There are internet search sites that are specialized for R searches, including search.r-project.org (which is the site used by RSiteSearch) and Rseek.org. If you Google 'R', go to help > posting guide. I hope this helps. I had similar problems and I know it can be frustrating getting the specific answer, but keep going!

Andrew :-)

I haven't done myself, but I guees you first need to import csv files ( "species_omission.csv") to R and then use the package pROC. The function AUC has an partial.auc argument that allows computing pAUC.

Check it here

https://cran.r-project.org/web/packages/pROC/pROC.pdf

Dear Andrew and Facundo,

Thank you both so much for your answers :)

While I was waiting for the answers here, I tried Niche Analyst software and it works great! I counted all required statistics without any problem. And it is more intuitive than R, so I think that others also can easily use it.

http://nichea.sourceforge.net/

Recall that AUC and by extension pAUC have been hardly criticized as a measure of model performance, and their use are discouraged. Some other measures such as information criteria are preferred.

http://onlinelibrary.wiley.com/doi/10.1111/j.1466-8238.2007.00358.x/abstract

https://www.researchgate.net/publication/256473817_Using_Maxent_to_model_the_historic_distributions_of_stonefly_species_in_Illinois_streams_The_effects_of_regularization_and_threshold_selections

https://www.ncbi.nlm.nih.gov/pubmed/21563566

Article Using Maxent to model the historic distributions of stonefly...

The problem is that in one of my articles reviewer had insisted on making this analysis... Therefore, what else should I do? :-/

I also calculate AIC, AICc and Kappa. I personally think that checking something with several methods is not correct. But sometimes it does not matter what we think, but what 'experts' say.

Nevertheless, thank you for showed me this article, because I can rely on it in the discussion :)

Oh! between a rock and a hard place. I've been also accepted an article, and the reviewer told me to compute AICc. The problem with computing several metrics is that different models may be (and oftenly are) selected, so I think it does not make much sense. You should select a priori one measure and justify its use. My advice is agree the reviewer, and justify why you use that measure. 

Indeed, this is what I am going to do :-)

We developped a webtool to easily analyze all sort of screening results, using all usual metrics in the field. Partial metrics (including pAUC) can be calculated on mouse-over, which may be convenient in your case. See http://stats.drugdesign.fr.

Webtool associated publication : Screening Explorer – An Interactive Tool for the Analysis of Screening Results (http://pubs.acs.org/doi/abs/10.1021/acs.jcim.6b00283)

Thank you for this information, and for sharing it here. It may be helpful for others looking for similar solutions

 Also, to calculate regular AUC, omission rates, and AICc, please check out the R package ENMeval. It iterates model building of Maxent models over many different parameter combinations (feature classes and regularization multipliers), and outputs a summary table showing you the result stats.

Use partialROC () available via ENMGadgets. Please send me a message and i can send all materials for that. Good Luck for everyone! i hope you managed to respond the reviewer Agnieszka and your paper is now published. 

You may try this online tool: http://shiny.conabio.gob.mx:3838/nichetoolb2/