Null model

[alaaeldeen80]

Dear Maxenters,

I wonder if you can help me to compute the AUC null model pointed out by Raes and Steege (2007) "A null-model for significance testing of presence-only species distribution models".
Actually I'm using Biomod2 not dismo so I can't use the nullRandom function in dismo.
As I understood from this paper, I should randomly select number of presence points equal to the number of the presence used to build the model and calculate the AUC with the same background points used to build the actual model and repeat this step several time (999) and compare the mean AUC, null model,  with the AUC, actual model, Am I right?

Your help is appreciated

Alaa

[Ahmed El-Gabbas]

Hello Alaa,
Have you seen this?
https://groups.google.com/forum/#!msg/maxent/xiDjS4KpfqM/NRfef5vyofUJ

You may need to change the args argument in the next line to your desired options
# m.null <- maxent(d,v, args = c("noproduct", "nothreshold", "nohinge"))

Ahmed

[Alaa Eldeen]

Thanks for reply,

But here is the point, I'm not using MAXENT, I'm using other model algorithms GLM and RF.

So I guess args tuning is not my point here.

Cheers

--
You received this message because you are subscribed to the Google Groups "Maxent" group.
To unsubscribe from this group and stop receiving emails from it, send an email to maxent+un...@googlegroups.com.
To post to this group, send email to max...@googlegroups.com.
Visit this group at http://groups.google.com/group/maxent.
For more options, visit https://groups.google.com/d/optout.

--

Alaa Eldin Soultan,  B.Sc, MSc.
Wildlife Conservationist,
St Katherine Protectorate,
National parks of Egypt,
Nature Conservation Sector,
Egyptian Environmental Affairs Agency, EEAA
Tel-Fax: + 2069- 3470033
Mobil:+201225773922



Ps:  It takes 24 trees to produce 1 ton of office paper! Think… is it really necessary to print this email?

[Husam El Alqamy]

Dear Alaa

whatever the modeling algorithm  you are using, Use ENMTools to create the random sets of points equal to your presence points, you will have CSV files equal to your duplicate random points, feed this into your modeling algorithm or software and then extract the AUC value of each run. build a list of these in R and use the histo() function to build your histogram. Your model is significant against chance if you have your AUC value near the left hand tail of your Random duplicate AUCs histogram. In other words the AUC of the test model id greater that the 75 percentile on the histogram produced by the 999 AUC values of the null model. This is applied in our Ibex habitat model that we presented in Spain in 2009. I hope this helps.

Regards  

--

You received this message because you are subscribed to the Google Groups "Maxent" group.
To unsubscribe from this group and stop receiving emails from it, send an email to maxent+un...@googlegroups.com.
To post to this group, send email to max...@googlegroups.com.
Visit this group at http://groups.google.com/group/maxent.
For more options, visit https://groups.google.com/d/optout.

--

Husam El Alqamy, B.Sc., M.Phil.
Sr. Biodiversity GIS Analyst ,
Environmental Information Sector, EIS
Environmental Agency Abu Dhabi,UAE
Antelope Specialist Group, ASG - IUCN

[Alaa Eldeen]

Thanks a million Husam

You made it so easy for me, I will do and keep you updated

Happy Eid

[Husam El Alqamy]

sorry Alaa.

There was a mistake in my previous message. previously I said "Your model is significant against chance if you have your AUC value near the left hand tail of your Random duplicate AUCs histogram" actually I was wrong. The AUC value should be on the RIGHT right tail of your histogram.

Regards

[matea.vukelic31]

Hi,i know it is a simple question but i just started to use ENMTools, when i try to build a null model using resample from raster,what should i put in the number of points per replicate? I have one species(different wolfs) and 25 000 data (gps coordinates)?

ReplyDelete

[Areej Jaradat]

 Dear AlaaEldin and Husam Alqamy

If we used a model were we set a random testing percentage, what would be the comparable setting used when running the null model?. I've seen papers generated the same number of occurrence points, but they were not clear if they set all random occurrence points as training only? doesn't  this contradict with the tested maxent model where they deducted part of the sample as a testing not training dataset? ? 

Or in such cases we normally set the null model to the same testing percentage as the tested model? Or we generate random points as the same number of the 'training' occurrence points that used in the study?

Also what is the recommended number of random datasets to produce using the ENMtools? and how many times it should be repeated (replicates) in maxent? 

. I searched the group, and the answers were not clear either on this point. I would appreciate it if someone can resolve this issue

Best

[Jamie M. Kass]

I think this paper might clear a lot of this up -- it proposes a null model like Raes & ter Steege's but uses the same validation dataset used to evaluate the empirical model:

Bohl, C. L., Kass, J. M., & Anderson, R. P. (2019). A new null model approach to quantify performance and significance for ecological niche models of species distributions. Journal of Biogeography, 46(6), 1101-1111.

The new ENMeval version implements this null model pretty easily. Check out the new vignette:

https://github.com/jamiemkass/ENMeval

Jamie M. Kass

JSPS Postdoctoral Scholar

Okinawa Institute of Science and Technology Graduate Univ.

[Husam El Alqamy]

The null hypothesis is that your model product is not different from any random model due to chance, The alternative hypothesis is that your model is significantly different from random chance and represent the niche envelope of your species as a result of relations between presence records and environmental variables.

Regards

B C T S  GIS	Hossameldin ELALKAMY, MPhill., PhD., RPBio. GIS Analyst  BC Timber Sales|  Prince George Ministry of Forests, Lands and Natural Resource Operations P. 778 690 8028 C. 778.896.3229|2000 Ospika Blvd. Prince George, BC,V8W 9M1
	| Profile

To view this discussion on the web visit https://groups.google.com/d/msgid/maxent/e6976c1d-66e8-41dd-9eb7-c3570a697fa2n%40googlegroups.com.