Hi Timothy,
I was trying to get a probability of the species based on MaxEnt function. As
per the theory of MaxEnt,I uses the sites where only species is present.
Therefore I selected only those rows where the species is present,
my data is : 'dd', in which- there are X,Y coordinates and 4 different kind of
species (sp1, sp2, sp3, sp4) and four environmental variables (env1-env4).
and trying to develop model using the sites in which the species present,
For example, sp1: to select the sites in which the species is present were
selected
spA<-subset(dd,dd$sp1==1)# select the rows in which species 1 is present
After that, I was trying to develop a model for the species, using following
m1 <- maxent(spA[,7:10], spA[,'sp1']) # develop a model for species 1 based on
the data where the species is present
After that, I wanted to use this model to predict using new data (new data set-
that is 'predi'). but, when I run but it R crashed.
m1.prediction<-predict(m1, predi) # apply the model for new data
Dennis also tried to help me but his R was also crashed. Any thing- did I
mistake from which our Rs were crashed? Any idea?
thanks
KG
--------------------------------------------------------------------------------------------------------
install.packages("maxent", dependencies=TRUE)
library(maxent)
dd<-structure(list(long = c(-75.747, -106.84, -105.27, -104.64, -103.71,
-102.72, -101.79, -100.8, -105.69, -104.67, -103.71, -102.69,
-101.66, -100.71, -99.685, -98.656, -97.627, -111.82, -110.2,
-109.55, -106.88, -106.84, -105.85, -104.82, -103.84, -102.8,
-101.83, -100.79, -99.742, -98.695, -97.744, -112.46, -111.8,
-110.79, -108.41, -107.74), lat = c(19.792, 21.576, 21.347, 21.243,
21.078, 21.185, 20.99, 21.067, 22.621, 22.763, 22.594, 22.704,
22.797, 22.581, 22.642, 22.685, 22.711, 24.153, 23.44, 23.385,
23.4, 23.702, 23.565, 23.71, 23.542, 23.654, 23.454, 23.533,
23.596, 23.64, 23.377, 24.796, 24.459, 24.399, 24.503, 24.429
), sp1 = c(0L, 1L, 0L, 1L, 1L, 1L, 0L, 0L, 1L, 0L, 1L, 0L, 0L,
0L, 0L, 0L, 0L, 0L, 0L, 1L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 1L, 0L,
0L, 0L, 1L, 0L, 1L, 1L, 0L), sp2 = c(0L, 0L, 0L, 0L, 0L, 0L,
0L, 1L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L,
0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 1L, 0L, 0L, 0L), sp3 = c(0L,
0L, 0L, 0L, 0L, 0L, 0L, 1L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L,
0L, 0L, 0L, 0L, 0L, 0L, 1L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L,
0L, 0L, 0L), sp4 = c(0L, 0L, 0L, 0L, 0L, 0L, 1L, 0L, 0L, 0L,
0L, 0L, 1L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L,
0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L), env1 = c(182.0037689,
163.3301239, 443.0214233, 1240.155273, 1774.867432, 1909.528809,
2153.244141, 1878.018433, 909.1315308, 1980.898438, 2271.118896,
2122.672852, 2033.41626, 1534.658447, 828.4759522, 222.3117523,
23.4761219, 79.52642822, 322.994751, 273.6637268, 35.5085907,
516.3795776, 2205.419434, 2118.727539, 2178.901123, 1995.210083,
2048.075928, 1824.84021, 1043.509888, 184.4526062, 12.6928978,
20.87172508, 126.5344544, 258.8852844, 3.28964257, 140.3287811
), env2 = c(1134L, 550L, 2111L, 2523L, 2156L, 1209L, 1107L, 2605L,
3176L, 2490L, 1360L, 801L, 1118L, 1484L, 2730L, 1309L, 104L,
197L, 2033L, 1339L, 567L, 2694L, 2708L, 1806L, 1344L, 912L, 1377L,
2346L, 3265L, 989L, 69L, 428L, 764L, 896L, 100L, 1521L), env3 = c(24.533,
24.928, 24.707, 21.052, 21.318, 18.428, 19.041, 17.743, 24.371,
19.689, 16.879, 16.528, 16.901, 18.015, 20.648, 25.31, 24.308,
22.528, 22.912, 22.001, 25.097, 25.391, 19.154, 14.943, 17.143,
16.898, 16.891, 17.563, 15.63, 24.354, 24.088, 22.527, 22.528,
22.126, 25.317, 25.643), env4 = c(0.047969, 0.003469, 0.003385,
0.002253, 0.000791, 0.001834, 0.008016, 0.009262, 0.003934, 0.002322,
0.00061, 0.000799, 6.4e-05, 0, 0, 0.000107, 0.003151, 0.018915,
0.015077, 0.004554, 0.003499, 0.002705, 0.003507, 0.001173, 0.000149,
0.000308, 0, 0, 0, 0.00074, 0.002845, 0.017047, 0.018915, 0.017111,
0.002417, 0.002668)), .Names = c("long", "lat", "sp1", "sp2",
"sp3", "sp4", "env1", "env2", "env3", "env4"), class = "data.frame", row.names
= c(NA,
-36L))
predi<-dd[,7:10]
spA<-subset(dd,dd$sp1==1)
m1 <- maxent(spA[,7:10], spA[,'sp1']) # for species 1
m1.prediction<-predict(m1, predi)# predict species 1 using new data set
spB<-subset(dd,dd$sp2==1)
m2 <- maxent(spB[,7:10], spB[,'sp2']) # for species 2
m2.prediction<-predict(m2, predi)
spC<-subset(dd,dd$sp3==1)
m3 <- maxent(spC,7:10], spC[,'sp3']) # for species 3
m3.prediction<-predict(m3, predi)
spD<-subset(dd,dd$sp4==1)
m4 <- maxent(spD[,7:10], spD[,'sp4']) # for species 4
m4.prediction<-predict(m4, predi)
prediction<-cbind(m1.prediction,m2.prediction,m3.prediction,m4.prediction)
-----------------------------------------------------------------------------------------------------------
> Date: Thu, 16 May 2013 17:53:56 -0700
> Subject: Re: [R] Why is it not working the MaxEnt analysis (package "maxent",
> and "dismo")?
> From: djmu...@gmail.com
> To: kristi.glo...@hotmail.com
>
> Hi:
>
> R crashed when I tried to run your first set of predictions, and it
> doesn't really surprise me.
>
> * predi consists of the variables env1-env4
>
> AFAICT, this code:
> spA<-subset(dd,dd$sp1==1)
> m1 <- maxent(spA[,7:10], spA[,'sp1']) # for species 1
> m1.prediction<-predict(m1, predi)
>
> attempts to select out the rows of dd for which sp1 == 1. Then you run
> the maxent function on the rows of predi for which sp1 == 1, where the
> second argument is 1. If you read the help page for maxent::maxent,
> you'll discover that the first argument is OK, but the second argument
> expects
> "A factor or vector of labels corresponding to each document in the
> feature_matrix"
> Your second argument is a numeric constant.
>
> What I didn't understand was why you would want to use predi as the
> set of values at which to get model predictions when a subset of that
> matrix was used as the response in the maxent() call that returns the
> fitted model.
>
> I don't know enough about what this package does to provide a solid
> rationale for what is right or wrong with your code, but I do know
> that whatever you did is 'bad' in the sense that it was sufficient to
> crash R. This often means some underlying C code threw a segmentation
> fault as a result of your inputs. If I had to guess, I would posit
> that the second argument is meant to be a (variable) response and you
> fed it a constant vector of 1's. What havoc that creates in the
> package code is beyond my understanding.
>
> I don't know if you got a crash on your end, but it's not something
> that should happen so you should contact the maintainer of the maxent
> package and report the problem, including the data and code that
> induced the crash.
>
> maintainer("maxent")
> [1] "Timothy P. Jurka <tpju...@ucdavis.edu>"
>
> He'll undoubtedly have a better idea of the source of the problem than
> me. I got it from the first block of code above using
>
> > sessionInfo()
> R version 3.0.0 (2013-04-03)
> Platform: x86_64-w64-mingw32/x64 (64-bit)
>
> locale:
> [1] LC_COLLATE=English_United States.1252
> [2] LC_CTYPE=English_United States.1252
> [3] LC_MONETARY=English_United States.1252
> [4] LC_NUMERIC=C
> [5] LC_TIME=English_United States.1252
>
> attached base packages:
> [1] stats graphics grDevices utils datasets methods base
>
> other attached packages:
> [1] maxent_1.3.3 tm_0.5-8.3 SparseM_0.99 Rcpp_0.10.3 dismo_0.8-11
> [6] raster_2.1-25 sp_1.0-9
>
> loaded via a namespace (and not attached):
> [1] grid_3.0.0 lattice_0.20-15 slam_0.1-28 tools_3.0.0
>
> in a fresh R session.
>
>
>
> Aside: I attempted to write a function that would cut down the amount
> of typing that you would need to do, something like
>
> predFun <- function(DF, var, ...)
> {
> require(maxent)
> s1 <- DF[DF[[var]] == 1, ]
> m <- maxent(s1[ , 7:10], s1[, var])
> predict(m, predi)
> }
>
> ## Note: subset() uses nonstandard evaluation, so it is not recommended
> ## for use inside of function bodies.
>
> ## Note 2: Whatever code fixes are recommended, make them in the function.
>
> vars <- paste0("sp", 1:4)
>
> # Run predFun recursively on each element of vars,
> # which becomes the x in the anonymous function
> # comprising the second argument of lapply()
>
> L <- lapply(vars, function(x) predFun(DF, x))
> do.call(cbind, L) # cbind the results
>
> Obviously untested, but it gives you an idea of how the game works.
>
> Dennis
>
> On Thu, May 16, 2013 at 2:48 PM, Kristi Glover
> <kristi.glo...@hotmail.com> wrote:
> > Hi R-user,
> > I was trying to compute probability of species distribution using MaxEnt.
> > the MaxEnt uses only teh sites in which species present,
> > I tried to run but did not work, would you give me some hints?
> >
> > I have given an example table and how I tried it.
> > ---------------------------------------------------
> > install.packages("maxent", dependencies=TRUE)
> > install.packages("dismo", dependencies=TRUE)
> > library(dismo)
> > library(maxent)
> >
> > dd<-structure(list(long = c(-75.747, -106.84, -105.27, -104.64, -103.71,
> > -102.72, -101.79, -100.8, -105.69, -104.67, -103.71, -102.69,
> > -101.66, -100.71, -99.685, -98.656, -97.627, -111.82, -110.2,
> > -109.55, -106.88, -106.84, -105.85, -104.82, -103.84, -102.8,
> > -101.83, -100.79, -99.742, -98.695, -97.744, -112.46, -111.8,
> > -110.79, -108.41, -107.74), lat = c(19.792, 21.576, 21.347, 21.243,
> > 21.078, 21.185, 20.99, 21.067, 22.621, 22.763, 22.594, 22.704,
> > 22.797, 22.581, 22.642, 22.685, 22.711, 24.153, 23.44, 23.385,
> > 23.4, 23.702, 23.565, 23.71, 23.542, 23.654, 23.454, 23.533,
> > 23.596, 23.64, 23.377, 24.796, 24.459, 24.399, 24.503, 24.429
> > ), sp1 = c(0L, 1L, 0L, 1L, 1L, 1L, 0L, 0L, 1L, 0L, 1L, 0L, 0L,
> > 0L, 0L, 0L, 0L, 0L, 0L, 1L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 1L, 0L,
> > 0L, 0L, 1L, 0L, 1L, 1L, 0L), sp2 = c(0L, 0L, 0L, 0L, 0L, 0L,
> > 0L, 1L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L,
> > 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 1L, 0L, 0L, 0L), sp3 = c(0L,
> > 0L, 0L, 0L, 0L, 0L, 0L, 1L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L,
> > 0L, 0L, 0L, 0L, 0L, 0L, 1L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L,
> > 0L, 0L, 0L), sp4 = c(0L, 0L, 0L, 0L, 0L, 0L, 1L, 0L, 0L, 0L,
> > 0L, 0L, 1L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L,
> > 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L, 0L), env1 = c(182.0037689,
> > 163.3301239, 443.0214233, 1240.155273, 1774.867432, 1909.528809,
> > 2153.244141, 1878.018433, 909.1315308, 1980.898438, 2271.118896,
> > 2122.672852, 2033.41626, 1534.658447, 828.4759522, 222.3117523,
> > 23.4761219, 79.52642822, 322.994751, 273.6637268, 35.5085907,
> > 516.3795776, 2205.419434, 2118.727539, 2178.901123, 1995.210083,
> > 2048.075928, 1824.84021, 1043.509888, 184.4526062, 12.6928978,
> > 20.87172508, 126.5344544, 258.8852844, 3.28964257, 140.3287811
> > ), env2 = c(1134L, 550L, 2111L, 2523L, 2156L, 1209L, 1107L, 2605L,
> > 3176L, 2490L, 1360L, 801L, 1118L, 1484L, 2730L, 1309L, 104L,
> > 197L, 2033L, 1339L, 567L, 2694L, 2708L, 1806L, 1344L, 912L, 1377L,
> > 2346L, 3265L, 989L, 69L, 428L, 764L, 896L, 100L, 1521L), env3 = c(24.533,
> > 24.928, 24.707, 21.052, 21.318, 18.428, 19.041, 17.743, 24.371,
> > 19.689, 16.879, 16.528, 16.901, 18.015, 20.648, 25.31, 24.308,
> > 22.528, 22.912, 22.001, 25.097, 25.391, 19.154, 14.943, 17.143,
> > 16.898, 16.891, 17.563, 15.63, 24.354, 24.088, 22.527, 22.528,
> > 22.126, 25.317, 25.643), env4 = c(0.047969, 0.003469, 0.003385,
> > 0.002253, 0.000791, 0.001834, 0.008016, 0.009262, 0.003934, 0.002322,
> > 0.00061, 0.000799, 6.4e-05, 0, 0, 0.000107, 0.003151, 0.018915,
> > 0.015077, 0.004554, 0.003499, 0.002705, 0.003507, 0.001173, 0.000149,
> > 0.000308, 0, 0, 0, 0.00074, 0.002845, 0.017047, 0.018915, 0.017111,
> > 0.002417, 0.002668)), .Names = c("long", "lat", "sp1", "sp2",
> > "sp3", "sp4", "env1", "env2", "env3", "env4"), class = "data.frame",
> > row.names = c(NA,
> > -36L))
> >
> > predi<-dd[,7:10]
> >
> > spA<-subset(dd,dd$sp1==1)
> > m1 <- maxent(spA[,7:10], spA[,'sp1']) # for species 1
> > m1.prediction<-predict(m1, predi)
> >
> > spB<-subset(dd,dd$sp2==1)
> > m2 <- maxent(spB[,7:10], spB[,'sp2']) # for species 2
> > m2.prediction<-predict(m2, predi)
> >
> > spC<-subset(dd,dd$sp3==1)
> > m3 <- maxent(spC,7:10], spC[,'sp3']) # for species 3
> > m3.prediction<-predict(m3, predi)
> >
> >
> > spD<-subset(dd,dd$sp4==1)
> > m4 <- maxent(spD[,7:10], spD[,'sp4']) # for species 4
> > m4.prediction<-predict(m4, predi)
> >
> > prediction<-cbind(m1.prediction,m2.prediction,m3.prediction,m4.prediction)
> >
> > But, it did not work. I could not figure it out why it is not working.
> > Another thing, How I can make a loop for several species?
> > cheers,
> > KG
> >
> >
> >
> >
> >
> > [[alternative HTML version deleted]]
> >
> > ______________________________________________
> > R-help@r-project.org mailing list
> > https://stat.ethz.ch/mailman/listinfo/r-help
> > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> > and provide commented, minimal, self-contained, reproducible code.
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
