ggplot：按月-年聚合多年数据，审美长度误差

Question

i've read every relevant aggregate() by month and lubridate question i could find but am still running into an error of aesthetic length. lots didn't work for me bc they grouped data by month but the dataframe only contained data from one year. i don't need the cumulative total of every January across time – i need it to be month- AND year-specific.

my sample data: (df is called "sales")

order_date_create     order_sum
2020-05-19            900
2020-08-29            500
2020-08-30            900
2021-02-01            200
2021-02-06            500

aggregating by month-year:

# aggregate by month (i used _moyr short for month year)
sales$bymonth <- aggregate(cbind(order_sum)~month(order_date_create),
                     data=sales,FUN=sum)
sales$order_moyr <- format(sales$order_date_create, '%m-%Y') # why does this get saved under values instead of data?

here's my ggplot:

# plot
ggplot(sales, aes(order_moyr, order_sum)) + 
  scale_x_date(limits = c(min, as.Date(now())), 
               breaks = "1 month", 
               labels = date_format("%m-%Y")) + 
  scale_y_continuous(labels = function(x) format(x, big.mark = "'", decimal.mark = ".", scientific = FALSE)) + 
  labs(x = "Date", y = "Sales Volume", title = "Sales by Month") +
  geom_bar(stat="identity")+ theme_economist(base_size = 10, base_family = "sans", horizontal = TRUE, dkpanel = FALSE) +  scale_colour_economist()

if i use x = order_date_create and y = order_sum it plots correctly, with month-year axis, but each bar is still daily sum . if i use x = order_moyr and y = bymonth , i get this error:

Error: Aesthetics must be either length 1 or the same as the data (48839): y

tangentially, if anyone knows how to use both scale::dollar AND format the thousands separator in the same scale_y_continous fcn it would be a great help. i've not found how to do both.

Answer 1

library(scales); library(lubridate); library(dplyr); 
library(ggthemes)
sales %>%
  count(order_moyr = floor_date(order_date_create, "month"), 
        wt = order_sum, name = "order_sum") %>%
                     
ggplot(aes(order_moyr, order_sum)) + 
  scale_x_date(breaks = "1 month",
               labels = date_format("%m-%Y")) +
  scale_y_continuous(labels = scales::dollar_format(big.mark = "'", 
                                             decimal.mark = ".")) + 
  labs(x = "Date", y = "Sales Volume", title = "Sales by Month") +
  geom_bar(stat="identity", width = 25)+ 
  theme_economist(base_size = 10, base_family = "sans", 
                  horizontal = TRUE, dkpanel = FALSE) +  
  scale_colour_economist()