Created
April 24, 2019 07:30
-
-
Save guidocor/26620367f6a4e478957ce06e32fd44b0 to your computer and use it in GitHub Desktop.
Clusters con la valoración media de los políticos. Encuesta preelectoral 2019
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
###################################################################### | |
# Cluster Analysis de la valoración | |
###################################################################### | |
rm(list=ls()); gc() | |
options(scipen=20) | |
if (!require("pacman")) install.packages("pacman") | |
pacman::p_load("tidyverse", "psych", "haven", "reshape2", "qgraph", "Rmisc", "Hmisc", "mclust", "corrplot") | |
# Sacamos del CIS la encuesta | |
dataset <- read_sav("3242.sav") | |
# Este códio ha sido posible gracias el siguiente post: | |
# https://willhipson.netlify.com/post/latent-profile/latent-profile/ | |
# Nos interesan las preguntas de simpatía, asi que vamos por la que empiezan por P11 | |
df = dataset %>% dplyr::select(starts_with("P11")) | |
colnames(df) <- c("Abascal", "Casado", "Garzón", "Iglesias", "Rivera", "Sánchez") | |
df$id = 1:nrow(df) | |
df = gather(df, key = "Simpatía", value = "valoración", 1:6) | |
df[df$valoración == 99 | df$valoración == 98 | df$valoración == 97, "valoración"] <- NA | |
df <- spread(df, key = "Simpatía", value = "valoración") | |
df <- select(df, -id) | |
df <- df[complete.cases(df),] | |
# Obtenemos el BIC de los modelos | |
BIC <- mclustBIC(df) | |
BIC %>% summary | |
plot(BIC) | |
# Aparecen 9 perfiles, pero por economía cognitiva modelaremos solo 6 | |
mod1 <- Mclust(df, modelNames = "EEV", G = 9, x = BIC) | |
summary(mod1) | |
# Creamos el data.frame de los parametros | |
means <- data.frame(mod1$parameters$mean, stringsAsFactors = FALSE) %>% | |
rownames_to_column() %>% | |
melt(id.vars = "rowname", variable.name = "Profile", value.name = "Mean") %>% | |
mutate(Mean = round(Mean, 2)) | |
# Creamos los porcentaes de cada grupo | |
n = data.frame(Profile = mod1$classification) %>% | |
group_by(Profile) %>% count("Profile") %>% ungroup() %>% | |
mutate(freq = round((freq/sum(freq))*100, 0) ) | |
n$Profile <- paste0("X", n$Profile) | |
# unimos | |
means <- left_join(means, n) | |
# A plotear! | |
means %>% | |
ggplot(aes(rowname, Mean, group = Profile, color = rowname)) + | |
geom_line(size = 1.25, alpha = 0.3, color = "black") + | |
geom_point(size = 4.25) + | |
labs(x = NULL, y = "Valoración media") + | |
theme_bw(base_size = 14) + | |
facet_wrap(~ Profile) + | |
theme(axis.text.x = element_text(angle = 45, hjust = 1), legend.position = "none") + | |
scale_x_discrete(limits = c("Abascal", "Casado", "Rivera", "Sánchez", "Iglesias", "Garzón")) + | |
geom_text(aes(x = 5, y = 5, label = paste0(freq, "%" )), color = "black") + | |
ggtitle("Perfiles basados en las valoraciones a los principales políticos \n N = 9544, Fuente: CIS Abril 2019") + | |
scale_color_manual(values = c("Abascal"= "green4", "Casado" = "dodgerblue", | |
"Rivera" = "orange", "Sánchez" = "firebrick", | |
"Iglesias" = "mediumpurple", "Garzón" = "red")) | |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment