hemingnm / enmwizard

Advanced Techniques for Ecological Niche Modeling Made Easy

Hi, I'm Neander M. Heming

I am a post doc researcher at Applied Ecology and Conservation Lab, UESC - Ilhéus, BA, Brazil.

My research interests include the effects of global warming on the distribution and conservation of organisms, methodological aspects of ecological niche modeling, and how ecological and environmental processes affect bird life histories. Methods include ecological niche models, R stats, R programming, spatial analyses. Currently working on threats that climate change poses to the biodiversity of cabrucas (traditional cocoa agroforests) of southern Bahia and how cabrucas can mitigate impacts of a warming climate.

Interests:

• Ecology
• Conservation
• Climate Change
• Ecological Niche Modeling

---

R packages developed or co-authored by me:

Package	A short summary	Downloads
	Randomization of presence/absence species distribution raster data
	Taxonomic (TD), functional (FD), and phylogenetic (PD) diversity for raster data
	Endemism and evolutionary diversity metrics for raster data

---

Main tools:

         

---

Find me:

Personal website

Lattes resume

Other places

enmwizard's Issues

Error in argument is of length zero

Dear SIR:
my study extent is world,but when I run this code,it is error!could you please tell me how to solve it?

proj.extent <- extent(c(-180, 180, -90, 90))
pred.cut.l <- cut_projarea_rst_mscn_b(proj.extent, predictors.l, occ.polys)

Creating projection area
Error in if (!datatype %in% c("LOG1S", "INT1S", "INT2S", "INT4S", "INT1U", :
argument is of length zero

Question about thrshld_b()

Good morning,
and first of all, thanks for this great package!
We were trying to reproduce the tutorial with our data and came across a small problem with the thrshld_b() function with the following error message:
mods.thrshld.lst <- thrshld_b(mxnt.mdls.preds.cf, thrshld.i = c(5,7))
“Error in FUN(X[[i]], ...) : tentative d’obtenir le slot “results” d’un objet d’une classe élémentaire (“NULL”) sans slots”.
Sorry, it is in French but it looks like the function is looking for a result slot within the object. Is there something related to the function or do we have to precise on which part the mxnt.mdls.preds.cf object we want to apply the threshold?
Thank you for your help

Error message with the ENMevaluate_b() function

Dear Neander,
thank you again for this great tool that ENM Wizard is. I am trying to use your package on a new dataset and got the following error message while trying to run ENMevaluate_b():

Error in ENMeval::ENMevaluate(occ.l[[i]], a.calib.l[[i]], bg.coords = bg.coords.l[[i]], :
unused arguments (aggregation.factor = aggregation.factor, kfolds = kfolds)

The exact command line I used was:

ENMeval.res.lst <- ENMevaluate_b(occ.locs, pred.cut, RMvalues = seq(0.5, 4.5, 0.5), fc = c("L", "P", "Q", "H", "LP", "LQ", "LH", "PQ", "PH", "QH", "LPQ", "LPH", "LQH", "PQH", "LPQH"), method="block", algorithm="maxent.jar")

I came across this closed issue which suggest that it may have something to do with a recent update of ENMeval. Do you think this could be the case?

Cheers,
Joris

Error: Error writing feature weights file

Hi,
sorry to bother you again but now it seems that I cannot save the output of the calibration models:
Error: Error writing feature weights file
Error in file(con, "r") : cannot open the connection
In addition: Warning message:
In file(con, "r") :
cannot open file '3_out.MaxEnt/Mdls.Bvarieg/cloglog/Mod_1.5_L/species.lambdas': No such file or directory
Indeed the folder /Mod_1.5_L/ exists but only contains the 'absent', 'maxent.log', 'maxentResults.csv' (which is empty) and 'presence' files.
Does any body know how to solve that?
All the best,
Joris

The evaluation of ensemble model performance and the calculation of environmental variable importance within the ensemble model

Hello, SIR:
This question is about how to evaluate ensemble model performance and calculate environmental variable importance.
When conducting the following steps, I utilized the parameters below:
mxnt.mdls.preds.lst<-calib_mdl_b(ENMeval.o.l = ENMeval.res.lst,
a.calib.l = clim.stack.list,
occ.l = occ.list,
mSel = "EBPM")
I obtained the following results in model performance and variable importance:
(1) metric.var.PermImportance.csv
sp sel.crit wc2_13 wc2_14 wc2_15 wc2_3 wc2_4 wc2_8 wrb_MostProba
1 R Mod.EBPM_1 3.3966 0 0 3.5917 17.5038 20.1878 18.6949
2 R Mod.EBPM_2 4.4336 1.446 2.0736 3.4377 16.4984 1.3751 45.5691
3 R Mod.EBPM_4 3.1912 9.8249 14.111 2.0451 5.4621 1.5338 36.9477
4 R Mod.EBPM_5 1.7422 2.8706 0 3.5272 11.5683 0.1122 59.874
5 R Mod.EBPM_3 2.4887 23.3218 1.7355 0 2.6066 0.0851 51.2607
(2) sel.mdls.smmr.Rho.csv
Optimality criteria FC RM AICc delta AICc wAIC parameters Rank AIC avg OR10 avg ORLPT avg AUC
2 EBPM_1 LQ 0.5 NA NA NA 17 12 0.25 0.083333333 0.986241667
10 EBPM_2 LQH 1 NA NA NA 29 20 0.416666667 0.083333333 0.983875
11 EBPM_4 LQHP 1 NA NA NA 31 21 0.416666667 0.083333333 0.983333333
16 EBPM_5 LQH 1.5 NA NA NA 21 25 0.333333333 0.083333333 0.982458333
18 EBPM_3 LQHPT 1.5 NA NA NA 23 27 0.333333333 0.083333333 0.983758333
But I want to obtain the overall evaluation results of this ensemble model, not individual evaluation results. What should I do?
Thanks!

Random test percentage setting

Hi
How to set the 'random test percentage'? It seems in the 'calib_mdl_b' funtion, but I cann't do it right....
Thanks in advance!

ENMevaluate_b wrong

Hi @HemingNM
I updated ENMeval package from v0.3 to v 2.0, then when funtion 'ENMevaluate_b ' was applied, it gave error message 'Error in ENMeval::ENMevaluate(occ.l[[i]], a.calib.l[[i]], bg.coords = bg.coords.l[[i]], : Parameter useless(aggregation.factor = aggregation.factor, kfolds = kfolds)'
I noticed the ENMeval package was updated recently(2021-05-19), is it the reason that which is incompatible with ENMwizard now?
Though it's ok to use former version of ENMeval.
BTW, sometimes the funtion 'thrshld_b' failed again, it's weird, because #1 have discussed it a lot and I've used it successfully many times...

Thanks in advance!

Evaluate model performance

Hi @HemingNM:
This question is about how to evaluate model performance.
May I ask how to get the true skill statistics(TSS) of the final model? Because I was asked to assess whether my final best model(selected based on "LowAIC") can be able to discriminate between background and presence points. As I know the ENMwizard only have three criterion to directly assess model performance in the output: "train.AUC", "avg.test.orMTP", and "avg.test.or10pct", but without TSS.
Thanks in advance!

How to fix the problem of no slot of name "occ.pts" for this object of class "ENMevaluation"

Hi Neander Marcel Heming,

I'm using your package ENMwizard. When I use the function calib_mdl_b. there is one problem "no slot of name "occ.pts" for this object of class "ENMevaluation" I checked the ENMevaluate_b function. The occ.pts related content is set as explanatory note. How can I fix this problem?

Error while running select_vars_b()

Hi!
I get the following error while running the function select_vars_b(). It seems that it does not impede to run the rest of the script, but I do not understand this error. In the folder, I have a filed called envData.Oav.grd but not envDataSel.Oav.grd. Any idea of what it could be?
Cheers,
Joris

Error in get(as.character(FUN), mode = "function", envir = envir) :
object '2_envData/area.calib/envDataSel.Oav.grd' of mode 'function' was not found

ENMeval.res.lst error

HELO,SIR: when I run this code:
ENMeval.res.lst <- ENMevaluate_b(occ.locs, pred.cut,
RMvalues = c(1, 1.5), fc = c("L", "LQ", "LP"),
method="block", algorithm="maxent.jar")
Error in ENMeval::ENMevaluate(occ.l[[i]], a.calib.l[[i]], bg.coords = bg.coords.l[[i]], :
unused arguments (aggregation.factor = aggregation.factor, kfolds = kfolds)
how to solve it?thanks!