search

Model-R / modleR

An ecological niche model workflow based on dismo
https://model-r.github.io/modleR
51 stars 24 forks source link

do_many,do_any, and final_model #79

Open eliascherenet opened 4 years ago

eliascherenet commented 4 years ago

I was running the following code: final_model(species_name = species[1], algorithms = NULL, #if null it will take all the algorithms in disk models_dir = test_folder, which_models = c("raw_mean", "bin_mean", "bin_consensus"), consensus_level = 0.5, uncertainty = TRUE, overwrite = TRUE)

ens <- ensemble_model

However, I found the following error _Error in final_model(species_name = species[1], algorithms = NULL, models_dir = test_folder, : could not find function "final_model"_ Error in ensemble_model(species_name = species[1], occurrences = occs, : could not find function "ensemble_model"

I need your help on how to fix this issue.

AndreaSanchezTapia commented 4 years ago

Hello @eliascherenet can you provide more details about your code? It seems as though modleR wasn't installed on your computer, or not loaded correctly.

eliascherenet commented 4 years ago

dear Andrea, Please find attached my code and i need your help where to set the code in order to calculate the uncertainty and sensitivity analysis.

Elias

On Wed, Oct 7, 2020 at 8:03 AM AndreaSanchezTapia notifications@github.com wrote:

Hello @eliascherenet https://github.com/eliascherenet can you provide more details about your code? It seems as though modleR wasn't installed on your computer, or not loaded correctly.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/Model-R/modleR/issues/79#issuecomment-704693462, or unsubscribe https://github.com/notifications/unsubscribe-auth/AD4I2LCVI7YVF6JPIJ2QOD3SJPZBHANCNFSM4SENBPZQ .

--

+++++++++++++++++++++++++++++++++++++++++++++ Elias Cherenet Weldemariam(MSc, BSc) Lecturer(GIS and Remote Sensing) Coordinator, Department of Geo-information Science (GIS) Haramaya University, Ethiopia Email: eliachent@gmail.com, elubdu1999@yahoo.com Skype: elias.weld, +251(0)966429087 ++++++++++++++++++++++++++++++++++++++++++++++

eliascherenet commented 4 years ago

dear Andrea,,

I hope you have got my attachment file in my previous email. Looking your help soon.

regards

On Wed, Oct 7, 2020 at 8:28 AM Elias Cherenet eliachent@gmail.com wrote:

dear Andrea, Please find attached my code and i need your help where to set the code in order to calculate the uncertainty and sensitivity analysis.

Elias

On Wed, Oct 7, 2020 at 8:03 AM AndreaSanchezTapia < notifications@github.com> wrote:

Hello @eliascherenet https://github.com/eliascherenet can you provide more details about your code? It seems as though modleR wasn't installed on your computer, or not loaded correctly.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/Model-R/modleR/issues/79#issuecomment-704693462, or unsubscribe https://github.com/notifications/unsubscribe-auth/AD4I2LCVI7YVF6JPIJ2QOD3SJPZBHANCNFSM4SENBPZQ .

--

+++++++++++++++++++++++++++++++++++++++++++++ Elias Cherenet Weldemariam(MSc, BSc) Lecturer(GIS and Remote Sensing) Coordinator, Department of Geo-information Science (GIS) Haramaya University, Ethiopia Email: eliachent@gmail.com, elubdu1999@yahoo.com Skype: elias.weld, +251(0)966429087 ++++++++++++++++++++++++++++++++++++++++++++++

--

+++++++++++++++++++++++++++++++++++++++++++++ Elias Cherenet Weldemariam(MSc, BSc) Lecturer(GIS and Remote Sensing) Coordinator, Department of Geo-information Science (GIS) Haramaya University, Ethiopia Email: eliachent@gmail.com, elubdu1999@yahoo.com Skype: elias.weld, +251(0)966429087 ++++++++++++++++++++++++++++++++++++++++++++++

AndreaSanchezTapia commented 4 years ago

Hello Elias, GitHub does not allow for attachments. Can you post a reproducible example of your code? Are you sure that modleR is correctly installed on your computer?

eliascherenet commented 4 years ago

dear Andrea, I found this " modelr" from the installation but didn't get the one mentioned in the GitHub" modleR " . I do not know if this is the same and have similar function.

please find also my code here:

library(maptools) ## For wrld_simpl library(raster) library(dismo) library(rgdal) library(raster) library(ggplot2) library(parallel) library(tidyr) library(dplyr)

file <- paste(system.file(package="dismo"), "/ex/sena.csv", sep="") file sp<- read.csv("C:/Users/user/Desktop/ex/sena.csv") head(sp) colnames(sp) class(sp)## this is data frame not spatial point, to change from datafram to spatial... setwd("C:/Users/user/Desktop/ex/wc2.1_10m_elev")

sp$species<- 1 # to add new columun

sp<-sp[,c('lon','lat','species')]## to combine the new column and other

column from original data

head(sp)

class(sp)

coordinates(sp)<- ~lon + lat

class(sp)

head(sp)

sp library(maptools) data(wrld_simpl) plot(wrld_simpl, xlim=c(27,27), ylim=c(-35,35), axes=TRUE,col="light yellow")

restore the box around the map

box()

plot points

points(sp$long, sp$lat, col='orange', pch=20, cex=0.75)

plot points again to add a border, for better visibility

points(sp$long, sp$lat, col='red', cex=0.75) coordinates(sp) <- ~long+lat crs(sp) <- crs(wrld_simpl) class(sp)

class(wrld_simpl)

ovr <- over(sp, wrld_simpl)

head(ovr)

tail(ovr)

cntr <- ovr$NAME

i <- which(is.na(cntr))

i

j <- which(cntr != sp$country)

cbind(cntr, sp$country)[j,]

plot(sp)

plot(wrld_simpl, add=T, border='blue', lwd=1)

points(sp[j, ], col='yellow', pch=20, cex=2)

sp<-sp[c('species')]

class(sp) head(sp) bio1 <- raster::getData('worldclim', var='bio', res=5) bio <- stack(bio1) bio

plot(bio)

elev <- raster("wc2.1_5m_elev.tif")

elev

plot(elev)

bio <- stack(bio1,elev)

bio <- raster::stack(bio1)

plot(bio)

bio e <- extent(bio)

p <- as(e, 'SpatialPolygons')

projection(p) <- crs(bio)

mask_raster <- function(raster, shape){ outr <- crop(raster, extent(shape)) outr <- mask(outr, p) return(outr) } bio<- mask_raster(raster= bio1, shape = p)

elev <- mask_raster(raster= elev, shape = p)

bio2 <- stack(bio1,elev) bio2

plot(bio2)

names(bio2)

#######################

r <- bio2[[c(1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20)]] names(r)

names(r) <- c("Temp","Prec")

############ myshp <- readOGR("C:/Users/user/Desktop/ex/Africa.shp")

plot(myshp)

bion2 <- raster(r, crs=proj4string(myshp))

r2 <- crop(r, extent(myshp)) bion3 <- mask(r2, myshp)

plot(bion3)

plot(myshp, add=TRUE, lwd=2)

bion3 <- stack(bion3) bion3 #################################

myshp <- readOGR("C:/Users/user/Desktop/ex/Africa.shp") plot(myshp) bion2 <- raster(r, crs=proj4string(myshp))

r2 <- crop(r, extent(myshp)) bion3 <- mask(r2, myshp)

plot(bion3)

plot(myshp, add=TRUE, lwd=2) bion3 <- stack(bion3) bion3 plot(bion3) plot(bion3[[1]])

############

install.packages("rgeos") e<-drawExtent() sp<-crop(sp,e) points(sp,col='red') biok<-crop(bion3,e) plot(biok) plot(biok[[1]])

install.packages("usdm")

library(usdm) ex<- raster::extract(biok,sp) head(ex)

v1<- vifstep(ex)

v2<- vifcor(bion3, th=0.7)

v1

v2

bm2<-exclude(biok,v1) head(bm2) plot(bm2[[1]])

vif(bion3)

#####################

e<-drawExtent()

sp<-crop(sp,e)

points(sp,col='blues9')

bio

install.packages("usdm")

library(usdm)

ex<- raster::extract(bio2,sp)

head(ex)

v1<- vifstep(bion3, th=10) v2<- vifcor(bion3, th=0.7) v1 v2 bm2<-exclude(bion3,v1)

plot(bm2[[1]])

to select important variables after VIF

bm2<-

bio2[[c('bio2','bio3','bio4','bio8','bio12','bio13','bio14','bio16','bio18','wc2.1_10m_elev')]]

names(bm2) <-

c('bio2','bio3','bio4','bio8','bio12','bio13','bio14','bio16','bio18','wc2.1_10m_elev')

bm2

plot(bm2)

bm2<-

bio2[[c('bio1','bio2','bio4','bio12','bio13','bio14','bio17','bio18','wc2.1_10m_elev')]]

names(bm2) <-

c('bio1','bio2','bio4','bio12','bio13','bio14','bio17','bio18','wc2.1_10m_elev')

bm2

plot(bm2)

##############333

bm2<-

bio2[[c('bio1','bio3','bio8','bio9','bio10','bio12','bio14','bio17','wc2.1_10m_elev')]]

names(bm2) <-

c('bio1','bio3','bio8','bio9','bio10','bio12','bio14','bio17','wc2.1_10m_elev')

bm2

plot(bm2)

################3333

bm2<-

bio2[[c('bio1','bio3','bio7','bio9','bio10','bio12','bio14','bio15','bio17','wc2.1_10m_elev')]]

names(bm2) <-

c('bio1','bio3','bio7','bio9','bio10','bio12','bio14','bio15','bio17','wc2.1_10m_elev')

bm2

plot(bm2)

###################### bm2<- bion3[[c('bio1','bio6','bio8','bio9','bio12','bio14','bio17','bio18','wc2.1_5m_elev')]] names(bm2) <- c('bio1','bio6','bio8','bio9','bio12','bio14','bio17','bio18','wc2.1_5m_elev') bm2

plot(bm2)

vif(bm2)

cliping raster

myshp <- readOGR("C:/Users/user/Desktop/ex/Africa.shp") plot(myshp) biom2 <- raster(bm2, crs=proj4string(myshp))

r2 <- crop(bm2, extent(myshp)) biom3 <- mask(r2, myshp)

plot(biom3)

plot(myshp, add=TRUE, lwd=2)

biom3 <- stack(biom3)

crop

points(sp,cex=0.5,pch=16)

proj4string(sp)<-projection(raster()) mapview(sp)

install.packages('mapview') library(mapview)

getmethodNames()

mapview(wrld_simpl) installAll()

library(sdm)

d <- sdmData(species~.,sp,predictors = bm2, bg=list(n=800), method='gRandom',remove=TRUE,seed=47) d m <- sdm(species~.,d, methods = c('glm','svm','rf','brt','gam','fda','mars','mda'), replications= c('boot'),n=10, parallelSetting=list(ncore=4, method='parallel'),seed=47)

m

p <-predict(m,bion3,'prediction2020.img', overwrite=TRUE)

p plot(p)

rcurve(m) plot(sdm::getVarImp(m,1:80)) getVarImp(m)

install.packages("shiny") library(shiny) installAll() gui(m)

library(rgdal) # for vector work; sp package should always load with rgdal. library (raster)

en<- ensemble(m,p,'ens2020.img',setting = list(method='weighted',stat='TSS',opt=2),overwrite=TRUE,uncertainty = TRUE,sensitivity = 0.9,consensus_level = 0.5)

en plot(en)

biof2<-raster::getData('CMIP5',var='bio',res=5,rcp=45, year=70,model='AC')

biof <- stack(biof2,elev) names(bio2) names(biof)

names(biof)<-names(bio2) names(biof)<-names(bio2) names(biof) ##########################

#####################

elev

plot(elev)

biof <- stack(biof2,elev)

biof

names(biof)

biok <- raster::stack(biof2)

plot(biok)

e <- extent(biok)

p <- as(e, 'SpatialPolygons')

projection(p) <- crs(biok)

mask_raster <- function(raster, shape){

outr <- crop(raster, extent(shape))

outr <- mask(outr, p)

return(outr)

}

biok<- mask_raster(raster= biof2, shape = p)

elev <- mask_raster(raster= elev, shape = p)

biol <- stack(biox,elev)

biol

plot(biol)

##############3

cliping to afrcia extent

myshp <- readOGR("C:/Users/user/Desktop/ex/Africa.shp")

plot(myshp)

biof3 <- raster(biof, crs=proj4string(myshp)) r2 <- crop(biof, extent(myshp)) biof3 <- mask(r2, myshp)

plot(biof3)

names(biof3)

plot(myshp, add=TRUE, lwd=2)

end of clipping

pf<-predict(m,biof3,'predictf4570.img',overwrite=TRUE)

plot(pf)

plot(stack(p,pf))

pf5<-calc(pf,mean)

plot(pf5, main='predicted future')

enf5<-ensemble(m,pf,'RCP 4.5 2070.img',overwrite=TRUE, setting = list(methods='weighted',stat='TSS',opt=2))

plot(enf5, main='average ensf')

enf55<-calc(enf5,mean)

plot(enf55, main= 'mean of ensf')

plot(stack(en,enf5)) plot(en) plot(enf5) plot(stack(en,enf5)) ch<-enf5-en plot(ch, main= 'Change') getEvaluation(m,stat='AUC') ##or ev<-getEvaluation(m,stat=c('AUC','TSS','threshold'), opt = 2) mean(ev$threshold) pa<-raster(en) pa[]<-ifelse(en[]>='0.4821103',1,0) plot(pa) pf<-raster(enf5) pf[]<-ifelse(enf5[]>='0.4821103',1,0) plot(pf,main='prediction70 after mean') pac<-pf-pa plot(pac, main='Pf-PA') cl<-colorRampPalette(c('red','gray','darkgreen')) plot(pac,col=cl(3),main='PF-PA')

library(sp) library(raster)

z<-raster::writeRaster(pac, "changemap.tif")

z <- raster(z) writeRaster(pac, filename="changemap.tif", overwrite=TRUE)

rcurve(m,id=2) plot(sdm::getVarImp(m,9)) getVarImp(m)

Elias

On Fri, Oct 9, 2020 at 10:01 PM AndreaSanchezTapia notifications@github.com wrote:

Hello Elias, GitHub does not allow for attachments. Can you post a reproducible example of your code? Are you sure that modleR is correctly installed on your computer?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/Model-R/modleR/issues/79#issuecomment-706352406, or unsubscribe https://github.com/notifications/unsubscribe-auth/AD4I2LCCJYAHS4Q75DTAW7DSJ5MY7ANCNFSM4SENBPZQ .

--

+++++++++++++++++++++++++++++++++++++++++++++ Elias Cherenet Weldemariam(MSc, BSc) Lecturer(GIS and Remote Sensing) Coordinator, Department of Geo-information Science (GIS) Haramaya University, Ethiopia Email: eliachent@gmail.com, elubdu1999@yahoo.com Skype: elias.weld, +251(0)966429087 ++++++++++++++++++++++++++++++++++++++++++++++

eliascherenet commented 4 years ago

dear Andrea, devtools::load_all("../../1_modleR") failed to run in my machine. and also the loading "modleR " package refused. I need your help in fixing these issues.

AndreaSanchezTapia commented 4 years ago

Hello Elias, when reading your script I can see that you are not using modleR but other packages, such as sdm and usdm. As far as modleR goes, if library(modleR) tells you there is no package, you must install it, via remotes::install_github() Check the README for installation: 

remotes::install_github("Model-R/modleR",
                        build = TRUE,
                        dependencies = TRUE,
                        build_opts = c("--no-resave-data", "--no-manual"),
                        build_vignettes = TRUE)

devtools::load_all() is a command used for development and it is not meant to be run, you can skip that.

eliascherenet commented 4 years ago

final_model(species_name = species[1], algorithms = NULL, #if null it will take all the algorithms in disk models_dir = test_folder, which_models = c("raw_mean", "bin_mean", "bin_consensus"), consensus_level = 0.5, uncertainty = TRUE, overwrite = TRUE) I tried to run the above code for testing purpose and i found the following error,

Abarema_langsdorffii Reading evaluation files for Abarema_langsdorffii in present Extracting data for Abarema_langsdorffii bioclim Reading models from .tif files Error in x[[1]] : subscript out of bounds
Your help on how to fix this is appreciated!!!!!!

eliascherenet commented 4 years ago

dear Andrea, Good afternoon, when i tried to run the final_model script I got this error. I would like to ask your quick help , on how to fix this problem. I tried to shorten the path but failed to execute.

Elias

[image: image.png]

On Mon, Oct 12, 2020 at 6:54 PM AndreaSanchezTapia notifications@github.com wrote:

Hello Elias, when reading your script I can see that you are not using modleR but other packages, such as sdm and usdm. As far as modleR goes, if library(modleR) tells you there is no package, you must install it, via remotes::install_github() Check the README for installation:

remotes::install_github("Model-R/modleR", build = TRUE, dependencies = TRUE, build_opts = c("--no-resave-data", "--no-manual"), build_vignettes = TRUE)

devtools::load_all() is a command used for development and it is not meant to be run, you can skip that.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/Model-R/modleR/issues/79#issuecomment-707203287, or unsubscribe https://github.com/notifications/unsubscribe-auth/AD4I2LFLX3SGUNPXHP6TBBDSKMRD5ANCNFSM4SENBPZQ .

--

+++++++++++++++++++++++++++++++++++++++++++++ Elias Cherenet Weldemariam(MSc, BSc) Lecturer(GIS and Remote Sensing) Coordinator, Department of Geo-information Science (GIS) Haramaya University, Ethiopia Email: eliachent@gmail.com, elubdu1999@yahoo.com Skype: elias.weld, +251(0)966429087 ++++++++++++++++++++++++++++++++++++++++++++++

AndreaSanchezTapia commented 4 years ago

Hi Elias, your attachment is not coming through. Could you send a .zip file with your example data to andreasancheztapia@gmail.com, please? Occurrences for one species, scripts and one or two explanatory variables should be enough. I would like to examine a reproducible example of your code.