OccupationPCAs/4_CheckDim.R at main · juchiyu/OccupationPCAs · GitHub

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
#==== Summary ====
# This file reads the results of PCAs and clustering analyses and create Figures 2-4
#

# ---- Clean up ----
rm(list = ls())
graphics.off()

# ---- Packages ----
library(ggplot2)
library(ggdendro)
library(tidyverse)
library(plyr)
library(ggrepel)
library(ggpubr)
library(gridExtra)
library(grid)
library(devtools)
library(readxl)
library(ExPosition)
library(InPosition)
library(ggplot2)
library(PTCA4CATA)
library(dplyr)
library(data4PCCAR)
library(factoextra)
library(dendextend)
library(stringr)

func.dir <- ("./functions/")
sapply(paste0(func.dir,list.files(func.dir, pattern = "\\.[Rr]$")), source)


# ---- Read.results ----
load("Results/from2_Dat4Plot.rda")

# ---- get.job.zones.color ----
job.zone.col <- c(`1` = "darkgoldenrod2",
                  `2` = "orange2",
                  `3` = "darkorange2",
                  `4` = "darkorange3",
                  `5` = "darkorange4")
jb.col_45 <- jb.col_123 <- jb.col_all <- list()
jb.col_45$gc <- jb.col_123$gc <- jb.col_all$gc <- job.zone.col %>% as.matrix

jb.col_all$oc <- recode(onet.jz_all, !!!job.zone.col)
jb.col_45$oc <- recode(onet.jz_45, !!!job.zone.col)
jb.col_123$oc <- recode(onet.jz_123, !!!job.zone.col)

# ---- fix all trait names (remove .) ----
rownames(pcares_all$fj) <- gsub("\\.", " ", rownames(pcares_all$fj))
rownames(pcares_45$fj) <- gsub("\\.", " ", rownames(pcares_45$fj))
rownames(pcares_123$fj) <- gsub("\\.", " ", rownames(pcares_123$fj))

# ---- Tables ----
TopBottomTable <- function(f, col.list, leDim = 1, show.n = 3, str4wrap = 30, column.name = "Occupations", row.label = c("Right", "Left")){
  # leDim = 1
  # show.n = 10

  order.f <- order(f[,leDim])
  order.fcol <- col.list$oc[order.f,]

  fmax <- f[order.f,leDim] %>% tail(show.n) %>% rev %>% round(2)
  fmax.name <- names(fmax) %>% str_wrap(str4wrap)
  fmax.col <- order.fcol %>% tail(show.n) %>% rev

  fmin <- f[order.f,leDim] %>% head(show.n) %>% round(2)
  fmin.name <- names(fmin) %>% str_wrap(str4wrap)
  fmin.col <- order.fcol %>% head(show.n)

  tab <- list()

  tab$top <- cbind(Scores = fmax,
                   fmax.name,
                   color = fmax.col) %>% as.data.frame

  tab$bottom <- cbind(Scores = c(fmin),
                      c(fmin.name),
                      color = c(fmin.col)) %>% as.data.frame

  rownames(tab$top) <- c(paste0(row.label[1]," ",1), c(2:show.n))
  rownames(tab$bottom) <- c(paste0(row.label[2], " ",1), paste0(" ",c(2:show.n)))

  colnames(tab$top)[2] <- column.name
  colnames(tab$bottom)[2] <- column.name

  return(tab)
}

## All job zones - Occupation

fi.tab_all.cp1 <- TopBottomTable(pcares_all$fi, occu.clust$all$col, show.n = 25)
fi.tab_all.cp2 <- TopBottomTable(pcares_all$fi, occu.clust$all$col, leDim = 2, row.label = vertical.label)
fi.tab_all.cp3 <- TopBottomTable(pcares_all$fi, occu.clust$all$col, leDim = 3, row.label = vertical.label)

## All job zones - traits

fj.tab_all.cp1 <- TopBottomTable(pcares_all$fj, trt.clust$all$col, column.name = "Job traits")
fj.tab_all.cp2 <- TopBottomTable(pcares_all$fj, trt.clust$all$col, column.name = "Job traits", leDim = 2, row.label = vertical.label)
fj.tab_all.cp3 <- TopBottomTable(pcares_all$fj, trt.clust$all$col, column.name = "Job traits", leDim = 3, row.label = vertical.label)

## Job Zones 45 - Occupation

fi.tab_jz45.cp1 <- TopBottomTable(pcares_45$fi, occu.clust$jz45$col)
fi.tab_jz45.cp2 <- TopBottomTable(pcares_45$fi, occu.clust$jz45$col, leDim = 2, row.label = vertical.label)
fi.tab_jz45.cp3 <- TopBottomTable(pcares_45$fi, occu.clust$jz45$col, leDim = 3, row.label = vertical.label)

## Job zones 45 - traits

fj.tab_jz45.cp1 <- TopBottomTable(pcares_45$fj, trt.clust$jz45$col, column.name = "Job traits")
fj.tab_jz45.cp2 <- TopBottomTable(pcares_45$fj, trt.clust$jz45$col, column.name = "Job traits", leDim = 2, row.label = vertical.label)
fj.tab_jz45.cp3 <- TopBottomTable(pcares_45$fj, trt.clust$jz45$col, column.name = "Job traits", leDim = 3, row.label = vertical.label)

## Job zones 123 - Occupation

fi.tab_jz123.cp1 <- TopBottomTable(pcares_123$fi, occu.clust$jz123$col)
fi.tab_jz123.cp2 <- TopBottomTable(pcares_123$fi, occu.clust$jz123$col, leDim = 2, row.label = vertical.label)
fi.tab_jz123.cp3 <- TopBottomTable(pcares_123$fi, occu.clust$jz123$col, leDim = 3, row.label = vertical.label)

## Job zones 123 - traits

fj.tab_jz123.cp1 <- TopBottomTable(pcares_123$fj, trt.clust$jz123$col, column.name = "Job traits")
fj.tab_jz123.cp2 <- TopBottomTable(pcares_123$fj, trt.clust$jz123$col, column.name = "Job traits", leDim = 2, row.label = vertical.label)
fj.tab_jz123.cp3 <- TopBottomTable(pcares_123$fj, trt.clust$jz123$col, column.name = "Job traits", leDim = 3, row.label = vertical.label)


#---- job.zone.plots ----
get.FacPlot <- function(eigres, pcares, design, col.list, xaxis = 1, yaxis = 2, TiLab = FALSE,
                        alpha.indiv.point = 0.1, alpha.mean.point = 0.9,
                        cex.indiv.point = 5, cex.mean.point = 10, cex.mean.text = 10, label.size = 20, title = "", title.size = 10,
                        print.list = FALSE, list.title = "occupations", str4wrap = 30, list.size = 10, list.ybase = 0, list.scale.y.ony = c(0.1,0,-0.1), list.scale.y.onx = c(0.2,0.25,0.30), list.xmin.scale = 0.15, list.scale.ymin = 0.1){
  jbplt <- PlotFactor(lambda = eigres$eig, tau = eigres$tau,
                      xaxis = xaxis, yaxis = yaxis,
                      f = pcares, design = design, title = title,
                      col.list = col.list, alpha.indiv.point = alpha.indiv.point, alpha.mean.point = alpha.mean.point,
                      cex.indiv.point = cex.indiv.point, cex.mean.point = cex.mean.point, cex.mean.text = cex.mean.text)
  if(isFALSE(TiLab)){
    plot.res <- jbplt$f.map$zeMap_background + jbplt$f.map$zeMap_dots +
      jbplt$mean.map$zeMap_dots + jbplt$mean.map$zeMap_text + jbplt$label +
      theme(text = element_text(size = label.size))
  }else{
    plot.res <- jbplt$f.map$zeMap_background + jbplt$f.map$zeMap_dots +
      jbplt$TI +
      jbplt$mean.map$zeMap_text + jbplt$label +
      theme(text = element_text(size = label.size))
  }

  if(print.list == TRUE){
    x.range <- jbplt$f.map$constraints$maxx - jbplt$f.map$constraints$minx
    y.range <- jbplt$f.map$constraints$maxy - jbplt$f.map$constraints$miny
    x.tab <- TopBottomTable(pcares, col.list, column.name = "text", leDim = xaxis, str4wrap = str4wrap)
    y.tab <- TopBottomTable(pcares, col.list, column.name = "text", leDim = yaxis, str4wrap = 100) #100
    # title.tab <- matrix(rep(c("NA", sprintf("~underline('Most contributing %s')", list.title), "black"),4), nrow = 4, byrow = T,
    #                      dimnames = list(c(), c("Score", "text", "color")))
    text2print <- rbind(x.tab$top, x.tab$bottom, y.tab$top, y.tab$bottom) %>% data.frame
    text2print$x <- c(rep(jbplt$f.map$constraints$maxx + 0.1*x.range, 3), # max on dim1
                      rep(jbplt$f.map$constraints$minx - list.xmin.scale*x.range, 3), # min on dim1
                      rep(0, 6) # max and min on dim2
    )
    text2print$y <- c(rep(c(list.scale.y.ony*y.range), 2) + list.ybase, # max and min on dim1
                      c(rep(jbplt$f.map$constraints$maxy , 3) + rev(list.scale.y.onx)*y.range), # max on dim2
                      c(rep(jbplt$f.map$constraints$miny-list.scale.ymin*y.range, 3) - list.scale.y.onx*y.range) # min on dim2
    )
    text2print$hjust <- c(rep("left",3),
                          rep("right",3),
                          rep("middle",3),
                          rep("middle",3))
    # title2print <- title.tab %>% data.frame
    # title2print$x <- c(jbplt$f.map$constraints$maxx + 0.1*x.range, # max on dim1
    #                    jbplt$f.map$constraints$minx - list.xmin.scale*x.range, # min on dim1
    #                    rep(0, 2)) # max and min on dim2
    # title2print$y <- c(rep((2*list.scale.y.ony[1]-list.scale.y.ony[2])*y.range + list.ybase,2),
    #                    jbplt$f.map$constraints$maxy + (2*list.scale.y.onx[3]-list.scale.y.onx[2])*y.range,
    #                    jbplt$f.map$constraints$miny-list.scale.ymin*y.range - (2*list.scale.y.onx[1] - list.scale.y.onx[2])*y.range
    #                    )
    # title2print$hjust <- c("left", "right", "middle", "middle")
    plot.res <- plot.res +
      geom_text(data = text2print, x = text2print$x, y = text2print$y,
                size = list.size, label = text2print$text,
                color = text2print$color, hjust = text2print$hjust) +
      # geom_text(data = title2print, x = title2print$x, y = title2print$y,
      #           size = list.size, label = title2print$text,
      #           color = title2print$color, hjust = title2print$hjust, parse = TRUE) +
      theme(plot.title = element_text(size = title.size), axis.title = element_text(size = title.size), plot.margin = unit(c(2.5,5,2.5,5), "cm")) + coord_fixed(ratio = 1, clip = "off")
  }
  return(plot.res)
}

# ---- fi.plots ----
fi.point.size = 5; fi.text.size = 5; title.size = 30
xaxis = 4; yaxis = 5

## General PCA
get.FacPlot(eigres_all, pcares_all$fi, design = occu.clust$all$list, occu.clust$all$col, xaxis = xaxis, yaxis = yaxis, cex.mean.point = fi.point.size, cex.indiv.point = fi.point.size-2, cex.mean.text = fi.text.size, title.size = title.size, title = "Occupation clusters", print.list = FALSE,
                         str4wrap = 30, list.xmin.scale = 0.2, list.scale.y.ony = c(0.2,0,-0.2), list.scale.y.onx = c(0.1,0.16,0.23)+0.05, list.scale.ymin = 0.02, list.size = 12)
## Cognitive PCA
get.FacPlot(eigres_45, pcares_45$fi, design = occu.clust$jz45$list, occu.clust$jz45$col, xaxis = xaxis, yaxis = yaxis, cex.mean.point = fi.point.size, cex.indiv.point = fi.point.size-2, cex.mean.text = fi.text.size, title.size = title.size, title = "Occupation clusters", print.list = FALSE,
                          str4wrap = 30, list.ybase = 3, list.xmin.scale = 0.2, list.scale.y.onx = c(0.1,0.16,0.22)+0.05, list.scale.ymin = 0.03, list.scale.y.ony = c(0.25,0,-0.25), list.size = 12)
## Labour PCA
get.FacPlot(eigres_123, pcares_123$fi, design = occu.clust$jz123$list, occu.clust$jz123$col, xaxis = xaxis, yaxis = yaxis,cex.mean.point = fi.point.size, cex.indiv.point = fi.point.size-2, cex.mean.text = fi.text.size, title.size = title.size, title = "Occupation clusters", print.list = FALSE,
                           str4wrap = 30, list.xmin.scale = 0.2, list.scale.ymin = 0.03, list.scale.y.onx = c(0.1,0.18,0.26)+0.05, list.scale.y.ony = c(0.15,0,-0.15), list.size = 12)
# ---- fj.plots ----
fj.point.size = 5; fj.text.size = 5

## General PCA
get.FacPlot(eigres_all, pcares_all$fj, design = trt.clust$all$list, trt.clust$all$col, xaxis = xaxis, yaxis = yaxis, cex.mean.point = fj.point.size, cex.indiv.point = fi.point.size-2, cex.mean.text = fj.text.size, TiLab = TRUE, title.size = title.size, title = "Trait clusters", print.list = FALSE, list.title = "traits",
                         str4wrap = 30, list.ybase = 24, list.scale.y.ony = c(0.15,0,-0.15), list.scale.y.onx = c(0.1,0.16,0.22)+0.05, list.scale.ymin = 0.05, list.size = 12)
## Cognitive PCA
get.FacPlot(eigres_45, pcares_45$fj, design = trt.clust$jz45$list, trt.clust$jz45$col, xaxis = xaxis, yaxis = yaxis, cex.mean.point = fj.point.size, cex.indiv.point = fi.point.size-2, cex.mean.text = fj.text.size, TiLab = TRUE, title.size = title.size, title = "Trait clusters", print.list = FALSE, list.title = "traits",
                          list.ybase = 10, list.xmin.scale = 0.2, list.scale.y.ony = c(0.15,0,-0.15), list.scale.ymin = 0.04, list.scale.y.onx = c(0.1,0.16,0.22)+0.05, list.size = 12)

## Labour PCA
get.FacPlot(eigres_123, pcares_123$fj, design = trt.clust$jz123$list, trt.clust$jz123$col, xaxis = xaxis, yaxis = yaxis,cex.mean.point = fj.point.size, cex.indiv.point = fi.point.size-2, cex.mean.text = fj.text.size, TiLab = TRUE, title.size = title.size, title = "Trait clusters", print.list = FALSE, list.title = "traits",
                           str4wrap = 25, list.ybase = 5, list.scale.ymin = 0.03, list.scale.y.ony = c(0.20,0,-0.20), list.scale.y.onx = c(0.1,0.17,0.24)+0.05, list.size = 12)