forked from idiv-biodiversity/2019_big_data_biogeography
-
Notifications
You must be signed in to change notification settings - Fork 0
/
fr_plots.Rmd
263 lines (208 loc) · 8.17 KB
/
fr_plots.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
---
output: html_document
---
```{r global_options, include=FALSE}
knitr::opts_chunk$set(fig.width=12, fig.height=8, eval = FALSE,
echo=TRUE, warning=FALSE, message=FALSE,
tidy = TRUE, collapse = TRUE,
results = 'hold')
```
# Background
Clear and appealing vizualization is essential to communicate scientific results. One of the main advantages of R is that it makes it easy to visualize data. In the last years, especially the grmmar of graphics framework has produced many different visualization methods.
# Objective
This exercise will provide you with scripts tpo plot speces richness maps and ancestral area reconstructions from BioGeoBEARS.
# Exercise
# Libraries
```{r}
library(ape)
library(ggtree)
library(tidyverse)
library(geiger)
library(stringr)
library(RColorBrewer)
library(colorspace)
library(jpeg)
library(viridis)
```
# Tutorial
## Plotting species richness maps
## PLotting results from INfomap bioregions
```{r}
#recode the bioregions names
bior.plot$id <- recode_factor(bior.plot$id,
"1 0" = "Northern South America",
"2 0" = "Caribbean",
"3 0" = "Central South America",
"4 0" = "Africa",
"5 0" = "Eastern South America",
"6 0" = "Australasia",
"7 0" = "North America",
"8 0" = "Madagascar"
)
bior.plot$id <- factor(bior.plot$id ,levels(bior.plot$id )[c(7, 2, 1, 3, 5, 4, 8, 6)])
proj4string(landmass) <- "+proj=longlat +ellps=WGS84 +datum=WGS84 +no_defs"
landmass <- spTransform(landmass, w3)
mapWorld <- fortify(landmass)
cols <- brewer.pal(n = 8, name = "Set1")#[-9]
#to lump SA
#cols[c(3:5)] <- cols[3]
ggplot()+
#mapWorld+
geom_polygon(data = mapWorld, aes(x = long, y = lat, group = group), fill = "grey80")+
geom_polygon(data = bior.plot, aes(x = long, y = lat, group = group, fill = id))+
scale_fill_manual(values = cols)+
#coord_map("cylequalarea", parameters = 30, xlim=c(-180, 180))+
theme_bw()+
coord_fixed()+
#coord_map("cylequalarea", parameters = 30, xlim=c(-180, 180))
theme(
legend.position = c(0,0),
legend.title = element_blank(),
legend.justification = c(-0.3,-0.3),
legend.text = element_text(size = 16),
axis.ticks = element_blank(),
axis.text = element_blank(),
axis.title = element_blank(),
panel.grid = element_blank()
)
```
# Plotting ancestral area reconstruction
## Load the data
```{r}
# The range probabilities
range_probabilities <- read.table("example_data/bombacoideae_DEC_node_probabilities.txt", header = TRUE)
# The phylogenetic tree
tr <- read.tree("example_data/bombacoideae_phylogeny.tre")
# The tip classification used for the DEC model
traits <- read.table("example_data/bombacoideae_biome_classification.txt", header = TRUE, skip = 1)
# A map of the biomes
map <- readJPEG("example_data/Fig1b_bioregionalization_reduced.jpg")
```
## Set the paramaeters for the plotting
```{r}
n_states <- 5
state_names <- c("A", "B", "C", "D", "E")
```
```{r}
#node states
states <- range_probabilities%>%
select(-c(1))%>%
mutate(node = parse_number(rownames(.)))
#only keep the most likely states and fuse others to "others"
plo.states <- list()
for(i in 1:nrow(states)){
sub <- data.frame(t(slice(states, i)))
ord <- rev(order(sub[,1]))
out <- data.frame(t(sub[ord,][1:2]))
names(out) <- rownames(sub)[ord][1:2]
plo.states[[i]] <- round(out,3)
}
plo.states <- bind_rows(plo.states)
plo.states[is.na(plo.states)] <- 0
plo.states$OTHERS <- 1-rowSums(plo.states[,-1])
plo.states <- plo.states[-c(1:Ntip(tr)),-1]
plo.states <- plo.states[, colSums(plo.states) > 0]
areas <- names(plo.states)
names(plo.states) <- seq(1:ncol(plo.states))
# write.csv(plo.states, "output/node.states.csv")
# plo.states <- read_csv("output/node.states.csv")
#tip states
tip.states <- traits
names(tip.states) <- c("id", "a")
tips <- data.frame(row.names= tip.states$id,
str_split_fixed(string = tip.states$a, n = n_states, pattern = ""))
names(tips) <- state_names
brc <- c(viridis(n = n_states), rainbow(n = (ncol(states) - n_states)))
for(i in 1:n_states){
tips[,i] <- as.numeric(tips[,i])
tips[tips[,i] != 1, i] <- "grey90"
tips[tips[,i] == 1, i] <- brc[i]
}
rownames(tips) <-tip.states$id
# #new tip labels
# tl <- read.csv("input/FINAL_Bombacoids_traits_input_1707.csv")[, c("id", "species")]
# rownames(tl) <- tl$id
# tl <- tl[tr$tip.label,]
#
# tr$tip.label <- gsub("_", " ", as.character(tl$species))
#
# tlp <- tl[tip.states$id,]
# rownames(tips) <- gsub("_", " ", as.character(tlp$species))
#pick nodes for pie label size
#from revells blog
getDescendants<-function(tree,node,curr=NULL){
if(is.null(curr)) curr<-vector()
daughters<-tree$edge[which(tree$edge[,1]==node),2]
curr<-c(curr,daughters)
w<-which(daughters>=length(tree$tip))
if(length(w)>0) for(i in 1:length(w))
curr<-getDescendants(tree,daughters[w[i]],curr)
return(curr)
}
#geological epochs
epo.lim <- data.frame(x = c(-2.58, -5.33, -23.03, -33.09, -56, -66))
epo.lab <- data.frame(x = c(-2.58/2, -2.58 - (5.33-2.58) / 2 ,
-5.33 - (23.03 - 5.33) / 2 ,
-23.03 - (33.09 - 23.03) / 2 ,
-33.09 - (56 - 33.09) / 2,
-56 - (66 - 56) / 2,
-66 - (max(node.depth.edgelength(tr) - 66) / 2)),
lab = c("Q.", "Pl.", "Miocene", "Oligocene", "Eocene", "Paleocene", "Late Cretaceous"))
epo.lab <- epo.lab[1:5,]
epo.lim <- epo.lim + max(node.depth.edgelength(tr))
epo.lab$x <- epo.lab$x + max(node.depth.edgelength(tr))
#check the nodes of which the descendants should be small instead of large
plot(tr, cex = 0.5)
nodelabels()
sno <- c(getDescendants(tr, 189), getDescendants(tr, 167), getDescendants(tr,127), getDescendants(tr, 114),
getDescendants(tr, 118), 107, 111, 112, 163, 164)
sno <- sno[!sno %in% 1:Ntip(tr)]
lno <- as.numeric(rownames(plo.states)[!as.numeric(rownames(plo.states)) %in% sno])
lno <- lno[!lno %in% 1:Ntip(tr)]
#lno <- lno[!lno %in% c(114,116, 117, 115, getDescendants(tr, 117))]#exclude the root and outgroup
#mix node colors
names(brc)[c(1:n_states)] <- state_names
node.col <- brc[areas]
names(node.col) <- areas
rains <- brc[-c(1:n_states)]
node.col[is.na(node.col)] <- rains
node.col["OTHERS"] <- "grey50"
#colours for the legend
leg <- node.col[order(nchar(names(node.col)))]
names(leg) <- c(names(states)[-length(names(states))], "OTHERS")
#base plotting
plot(tr, label.offset = 9.5, x.lim = 116, cex = 0.6, y.lim = 107)
segments(x0 = epo.lim$x, y0 = -8, x1 = epo.lim$x, y1 = 105, col = "grey50")
text(x = epo.lab$x, y = -1.5, labels = epo.lab$lab, cex = 0.6)
#tip labels
tiplabels(pch = 22, bg = tips[tr$tip.label,1], adj = 3.5, col = tips[tr$tip.label,1], cex = 0.8)
tiplabels(pch = 22, bg = tips[tr$tip.label,1], adj = 4.5, col = tips[tr$tip.label,1], cex = 0.8)
tiplabels(pch = 22, bg = tips[tr$tip.label,2], adj = 5.5, col = tips[tr$tip.label,2], cex = 0.8)
tiplabels(pch = 22, bg = tips[tr$tip.label,2], adj = 6.5, col = tips[tr$tip.label,2], cex = 0.8)
tiplabels(pch = 22, bg = tips[tr$tip.label,3], adj = 7.5, col = tips[tr$tip.label,3], cex = 0.8)
tiplabels(pch = 22, bg = tips[tr$tip.label,3], adj = 8.5, col = tips[tr$tip.label,3], cex = 0.8)
tiplabels(pch = 22, bg = tips[tr$tip.label,4], adj = 9.5, col = tips[tr$tip.label,4], cex = 0.8)
tiplabels(pch = 22, bg = tips[tr$tip.label,4], adj = 10.5, col = tips[tr$tip.label,4], cex = 0.8)
# tip label heading
text(labels = c("MBF", "DBF", "SAV"),
x = seq(60, 64, by = 2), y = 104, cex = 0.5, srt = 90, adj = 0)
par(fg = "transparent")
# node label pie-charts
nodelabels(pie= as.matrix(plo.states[as.numeric(rownames(plo.states)) %in% lno,]),
node = lno, piecol=node.col, cex = 1)
#for whatever reason this works, but vectorized nodelabels does not
for(i in 1:length(sno)){
print(i)
nodelabels(pie= as.matrix(plo.states[as.numeric(rownames(plo.states)) %in% sno[i],]),
node = sno[i], piecol=node.col, cex = 0.5)
}
par(fg = "black")
axisPhylo(1)
rasterImage(map,
xleft = -5,
ybottom = 95,
xright = 45 ,
ytop = 90 + 50 * (615/1027) * (8 / 11))
legend(x = -5, y = 100, ncol = 1, legend = names(leg),
fill = leg, cex = 0.9, bty = "n")
```