Adding labels to percentage stacked barplot ggplot2-CodePudding

I am new to ggplot and looking to get some help for a dataset I am making visualizations for.

Here is my current code:

#create plot
plot <- ggplot(newDoto, aes(y = pid3lean, weight = weight, fill = factor(Q29_1String, levels = c("Strongly disagree","Somewhat disagree", "Neither agree nor disagree", "Somewhat agree", "Strongly agree"))))   geom_bar(position = "fill", width = .732) 
#fix colors
plot <- plot   scale_fill_manual(values = c("Strongly disagree" = "#7D0000", "Somewhat disagree" = "#D70000","Neither agree nor disagree" = "#C0BEB8", "Somewhat agree" = "#008DCA", "Strongly agree" = "#00405B")) 
#fix grid
plot <- plot   guides(fill=guide_legend(title="29"))   theme_bw()   theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank())   theme(panel.border = element_blank())   theme(axis.ticks = element_blank())   theme(axis.title.y=element_blank())   theme(axis.title.x=element_blank())   theme(axis.text.x=element_blank())   theme(text=element_text(size=19,  family="serif"))   theme(axis.text.y = element_text(color="black"))   theme(legend.position = "top")   theme(legend.text=element_text(size=12)) 
#plot graph
plot

This creates this bar chart:

Right now the problem I am having is trying to add percentage labels on these bars. I want to add text that shows the percentage of each segment, centered and in white letters.

Unfortunately, I have been having some trouble adding geom_text, as it frequently gives me errors because I don't have an x variable and I'm not sure how to fix it, as the way I used fill is sort of peculiar compared to other ways I've seen it done with both x and y variables. I don't really know what I would even add for an x variable considering that the fill is the percentage for each type of response (different response types shown in levels).

Any help would be appreciated! Happy to answer any questions about the dataset if that is important.

Here is an example of what the two relevant columns look like (didn't use head because there's so many variables in this dataset). Basically they show which party a respondent is a part of and if they strongly agree, somewhat agree, etc.

Here is the output of dput for the two variables:

structure(list(pid3lean = structure(c("Democrats", "Democrats", 
"Democrats", "Democrats", "Independents", "Democrats", "Republicans", 
"Independents", "Republicans", "Democrats", "Democrats", "Independents", 
"Democrats", "Republicans", "Democrats", "Democrats", "Democrats", 
"Democrats", "Democrats", "Republicans"), label = "pid3lean", format.spss = "A13", display_width = 15L), 
    Q29_1String = structure(c(1L, 1L, 2L, 2L, 1L, 1L, 5L, 4L, 
    1L, 1L, 2L, 5L, 1L, 5L, 1L, 1L, 1L, 5L, 1L, 3L), .Label = c("Strongly agree", 
    "Somewhat agree", "Neither agree nor disagree", "Somewhat disagree", 
    "Strongly disagree"), class = "factor")), row.names = c(NA, 
-20L), class = c("tbl_df", "tbl", "data.frame"))

CodePudding user response：

To put the percentages in the middle of the bars, use position_fill(vjust = 0.5) and compute the proportions in the geom_text.

library(ggplot2)

colors <- c("#00405b", "#008dca", "#c0beb8", "#d70000", "#7d0000")
colors <- setNames(colors, levels(newDoto$Q29_1String))

ggplot(newDoto, aes(pid3lean, fill = Q29_1String))  
  geom_bar(position = position_fill())  
  geom_text(aes(label = paste0(..count../sum(..count..)*100, "%")),
            stat = "count",
            colour = "white",
            position = position_fill(vjust = 0.5))  
  scale_fill_manual(values = colors)  
  coord_flip()

Package scales has functions to format the percentages automatically.

ggplot(newDoto, aes(pid3lean, fill = Q29_1String))  
  geom_bar(position = position_fill())  
  geom_text(aes(label = scales::percent(..count../sum(..count..))),
            stat = "count",
            colour = "white",
            position = position_fill(vjust = 0.5))  
  scale_fill_manual(values = colors)  
  coord_flip()

CodePudding user response：

Here is an alternative approach:

Here we do the stats in the dataframe (calculate the percentages and change class to factor of Q29_1String
using geom_col
then use coord_flip
tweak the theme part

library(tidyverse)

df %>% 
  group_by(pid3lean) %>% 
  count(Q29_1String) %>% 
  ungroup() %>% 
  mutate(pct = n/sum(n)) %>% 
  mutate(Q29_1String = as.factor(Q29_1String)) %>% 
  ggplot(aes(x = pid3lean, y = pct, fill = Q29_1String))  
  geom_col(position = "fill", width = .732)  
  scale_fill_manual(values = c("Strongly disagree" = "#7D0000", "Somewhat disagree" = "#D70000","Neither agree nor disagree" = "#C0BEB8", "Somewhat agree" = "#008DCA", "Strongly agree" = "#00405B"))  
  coord_flip() 
  geom_text(aes(label = scales::percent(pct)), 
            position = position_fill(vjust = 0.5),size=5, color="white",
            )   guides(fill=guide_legend(title="29"))   
  theme_bw()   
  theme(panel.grid.major = element_blank(), 
        panel.grid.minor = element_blank(),
        panel.border = element_blank(), 
        axis.ticks = element_blank(), 
        axis.title.y=element_blank(), 
        axis.title.x=element_blank(), 
        axis.text.x=element_blank(), 
        text=element_text(size=19,  family="serif"), 
        axis.text.y = element_text(color="black"),
        legend.position = "top",
        legend.text=element_text(size=12)
        )

CodePudding user response：

You'll first need to calculate percentages using dplyr package:

library(dplyr)
newDoto <- newDoto %>% group_by(pid3lean) %>%
  count(Q29_1String) %>%
  mutate(perc = n/sum(n)) %>%
  select(-n)

With your existing code, you can just add the following line at the end of your code:

plot   
  geom_text(stat = 'count', aes(label = perc), position = position_fill(vjust = 0.5), size = 3, color = "white")