r - How do I prevent x labels from overlapping my bars in a barplot?

Question

Here is the code:

barplot(colMeans(sample_data, na.rm = TRUE),
    las = 1,
    main = "Main Title",
    xlab = "Variable",
    ylab = "How Characteristic",
    col = rainbow(20),
    cex.names = 0.9,
    horiz = FALSE)

A sample data set is available here:

https://github.com/akaEmma/public_data/blob/master/sample_data.csv Or you can type some of it in yourself. These are the variable names:

Love of Chocolate,Asian Knowledge,Stable Cleanliness,Love of People,Attention,Ethics,Aggression,Swimming,Style Points,Felinity

And here are some of the data that go with the names:

8.67    9   6.25    7.33    6.33    5       6.67    5   5.25
8       3   6       6.67    8       7       7.67    4.5 5.25
7.33    7.5 5.75    8.67    8.67    8       5.33    2.5 3
8       6.5 6       6.33    8.33    5.33    5.67    6   6.5
6       5.5 5.25    5.33    5       4.67    4       4   3.5
7.67    7   6       4.67    7.33    5.67    7.67    5   3.75
8.67    8   7.5     5.67    7.33    5       8.33    7   7.75

If I use the code above I get the following (ignore the periods; they aren't important):

If I create a larger plot (like fill my screen with it) I get this:

(ignore the missing label; I accidentally left it off and it's supposed to be "Felinity," whatever that is)

This sort of bar chart is for a PowerPoint on a huge screen, so I can go very small with the labels.

Here is what I want: I want clean pretty labels, one per bar, and since this is a wish list, I want the labels to adjust their own sizes so that they are small enough to fit one per bar, and I want them to be at the right vertical point so that they do not overlap with the bars. Any ideas?

Go crazy. I want beautiful bar charts and I have to make a lot of them, so twiddling for each one is simply not an option. This has to work every time with data files of this type regardless of the length of the variable names.

Thanks!

标签： rpowerpoint

Answer 1

请注意theme_set和theme_tufte是ggplot2特定功能。

使用ggplot2你可以做这样的事情

df <- read.csv("https://raw.githubusercontent.com/akaEmma/public_data/master/sample_data.csv")

library(tidyverse)
library(ggthemes)
df %>%
    gather(key, value) %>%
    group_by(key) %>%
    summarise(mean.value = mean(value, na.rm = T)) %>%
    mutate(key = factor(key, levels = key[rev(order(mean.value))])) %>%
    ggplot(aes(key, mean.value, fill = as.numeric(key))) +
    geom_col() +
    theme_tufte() +
    scale_fill_gradientn(colours = rainbow(5), guide = F) +
    theme(axis.text.x = element_text(size = 6)) + 
    labs(x = "", y = "How characteristic")