In joebrew/vilaweb: Reproducible data analysis for www.vilaweb.cat

# Basic knitr options
library(knitr)
opts_chunk$set(comment = NA, 
               echo = FALSE, 
               warning = FALSE, 
               message = FALSE, 
               error = TRUE, 
               cache = FALSE,
               fig.width = 8.64,
               fig.height = 4.86,
               fig.path = 'figures/')

# Libraries
library(vilaweb)
library(tidyverse)
library(ggplot2)
library(tidyr)
library(databrew)
# library(cowplot)
library(ggthemes)

# Source document: http://datos.cis.es/pdf/Es3257ccaa_A.pdf

# Define ccaas
ccaas <- c('Andalucía',
           'Aragón',
           'Asturias',
           'Balears', 
           'Canarias', 
           'Cantabria', 
           'Castilla-\nLa Mancha',
           'Castilla\ny León',
           'Cataluña',
           'Comunitat\nValenciana',
           'Extremadura','Galicia',
           'Madrid',
           'Murcia',
           'Navarra',
           'País Vasco',
           'La Rioja',
           'Ceuta',
           'Melilla')

val_plot <- function(vals = c(3.1, 3, 2.6, 3.6, 3.4, 3.7, 2.4, 2.9, 4, 3.5, 2.7, 3.1, 3.1, 2.7, 3.9, 4.3, 2.6, 1.5, 2.2),
                     order_it = TRUE,
                     remove_small = TRUE){
  df <- tibble(ccaa = ccaas,
               val = vals)
  if(remove_small){
    df <- df %>%
      filter(!ccaa %in% c('Melilla', 'Ceuta'))
  }
  if(order_it){
    df <- df %>% arrange(desc(ccaa)) %>%
      arrange(desc(val))
    df$ccaa <- factor(df$ccaa, levels = df$ccaa)
  }
  ggplot(data = df,
         aes(x = ccaa,
             y = val)) +
    geom_point(size = 3) +
    geom_segment(aes(xend = ccaa,
                     yend = 0),
                 alpha = 0.6) +
        theme_databrew() +
    theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5,
                                     size = 15),
          plot.title = element_text(hjust = 0.5, size = 22),
          plot.caption = element_text(size = 14)) +
    scale_y_continuous(name ='Valoración (1 = muy mal, 10 = muy bien)', breaks = NULL) +
      geom_text(aes(y = val, label = val),
                size = 5,
                nudge_y = max(df$val) * 0.1,
                alpha = 0.5) +
        labs(x = '',
             caption = 'Datos: CIS | Gráfico: @joethebrew')  
}

val_plot() +
  labs(title = 'Valoración de Pablo Iglesias por Comunidad Autónoma')

val_plot(vals = c(2.5, 2.2, 2.4, 2.2, 2.4, 1.9, 2.7, 2.3, 1.4, 2.3, 2.3, 1.9, 2.6, 2.6, 1.5, 1.4, 2.2, 2.7, 3.8)) +
  labs(title = 'Valoración de Santiago Abascal por Comunidad Autónoma')

val_plot(vals = c(3.5, 3.5, 3.2, 3, 3.6, 2.7, 3.7, 4.1, 1.8, 3.2, 3.9, 3.2, 3.3, 3.9, 2.2, 2, 4.4, 3.3, 5.8)) +
  labs(title = 'Valoración de Pablo Casado por Comunidad Autónoma')

val_plot(vals = c(3.8, 3.6, 3.4, 3.1, 3.8, 3.9, 3.7, 4, 2.2, 3.7, 3.6, 3.2, 3.7, 3.7, 2.4, 2.2, 3.2, 3.7, 6.2)) +
  labs(title = 'Valoración de Albert Rivera por Comunidad Autónoma')

val_plot(vals = c(4.8, 4.3, 4.1, 4.9, 5.2, 5.2, 4.3, 4.2, 4.2, 5, 4.9, 4.8, 4.5, 3.9, 4.2, 4.5, 3.7, 4.4, 5.2)) +
  labs(title = 'Valoración de Pedro Sánchez por Comunidad Autónoma')

religion <- tibble(religion = rep(c('Católico practicante',
                                'Católico no practicante',
                                'Creyente de otra religión',
                                'Agnóstico',
                                'Indiferente, no creyente',
                                'Ateo'), each = 19),
                   ccaa = rep(ccaas, 6),
                   val = c(23.6, 29.3, 37.8, 24.2, 22.7, 27.5, 34.8, 33.1, 10.9, 13.8, 25, 33.5, 20.4, 36.2, 28.2, 16.9, 40, 20, 20,
                           50.8, 43.9, 43.2, 33.3, 52.3, 45, 38.6, 40.8, 41.8, 49.7, 57.9, 48.5, 39.7, 42.5, 28.2, 37.8, 50, 40, 40,
                           2.4, 1.2, 0, 3, 2.3, 0, 1.5, 1.2, 2.2, 3.3, 1.3, 1.5, 3.1, 0, 2.6, 2.7, 0, 20, 40,
                           7, 4.9, 4.1, 9.1, 6.1, 7.5, 10.6, 8.9, 6, 5.3, 3.9, 2.6, 13.5, 6.2, 10.3, 10.8, 10, 0, 0,
                           8.7, 3.7, 1.4, 16.7, 7.6, 7.5, 4.5, 6.5, 11.3, 11.2, 1.3, 4.1, 6.9, 7.5, 20.5, 12.8, 0, 0, 0,
                           6.3, 9.8, 13.5, 13.6, 8.3, 10, 9.8, 8.9, 26.4, 15.8, 10.5, 9.3, 15, 7.5, 7.7, 18.9, 0, 20, 0 ))

religion <- religion %>%
  mutate(religion = ifelse(religion %in% c("Agnóstico", "Ateo", "Indiferente, no creyente"),
                           'Ateo, agnóstico,\nindiferente, o\nno creyente', religion)) %>%
  mutate(religion = ifelse(religion == 'Creyente de otra religión', 'Otra\nreligión',
                           ifelse(religion == 'Católico practicante', 'Católico\npracticante',
                                  ifelse(religion == 'Católico no practicante', 'Católico\nno practicante', religion ))))
religion$religion <- paste0('\n', religion$religion, '\n')
religion <- religion %>%
  group_by(religion, ccaa) %>%
  summarise(val = sum(val)) %>%
  group_by(ccaa) %>%
  mutate(p = val / sum(val) * 100) %>%
  ungroup %>%
  filter(!ccaa %in% c('Melilla', 'Ceuta'))
religion$religion <- factor(religion$religion, levels = rev(sort(unique(religion$religion))))

sample_size <- tibble(ccaa = ccaas,
                      n = c(543, 82, 74, 66, 132, 40, 132, 169, 450, 304, 76, 194, 393, 80, 39, 148, 20, 5, 5))

gr <- religion %>% 
  left_join(sample_size) %>%
  mutate(people = (p/100) * n) %>%
  group_by(catalunya = ccaa == 'Cataluña', 
           religion) %>%
  summarise(people = sum(people)) %>%
  group_by(catalunya) %>%
  mutate(p = people / sum(people) * 100)


cols <- RColorBrewer::brewer.pal(n = 4, name = 'Spectral')
cols[1] <- 'darkgrey'
ggplot(data = religion,
       aes(x = ccaa,
           y = p,
           fill = religion)) +
  geom_bar(stat = 'identity', position = position_stack()) +
  theme_databrew() +
   theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5,
                                     size = 12),
          plot.title = element_text(hjust = 0.5, size = 22),
          plot.caption = element_text(size = 14),
         legend.text = element_text(size = 13)) +
    scale_y_continuous(name ='Porcentaje', breaks = seq(0, 100, 10)) +
        labs(x = '',
             caption = 'Datos: CIS Julio 2019 | Excluidos los N.C. Tamaño de muestra: 2.952.\nGráfico: @joethebrew ')  +
  theme(legend.position = 'right') +
  scale_fill_manual(name = '',
                    values = cols) +
  geom_text(data = religion %>% filter(p >= 5),
            aes(label = round(p, digits = 1)),
            color = 'white',
            size = 2.5,
            alpha = 0.7,
            position = position_stack(vjust = 0.5)) +
  labs(title = 'Religión por comunidad autónoma')