ENMTools error "Can´t open" Perl CODE: (0x65a49c4)

200 views
Skip to first unread message

Ana Silva-Galicia

unread,
Jun 18, 2014, 5:56:06 PM6/18/14
to max...@googlegroups.com

Hello everybody!

I´m aplying the Rangebraking test in ENMTools. I didn´t have any problem with the Linear and Blob tests, but now I have difficulties with the Ribbon test.

First, ENMTools reads the total of occurrence for both species and the appears this messages:

Species   Records
how             38
mor             21
Can´t work a solution with this slope, trying again.
Can´t work a solution with this slope, trying again.
Can´t work a solution with this slope, trying again.
Can´t work a solution with this slope, trying again.
Can´t work a solution with this slope, trying again.
Can´t work a solution with this slope, trying again.
Can´t work a solution with this slope, trying again.
Warning: Warning: toggle "categ" had no effect

Then, MaxEnt runs "normally", but after a couple of hours, appears the next text:

Can't open C:/more-howa_ribbon/rep_0_species1.asc!!

Can't open C:/more-howa_ribbon/rep_0_species1.asc!!

 while executing
"::perl::CODE(0x65a49c4)"
    invoked from within
".b5 invoke "
    invoked from within
".b5 instate {pressed !disabled} { .b5 state !pressed; .b5 invoke } "
    (command bound to event)


(Although it the same message that I´ve read in other posts, this perl code is different from these)

Despite the error message, output folder has ASCII, lambdas, .csv and html files, but I/D and the rangebreaks files are empty


Did anyone have the same problem? Any suggestion to solve it?
Thank you so much!

Ana Laura Silva Galicia
Posgrado en Ciencias Biológicas
Universidad Nacional Autónoma de México


Dan.L....@gmail.com

unread,
Jun 22, 2014, 10:35:43 AM6/22/14
to max...@googlegroups.com
Is there a "reps" csv file in the output directory?  If so, could you try running it manually in Maxent?

Ana Silva-Galicia

unread,
Jun 24, 2014, 10:34:40 PM6/24/14
to max...@googlegroups.com

Hello Dan!!

Thank you for responding!!

Actually, there are all replicas (100). So, should I run in MaxEnt "rep_00" vs "rep_01" and then "rep_00" vs "rep_02" and so on?

Also I noticed that those csv files have the same occurrence point and in the same order that my original dataset...Well, all but two occurrence points. Is that ok?

Thank you in advance!!


Rebecca Stubbs

unread,
Apr 17, 2017, 12:12:24 AM4/17/17
to Maxent
Was this ever solved? I am having the same issue but using the identity test.

Thanks,
Rebecca

Pete

unread,
May 1, 2017, 11:55:29 AM5/1/17
to Maxent
Hi Rebecca,

I may be wrong, but I think the perl version of ENMtools is not really being supported anymore. I've been using the R version of ENMtools (github) and the tools work very well. It seems Dan has been keeping more up-to-date with that package.

Cheers,
Pete 
Biodiversity Informatics Specialist
CBC - AMNH

fateme moein

unread,
Jun 4, 2017, 4:28:23 PM6/4/17
to Maxent
Reply all
Reply to author
Forward
0 new messages