R plot: Quarterly date not in sequential order-CodePudding

sorry for the lack of reproducibility example that I am about to display but I am sure this issue has happened quite a lot when building plots. I have two time series that I want to display (one as a barplot and the other as a line). I also want to use two different Y-axes scales. To do so I use the following code:

Color <- "#FF6347"
priceColor <- rgb(0.2, 0.6, 0.9, 1)
coeff <- 0.035

ggplot(data, aes(x=time, group =1))  
  geom_bar( aes(y=YseriesA), stat="identity", size=.1, fill=Color, color="black", alpha=.4)   
  geom_line( aes(y=YseriesB / coeff), size=1.2, color=priceColor)  
  scale_y_continuous(
    # Features of the first axis
    name = "B",
    # Add a second axis and specify its features
    sec.axis = sec_axis(~.*coeff, name="A")
  )   
  theme_ipsum()  
  theme(
    axis.title.y = element_text(color = Color, size=13),
    axis.title.y.right = element_text(color = priceColor, size=13)
  )  
  theme(axis.text.x = element_text(angle = 90))  
  ggtitle("Title")

Most importantly, the variable time I have produced with the following code:

data$time <- seq(from = zoo::as.yearqtr("2000-01-01"), to = zoo::as.yearqtr("2019-01-01"), by = 1/4) %>%
  format("Q%q-%Y")

My question, the issue is the following. Why on earth is time not ordered from earlier to later in the X-axes but displays this weird order: Q1-2000, Q1-2001, Q1-2002..... Any help much much appreciated and once again my most sincere apologies for the lack of reproducibility.

CodePudding user response：

It is fully linked to the order of the variable. Here the order is alphabetic, this is why 'Q1-2018' is prior to 'Q2-2000'. You could either rename your modalities in this format: 2018-Q1 or use the factor function with level argument to specify the right order you want.

The following code is an example, you can use all modalities from your variable time.

### Simulation of data and checking order
data <- data.frame(time=c("Q1-2018", "Q2-2010", "Q3-2000"))
factor(data$time)
# -> Levels: Q1-2018 Q2-2010 Q3-2000

### Reordering levels and checking order
data$time <- factor(data$time, levels=c("Q3-2000", "Q2-2010", "Q1-2018"))
factor(data$time)
# -> Levels: Q3-2000 Q2-2010 Q1-2018

Another solution is to use the as.yearqtr function on your dates.

yourDates <- c("2018-02-02", "2010-05-05", "2000-10-10")
data <- as.yearqtr(yourDates, format = "%Y-%m-%d")
factor(data)
# -> Levels: 2000 Q4 2010 Q2 2018 Q1

CodePudding user response：

You could use zoo's scale_x_yearqtr (now that you're using that class!):

library(ggplot2)
library(zoo)

df |>
  ggplot(aes(x = date, y = y))   
  geom_line()  
  theme(axis.text.x = element_text(angle = 90))  
  zoo::scale_x_yearqtr(n = 40,
                         format = "Q%q-%Y")

Output

Data (please provide some another time - it doesn't need to be perfect):

library(tibble)
library(zoo)

set.seed(1)
df <- tibble(y = rnorm(77),
             date = seq(from = zoo::as.yearqtr("2000-01-01"),
                        to   = zoo::as.yearqtr("2019-01-01"),
                        by   = 1/4)
             )