shiny & ggplot: customizing date_breaks on x-axis based on selectInput choices

[Piranha]

Hello,
I am trying to create a shiny app for my colleagues to share some data on reported financial information of publicly traded companies. A small example of the data is reproduced below.

Data & Code:

library(shiny)
library(tidyverse)

data <- tibble(Qtr = as.Date(c("2018-02-01", "2018-05-01", "2018-08-01", "2018-11-01", "2019-02-01","2019-05-01","2019-08-01",
                     "2018-03-01", "2018-06-01", "2018-09-01", "2018-12-01", "2019-03-01", "2019-06-01", "2019-09-01")),
             Ticker = c("NKE", "NKE", "NKE", "NKE", "NKE", "NKE", "NKE", "UAA", "UAA", "UAA", "UAA", "UAA", "UAA", "UAA"),
             Revenue_YY = c(6.50, 12.80,  9.60,  9.50,  6.90,  4.00,  7.10,  
                            5.80,  7.60,  2.40,  1.50,  1.60,  1.45, -0.90))
data

# ui
ui <- fluidPage(
        selectInput("Ticker", "select Ticker", choices=c("NKE", "UAA"), selected="NKE"),
        mainPanel(
          plotOutput("Ticker_Chart")
          )
)

# server
server <- function(input, output) {
     
  output$Ticker_Chart <- renderPlot({
      chart <- data %>% 
        filter(Ticker==input$Ticker) %>% 
        ggplot(aes(x=Qtr)) +
        geom_line(aes(y=Revenue_YY), size=1) +
        geom_hline(yintercept=0) +
        theme_bw() +
        theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
        labs(y="Y/Y Change (%)", x="Quarter") +
        ggtitle(paste0("Y/Y Revenue Trend: ", input$Ticker)) +
        scale_x_date(date_breaks = "month", date_labels = "%b-%y")
      chart
     })
}

# Run the application 
shinyApp(ui = ui, server = server)

The problem: I am struggling with customizing the tick marks on the x-axis using the date_breaks argument inside scale_x_date.

At the moment, I am using monthly ticks marks:

        scale_x_date(date_breaks = "month", date_labels = "%b-%y")

But this is really not accurate, since the data here is in quarterly frequency. More importantly, the NKE data follows the Feb/May/Aug/Nov quarters, but the UAA data follows the Mar/Jun/Sep/Dec quarters.

I have tried using "3 months" as the date break, but that gives me weirdly out of sync axis ticks.

      scale_x_date(date_breaks = "3 months", date_labels = "%b-%y")

What I need looks like this:

I defined two new vectors with the appropriate tick marks:

ticks_aug_qtr <- seq.Date(as.Date("2018-02-01"), to=as.Date("2019-08-01"), by="quarter")
ticks_sep_qtr <- seq.Date(as.Date("2018-03-01"), to=as.Date("2019-09-01"), by="quarter")

...and then included the appropriate vector inside scale_x_date using "breaks" argument.

      scale_x_date(breaks = ticks_aug_qtr, date_labels = "%b-%y")

But I dont know how to dynamically change the breaks based on the selectInput Ticker choice.

Any help/guidance would be much appreciated.

Thanks!!

[raytong]

Hi @Piranha. You may use switch function to assign the breaks as follow. And you may report the issue of scale_x_date.

library(shiny)
library(tidyverse)

data <- tibble(Qtr = as.Date(c("2018-02-01", "2018-05-01", "2018-08-01", "2018-11-01", "2019-02-01","2019-05-01","2019-08-01",
                               "2018-03-01", "2018-06-01", "2018-09-01", "2018-12-01", "2019-03-01", "2019-06-01", "2019-09-01")),
               Ticker = c("NKE", "NKE", "NKE", "NKE", "NKE", "NKE", "NKE", "UAA", "UAA", "UAA", "UAA", "UAA", "UAA", "UAA"),
               Revenue_YY = c(6.50, 12.80,  9.60,  9.50,  6.90,  4.00,  7.10,  
                              5.80,  7.60,  2.40,  1.50,  1.60,  1.45, -0.90))
data

# ui
ui <- fluidPage(
  selectInput("Ticker", "select Ticker", choices=c("NKE", "UAA"), selected="NKE"),
  mainPanel(
    plotOutput("Ticker_Chart")
  )
)

# server
server <- function(input, output) {
  
  output$Ticker_Chart <- renderPlot({
    
    breaks <- switch(input$Ticker,
                     NKE = seq.Date(as.Date("2018-02-01"), to=as.Date("2019-08-01"), by="quarter"),
                     UAA = seq.Date(as.Date("2018-03-01"), to=as.Date("2019-09-01"), by="quarter"))
    
    chart <- data %>% 
      filter(Ticker==input$Ticker) %>% 
      ggplot(aes(x=Qtr)) +
      geom_line(aes(y=Revenue_YY), size=1) +
      geom_hline(yintercept=0) +
      theme_bw() +
      theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
      labs(y="Y/Y Change (%)", x="Quarter") +
      ggtitle(paste0("Y/Y Revenue Trend: ", input$Ticker)) +
      scale_x_date(breaks = breaks, date_labels = "%b-%y")
    chart
  })
}

# Run the application 
shinyApp(ui = ui, server = server)

[Piranha]

Thank you! This is an excellent solution.