使用ggplot2在条形图中并排绘制两个变量

Question

I think I need to use melt function but I'm not sure how to do so? Sample data, code, and resulting graph below. Basically, the "cnt" column is made up of the "registered" plus the "casual" for each row. I want to display the total "registered" vs the total 'casual" per month, instead of overall total "cnt"

example data

#Bar Chart
bar <- ggplot(data=subset(bikesharedailydata, !is.na(mnth)), aes(x=mnth, y=cnt)) +
  geom_bar(stat="identity", position="dodge") +
  coord_flip() +
  labs(title="My Bar Chart", subtitle = "Total Renters per Month", caption = "Caption", x = "Month", y = "Total Renters") +
  mychartattributes

Answer 1

To "melt" your data use reshape2::melt :

library(ggplot2)
library(reshape2)

# Subset your data
d <- subset(bikesharedailydata, !is.na(mnth))
# Select columns that you will plot
d <- d[, c("mnth", "registered", "casual")]
# Melt according to month
d <- melt(d, "mnth")

# Set fill by variable (registered or casual)
ggplot(d, aes(mnth, value, fill = variable)) + 
    geom_bar(stat = "identity", position = "dodge") + 
    coord_flip() + 
    labs(title="My Bar Chart", subtitle = "Total Renters per Month", 
         caption = "Caption", 
         x = "Month", y = "Total Renters")

Answer 2

Use tidyr and dplyr:

set.seed(1000)
library(dplyr)
library(tidyr)
library(ggplot2)

bikesharedailydata <- data.frame(month = month.abb, registered = rpois(n = 12, lambda = 2), casual =rpois(12, lambda = 1))


bikesharedailydata %>% gather(key="type", value = "count", -month) %>%
  ggplot(aes(x=month, y=count, fill = type))+geom_bar(stat = "identity", position = "dodge")