How I can create species frequency maps using raster and rgdal? - r

Question

I came across this code on GitHub the other day. I am going to create two species frequency maps (not a simple distribution map). One map deals with species frequency in different biomes, and the other one shows the frequency in different eco-regions. I have almost 500 species and want to combine the occurrences of these 500 in one big file and create the maps using that. The following code has been used for another project in which the researchers have developed two maps for the frequency of many plant species in various biomes and eco-regions on Earth. Can anybody tell me how I can turn this code into the one I need?
Which parts of this do I have to remove? Which codes do I need to add?
Frequency map of species in different biomes and eco-regions (Reference: Rice, A., Šmarda, P., Novosolov, M., Drori, M., Glick, L., Sabath, N., ... & Mayrose, I. (2019). The global biogeography of polyploid plants. Nature Ecology & Evolution, 3(2), 265-273.)
require("raster")
require("rgdal")
require("plyr")
# args: (1) data_file (output from previous step, final_species_data.csv). (2) ecoregions (a file of ecoregion names and ids, attributes_and_data/eco_id_with_names.csv) (3) output1 (for biomes) (4) output2 (for ecoregions)
# GLOBAL ARGUMENTS
args <- commandArgs(TRUE)
for(i in 1:length(args)){
eval(parse(text = args[[i]]))
}
get_biomes = function (data){
all_biomes = data$biome
all_biomes = gsub("[[:space:]]", "", all_biomes)
splitter = unlist(strsplit(all_biomes,","))
odd_ind = seq(1,length(unlist(strsplit(splitter,"_"))),2)
even_ind = seq(2,length(unlist(strsplit(splitter,"_"))),2)
biomes = unlist(strsplit(splitter,"_"))[odd_ind]
occurrences = unlist(strsplit(splitter,"_"))[even_ind]
return (list(biomes,occurrences))
}
get_ecoregions = function (data){
all_ecoregions = data$eco_id
all_ecoregions = gsub("[[:space:]]", "", all_ecoregions)
splitter = unlist(strsplit(all_ecoregions,","))
odd_ind = seq(1,length(unlist(strsplit(splitter,"_"))),2)
even_ind = seq(2,length(unlist(strsplit(splitter,"_"))),2)
ecoregions = unlist(strsplit(splitter,"_"))[odd_ind]
occurrences = unlist(strsplit(splitter,"_"))[even_ind]
return(list(ecoregions,occurrences))
}
data = read.csv(data_file,stringsAsFactors=F, na.strings = c(NA,""))
ecoregions_data = read.csv(ecoregions,stringsAsFactors=F)
factors = 23
eco_data = data.frame(matrix(nrow = nrow(ecoregions_data), ncol = factors))
names(eco_data) = c("eco_id","name","occurrences","sp","dp","pp","na","annual","herb_p","woody","unclass_herb","unclass_per","no_lifeform",
"mixed","unresolved","conflict","basal_m","higher_m","angiosperm","basal_d","rosids","asterids","no_tax")
eco_data$eco_id = ecoregions_data$eco_id
eco_data$name = ecoregions_data$eco_name
biome_data = data.frame(matrix(nrow = 14, ncol = factors))
names(biome_data) = c("biome","name","occurrences","sp","dp","pp","na","annual","herb_p","woody","unclass_herb","unclass_per","no_lifeform",
"mixed","unresolved","conflict","basal_m","higher_m","angiosperm","basal_d","rosids","asterids","no_tax")
biome_data$biome = seq(1,14,1)
# initialize data frames as zeros
eco_data[,3:ncol(eco_data)] = 0
biome_data[,3:ncol(biome_data)] = 0
# update count in table
update_count = function(row,original_data_row,curr_occ){
row$occurrences = row$occurrences + curr_occ
row$sp = row$sp + 1
if (is.na(original_data_row$ploidy)){ row$na = row$na + 1 }
else if (original_data_row$ploidy==0){ row$dp = row$dp + 1 }
else if (original_data_row$ploidy==1){ row$pp = row$pp + 1 }
if (is.na(original_data_row$lifeform)) { row$no_lifeform = row$no_lifeform + 1 }
else if (original_data_row$lifeform == "Annual") { row$annual = row$annual + 1 }
else if (original_data_row$lifeform == "Perennial herb") { row$herb_p = row$herb_p + 1 }
else if (original_data_row$lifeform == "Woody") { row$woody = row$woody + 1 }
else if (original_data_row$lifeform == "Unclassified herb") { row$unclass_herb = row$unclass_herb + 1 }
else if (original_data_row$lifeform == "Unclassified perennial") { row$unclass_per = row$unclass_per + 1 }
else if (original_data_row$lifeform == "mixed") { row$mixed = row$mixed + 1 }
else if (original_data_row$lifeform == "unresolved") { row$unresolved = row$unresolved + 1 }
else if (original_data_row$lifeform =="Conflict") { row$conflict = row$conflict + 1 }
if (is.na(original_data_row$Wood_major_group)) { row$no_tax = row$no_tax + 1 }
else if (original_data_row$Wood_major_group =="Basal monocots (non-commelinid monocots)") { row$basal_m = row$basal_m + 1 }
else if (original_data_row$Wood_major_group =="Higher monocots (commelinids)") { row$higher_m = row$higher_m + 1 }
else if (original_data_row$Wood_major_group =="Basal angiosperms") { row$angiosperm = row$angiosperm + 1 }
else if (original_data_row$Wood_major_group =="Basal dicots (non-asterid +non-rosid dicots)") { row$basal_d = row$basal_d + 1 }
else if (original_data_row$Wood_major_group =="Dicots - core rosids") { row$rosids = row$rosids + 1 }
else if (original_data_row$Wood_major_group =="Dicots - core asterids") { row$asterids = row$asterids + 1 }
return (row)
}
for (i in 1:nrow(data)){
print(i)
res = get_biomes(data[i,]) # get species' biomes
biomes = res[[1]]; occurrences = res[[2]]
remove_ind = NULL
if ("98" %in% biomes){ # remove "98" from biomes
remove_ind = which(biomes=="98")
biomes = biomes[-remove_ind]
occurrences = occurrences[-remove_ind]
}
if ("99" %in% biomes){ # remove "99" from biomes
remove_ind = which(biomes=="99")
biomes = biomes[-remove_ind]
occurrences = occurrences[-remove_ind]
}
# do not consider species with less than 5 occurrences
remove_ind = NULL
if (length(which(as.numeric(occurrences)<5))>0){
remove_ind = which(as.numeric(occurrences)<5)
occurrences = occurrences[-remove_ind]
biomes = biomes[-remove_ind]
}
data$biomes[i] = length(biomes)
for (b in 1:length(biomes)){ # iterate over all biomes
row_ind = which(biome_data$biome==biomes[b])
updated_row = update_count(biome_data[row_ind,],data[i,],as.numeric(occurrences[b]))
biome_data[row_ind,] = updated_row
}
remove_ind = NULL
if (is.na(data$eco_id[i])){ # possible that there are biomes without ecoregions
next
}
res = get_ecoregions(data[i,]) # get species' ecoregions
ecoregions = res[[1]]; occurrences = res[[2]]
if ("-9998" %in% ecoregions){ # remove "-9998" from biomes
remove_ind = which(ecoregions=="-9998")
ecoregions = ecoregions[-remove_ind]
occurrences = occurrences[-remove_ind]
}
if ("-9999" %in% ecoregions){ # remove "-9999" from biomes
remove_ind = which(ecoregions=="-9999")
ecoregions = ecoregions[-remove_ind]
occurrences = occurrences[-remove_ind]
}
# do not consider species with less than 5 occurrences
remove_ind = NULL
if (length(which(as.numeric(occurrences)<5))>0){
remove_ind = which(as.numeric(occurrences)<5)
occurrences = occurrences[-remove_ind]
ecoregions = ecoregions[-remove_ind]
}
data$ecoregions[i] = length(ecoregions)
for (e in 1:length(ecoregions)){
row_ind = which(eco_data$eco_id==ecoregions[e])
updated_row = update_count(eco_data[row_ind,],data[i,],as.numeric(occurrences[e]))
eco_data[row_ind,] = updated_row
}
}
# add biome number to each ecoregion
biome_data$pp_perc = biome_data$pp/biome_data$sp*100 # pp_perc in each ecoregion
biome_data$pp_perc_resolved = biome_data$pp/(biome_data$pp + biome_data$dp)*100
eco_data$pp_perc_sp = eco_data$pp/eco_data$sp*100 # pp_perc in each ecoregion
eco_data$pp_perc_resolved = eco_data$pp/(eco_data$pp + eco_data$dp)*100
write.csv(file = output1,biome_data,row.names=F)
write.csv(file = output2,eco_data,row.names=F)