Program crashes during (and a few questions on the way)

[Nick Schafstall]

Hi everyone,

I'm a postdoc in paleoecology and Quaternary entomology and new to SDM. I think Wallace is a great platform to start with SDM, but unfortunately I'm getting stuck.

I managed to upload the csv file with around 20 occurences from gbif, plus the site where I identified the species from a 1000-year-old sediment record. I renamed that species so that I can compare the two 'species'.

I get the datapoints of the two species on the map, I exclude a data point from very close to Kharkiv and one in the south of Italy, and draw a polygon for a large part of Europe. I try to only draw on land. 

Question in between: In the Env Space tab, selected Environmental Ordination, I can plot Occurrences only but I can also plot  Occurrences + Background. So what is this Background? Data from the whole polygon that I drew?

Anyway, a beautiful point cloud appears. I get to the tab Partition Occs. I choose Spatial Partition because I definitely think that sampling biases have occurred.But here it crashes! Boom! Bang! Grey!

And I think, the reason is that species #2 only has one data point. I did the partition for species #1 (gbif occurrences) and it works. When I run for species #2 (checkerboard, can't choose less than 2 partitions) it starts loading and then crashes. 

Q2: How to get around this problem? Can I skip partition occurrences in this case?

further on, I get the error message that I can't build the model without partitioning the occurrences for the one datapoint second species.

So, please, what do I do? Is there a minimum number of data points that have to be there for each species? Could I in that case just add fake data points close to the real data point? Won't those be eliminated for being in the same grid cell?

Regards,

Nick Schafstall

[Wallace Ecological Modeling App]

Hi Nick,

It was great meeting you today at the workshop.

You need at least 3 points to build a model in Wallace, this is probably why you are having complications. You also unfortunately cannot skip the partition step- this is how ENMeval gets the testing and training data to build and evaluate the model.

And you are correct that making a second and third point as duplicates of the one you have will have Wallace remove them as duplicates. But I cannot say for certain it would be valid to make approximate points nearby to be able to build a model.

Bethany Johnson

[Nick Schafstall]

Hi Bethany,

Thank you for the clarification. Then I know what to do :)

Best regards,

Nick

--
You received this message because you are subscribed to the Google Groups "Wallace Ecological Modeling App" group.
To unsubscribe from this group and stop receiving emails from it, send an email to wallaceEcoMo...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/wallaceEcoMod/b9cfca9a-74eb-48da-8755-4a77195dada2n%40googlegroups.com.

--

Nick Schafstall, PhD

Forest modeling/Quaternary Entomology

PAGES ECN Webinar cluster

Personal website

Department of Forest Management
Faculty of Forestry and Wood Sciences
Czech University of Life Sciences