-
Notifications
You must be signed in to change notification settings - Fork 0
/
02_analysis.R
263 lines (216 loc) · 8.83 KB
/
02_analysis.R
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
# Package -----------------------------------------------------------------
library(flextable)
library(officer)
library(tidyverse)
library(ggrepel)
library(patchwork)
library(FactoMineR)
library(factoextra)
library(missMDA)
library(fclust)
library(segregation)
# Load data ---------------------------------------------
if(!exists("AUX")) AUX <- list()
AUX$GRUPO <- read.csv(here::here("analysis/data/dimensions.csv"),
encoding = "latin1") |>
mutate(CLASE = factor(CLASE, levels = c("s", "n")),
DIM_LAB = factor(DIM_LAB, levels = unique(DIM_LAB) ) )
SCHOOLS <- read.csv(here::here("analysis/data/schools.csv"),
encoding = "latin1") |>
mutate(across(.cols = c(LQ_PRI, LQ_SUP),
.fns = ~factor(.x, levels = c( "Alto", "Medio", "Bajo"))),
COOP = factor(COOP, levels = c("c/PJ","s/PJ", "NO") ),
SUBS = factor(SUBS, levels = c("Estatal", "NO", "SI") ),
TIPO_HABITAT = factor(TIPO_HABITAT,
levels = c("Ciudad Central",
"Centro Administrativo y de Negocios",
"Residencial alto",
"Residencial medio",
"Residencial bajo",
"Conjunto Habitacional",
"Popular de Origen Informal")),
SECTOR = factor(SECTOR, levels = c("Estatal", "Privado"))
)
SEC_aux <- SCHOOLS |> select(-starts_with("N_"))
SEC_seg <- SCHOOLS |>select(ID_s, starts_with("N_"))
AUX$SCHOOLS <- SCHOOLS
rm(SCHOOLS)
# Grupos ------------------------------------------------------------------
GRUPO <- AUX$GRUPO |>
arrange(ACTIVA, DIM_LAB, CLASE) |>
count(ACTIVA, DIM_LAB, CLASE) |>
mutate(GRUPO = case_when(ACTIVA == 1 ~ as.character(DIM_LAB),
CLASE == "s" ~ paste0(DIM_LAB, " (sup.)"),
CLASE == "n" ~ paste0(DIM_LAB, " (sup. cualit.)") ),
id_g = 1:n())
# Impute missing values with PCA ------------------------
# Active variables names
varact <- AUX$GRUPO |>
filter(Tipo == "x") |>
select(VAR) |>
simplify() |>
unname()
# Impute marker
imputado <- is.na(SEC_aux[varact]) %>% apply(., 1, sum)
# Imputation
SEC_aux[varact] <- SEC_aux[varact] |>
imputePCA(scale = T, ncp = 4) |>
pluck("completeObs")
AUX$IMPUTA <- data.frame(ID_s = SEC_aux$ID_s, IMPUTADO = imputado)
rm(varact, imputado)
# MFA para segmentación --------------------------------------
SEC_MFA <- SEC_aux |>
select(-ID_s) |>
MFA(group = GRUPO$n, ncp = 10,
type = as.character(GRUPO$CLASE),
name.group = GRUPO$GRUPO,
num.group.sup = GRUPO$id_g[GRUPO$ACTIVA != 1],
graph = F)
rm(GRUPO)
# Armado de fuzzy cluster -----------------------------------------
SEC_CL <- Fclust(SEC_MFA$ind$coord[ , 1:4],
type = "gk",
k = 4, noise = T)
CL_ORDEN <- order(SEC_CL$H[, 1])
SEC_aux <- cbind(SEC_aux, cl = SEC_CL$clus[ , "Cluster"],
cl_p = SEC_CL$clus[ , "Membership degree"],
SEC_CL$U[ , CL_ORDEN]) %>%
mutate(cl = factor(cl, levels = CL_ORDEN,
labels = 1:4) )
colnames(SEC_aux)[str_starts(colnames(SEC_aux), "Clus.")] <- paste0("Clus.", 1:4)
# Indices de calidad de los agrupamientos ---------------------------------
# el indice PC y MPC son mediciones de "difusidad" (fuzzy)
I <- Fclust.index(SEC_CL, alpha = 1)
# Descripción de clusters:
D <- SEC_aux %>%
group_by(cl) %>%
summarise(Cl.Size = n(),
Cl.Student = sum(n),
No.Asig = sum(cl_p <= 0.5),
P.Asig = No.Asig/Cl.Size * 100,
deg.Min = min(cl_p),
deg.Max = max(cl_p),
deg.Av = mean(cl_p))
# Distancia entre cluster:
H <- SEC_CL$H[CL_ORDEN, ]
rownames(H) <- paste0 ("Clus ", 1:4)
HD <- round(dist(H), 2)
# Comparacion entre Sector y cluster
CSec <- Fclust.compare(VC = SEC_aux$SECTOR,
SEC_aux |>
select(starts_with("Clus.")) ) %>%
round(digits = 3)
# Tabla (Sub)Sector - cluster
t1 <- cbind(round(prop.table(xtabs(n ~ cl + SUBS, data = SEC_aux), 1) * 100, 1),
Total = xtabs(n ~ cl, data = SEC_aux) )
t1 <- t1 %>%
as_tibble(rownames = "Circuito") %>%
mutate(Circuito = paste("Clúster", Circuito),
Privado = NO + SI) %>%
select(Circuito, Estatal, Privado, SI, NO, Total)
CLUSTER <- list(H = H, Index = I, Descrip = D,
Distan = HD, Ind_Sec = CSec, Tab_Sec = t1)
rm(CL_ORDEN, CSec, I, D, H, HD, SEC_CL, t1)
# Descripción de los cluster ----------------------------------------------
dat <- SEC_aux %>%
select(-ID_s, -SECTOR, -COOP, -LQ_PRI,
-cl_p, -starts_with("Clus.")) |>
select(cl, everything()) |>
mutate(SUBS = factor(SUBS,
levels = c("Estatal", "SI", "NO"),
labels = c("Estatal", "Priv. Con Subsidio",
"Priv. Independiente")))
# Generar descripción de variables
CL_DESC <- catdes(dat, 1)
# Agregar Nombres Human-readable a variables
for(i in c("test.chi2", "quanti.var")){
CL_DESC[[i]] <- CL_DESC[[i]] %>%
as_tibble(rownames = "VAR") %>%
left_join(AUX$GRUPO |> select(VAR, IND_LAB),
by = "VAR")
}
# Agregar Nombres Human-readable a descripción de
# clúster por variables cuantitativas
for(i in 1:length(CL_DESC$quanti)){
CL_DESC$quanti[[i]] <- CL_DESC$quanti[[i]] %>%
as_tibble(rownames = "VAR") %>%
left_join(AUX$GRUPO |> select(VAR, IND_LAB),
by = "VAR")
}
names(CL_DESC$quanti) <- paste0("Clus.", 1:4)
# Agregar Nombres Human-readable a descripción de
# clúster por variables cualitativas
for(i in 1:length(CL_DESC$category )){
CL_DESC$category[[i]] <- CL_DESC$category[[i]] %>%
as_tibble(rownames = "V") %>%
mutate(VAR = str_split(V, "=", simplify = T)[ , 1],
CAT = str_split(V, "=", simplify = T)[ , 2],
CAT = case_when(CAT %in% c("FALSE", "NO") ~ "No",
CAT %in% c("TRUE", "SI") ~ "Sí",
T ~ CAT) ) %>%
left_join(AUX$GRUPO |> select(VAR, IND_LAB),
by = "VAR") %>%
filter(CAT != "NA" &
!(str_starts(VAR, "T_" ) & CAT == "No") &
!(str_starts(VAR, "TIT_" ) & CAT == "No") &
!(str_starts(VAR, "CONTINUIDAD_" ) &
CAT == "No")) %>%
mutate(IND_LAB = paste(IND_LAB, CAT, sep = ": "))
}
names(CL_DESC$category) <- paste0("Clus.", 1:4)
# Principales características por cluster
C_CUANTI <- C_CUALI <- data.frame()
for(i in 1:4){
#cat(paste("\n", "Circuito", i, "\n"))
c <- paste("Clúster", i)
b <- CL_DESC$quanti[[i]]
C_CUANTI[1:10, c] <- b[c(1:5, ((nrow(b)-4):nrow(b))), ] %>%
mutate(X_ = `Mean in category`,
X_ = ifelse(str_starts(IND_LAB, "%"),
X_ * 100, X_),
X_ = ifelse(X_ > 1, round(X_, 1), round(X_, 3)),
IND_LAB = paste(IND_LAB,
round(X_, 3),
sep = ": ")) %>%
select(IND_LAB)
b <- CL_DESC$category[[i]]
C_CUALI[1:10, c] <- b[c(1:5, ((nrow(b)-4):nrow(b))), ] %>%
mutate(IND_LAB = paste(IND_LAB, " (",
round(`Mod/Cla`, 1), "%)",
sep = "")) %>%
select(IND_LAB)
}
SEC_MFA$CL_DESC <- CL_DESC
rm(dat, i, c, b, CL_DESC)
# Segregación -------------------------------------------------------------
s_aux <- SEC_seg |>
rename_with(.cols = -ID_s,
.fn = ~str_remove(.x, pattern = "N_") ) |>
left_join(SEC_aux |> select(ID_s, SECTOR, cl, SUBS),
by = "ID_s") |>
filter(!is.na(cl)) |>
pivot_longer(cols = PRI:SUP, names_to = "group",
values_to = "peso")
# Global = Between Seg + Within Seg
SEGRE <- list()
SEGRE$t_f <- cbind(I = c("Sólo Agrupamiento", "Sólo Sector",
"Sólo Subsector",
"Agrupamiento + Sector"),
rbind(f_seg(db = s_aux, VAR = "cl"),
f_seg(db = s_aux, VAR = "SECTOR"),
f_seg(db = s_aux, VAR = "SUBS"),
f_seg(db = s_aux, VAR = c("cl", "SUBS")))) |>
mutate(Prop_ENTRE = round(Between / Global * 100, 1),
Prop_DENTRO = round(Within / Global * 100, 1)) %>%
select(I, Prop_ENTRE, Prop_DENTRO)
# Local
SEGRE$t_l <- mutual_local(s_aux, group = "group", unit = "cl",
weight = "peso", se =F, wide = T) %>%
select(cl, p, Inter = ls) %>%
mutate(cl = paste("Clúster", cl),
Intra = map_dbl(1:4,
\(x) mutual_total(data = s_aux[s_aux$cl == x, ],
group = "group", unit = "ID_s",
weight = "peso", se = F)$est[1] )
)
rm(s_aux, SEC_seg)