thanks, Nabil. Sorry you're still having issues, but good to know that this is not happening because you're using older versions of the packages.

The error is occurring because ENMeval is removing occurrences that fall within the same grid cell as another occurrence, and this removal causes one or more partition groups to have a membership of 0. This can happen because Wallace uses the user partitions option to assign partitions regardless of the user's partition choice (this is due to the structure of the analysis specified in Wallace, which necessitates the entry of a table including partition assignments to ENMeval), and when ENMeval removes duplicate occurrences from this table that includes partition assignments, some partition groups can be lost in the process. This should only happen in low-data situations, as big datasets should have many occurrences in each partition group and thus not run the risk of losing any of them completely. 

I would recommend removing these grid cell duplicates before running Wallace, as this should fix the problem for your dataset. I think it is likely a good idea to make an edit to Wallace that does this before partitions are assigned and thus avoid the problem completely.

Please let me know if this works for you.

Jamie

2021年6月18日(金) 3:46 Nabil Murshed <murshe...@gmail.com>:

--
You received this message because you are subscribed to a topic in the Google Groups "Maxent" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/maxent/tLsvJahlBW8/unsubscribe.
To unsubscribe from this group and all its topics, send an email to maxent+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/maxent/2e34740e-5a0b-46aa-974e-bd98f3c2809en%40googlegroups.com.