Plots Don't Get Updated From My Date Range in RShiny

#1

I am trying to make a shiny app that first imports data in the first tab and plots a ggplot based on the imported data and the user-chosen data range in the second tab. The below code doesn't give me an error, but my plot is not getting updated as I change my date range.

library(shiny)
library(ggplot2)
library(dplyr)

df <- data.frame("date" = c("2020-01-01", "2020-01-01", "2020-01-02", "2020-01-02"), "id" = c("A", "A", "B", "B"), "number" = c(1, 3, 2, 5))
df

write.csv(df, "df.csv", row.names = F)

if (interactive()) {
ui <- fluidPage(
      fileInput("file", "Upload Your File:"),
      uiOutput("daterange"),
      plotOutput("plot")
    )       

server <- function(input, output, session) {
  df <- reactive({
      req(input$file)
      inFile <- input$file
      read.csv(inFile$datapath, stringsAsFactors = F)
    })
  
 output$daterange <- renderUI({
    dateRangeInput("daterange", "Select the date range:",
                   start = as.Date(min(df()$date), format = "%m/%d/%Y"),
                   end = as.Date(max(df()$date), format = "%m/%d/%Y"), 
                   min = as.Date(min(df()$date), format = "%m/%d/%Y"),
                   max = as.Date(max(df()$date), format = "%m/%d/%Y")
                   )
  })


output$plot <- renderPlot({
 dfplot <- subset(df(), date >= input$daterange[1] & date <= input$daterange[2])

 g <- ggplot(df(), aes(id, number)) +
      geom_boxplot() +
      theme_classic()
  print(g)
  })
}

shinyApp(ui=ui, server = server)
}
#2

You are plotting using the original unfiltered data i.e. df() not the filtered data dfplot

#3

I changed my code to:

g <- ggplot(dfplot, aes(id, number))

Now the plots are still not showing at all.

#5

add some print statements in your function to show you what is happening

print( input$daterange[1] )
print( input$daterange[2] )
dfplot <- subset(df(), date >= input$daterange[1] & date <= input$daterange[2]) 
print(head(dfplot,n=10))
g <- ggplot(dfplot, aes(id, number))
#6

It works for me

library(shiny)
library(ggplot2)

df <- data.frame("date" = c("2020-01-01", "2020-01-01", "2020-01-02", "2020-01-02"), "id" = c("A", "A", "B", "B"), "number" = c(1, 3, 2, 5))

write.csv(df, "df.csv", row.names = F)

if (interactive()) {
  ui <- fluidPage(
    fileInput("file", "Upload Your File:"),
    uiOutput("daterange"),
    plotOutput("plot")
  )       
  
  server <- function(input, output, session) {
    df <- reactive({
      req(input$file)
      inFile <- input$file
      read.csv(inFile$datapath, stringsAsFactors = F)
    })
    
    output$daterange <- renderUI({
      dateRangeInput("daterange", "Select the date range:",
                     start = as.Date(min(df()$date), format = "%m/%d/%Y"),
                     end = as.Date(max(df()$date), format = "%m/%d/%Y"), 
                     min = as.Date(min(df()$date), format = "%m/%d/%Y"),
                     max = as.Date(max(df()$date), format = "%m/%d/%Y")
      )
    })
    
    
    output$plot <- renderPlot({
      dfplot <- subset(df(), date >= input$daterange[1] & date <= input$daterange[2])
      
      ggplot(dfplot, aes(id, number)) +
        geom_boxplot() +
        theme_classic()
    })
  }
  
  shinyApp(ui=ui, server = server)
}

image
image1864×1246 88.5 KB

#7

Well, it doesn't work for me, but I actually figured out a way by removing format = %m/%d/%Y i.e.

    dates <- as.Date(df()$date)
    dateRangeInput("daterange", "Select the date range:",
                   start = min(dates), end = max(dates),
                   min = min(dates), max = max(dates)
                   )

However, if I change for example a date from 1/2/2020 to 1/3/2020 and save the csv file, now the date range and plots don't show up again. This is making me really confusing. So, this shiny app only works with the data frame that was made within R, not the csv files that was made outside of R.

#8

We would need more realistic sample data to help you with that

#9

For example, if I make a data frame inside R and then write this as a csv file, then everything works fine:

df <- data.frame("date" = c("2020-01-01", "2020-01-01", "2020-01-02", "2020-01-02"), "id" = c("A", "A", "B", "B"), "number" = c(1, 3, 2, 5))

However, I I try to change the date in the last row (1/2/2020 -> 1/3/2020), then the date range and plots don't show anything.
Capture

#10

how are you reading the csv back in? unless you are using something sophisticated like the readr package, an apparent date field will probably be read in as a character string.

months(c(lubridate::ymd("2020-01-01"),"2020-01-01"))

in the above I show that the base months function (to tell you the longform word of the month of the date) fails when passing a raw string , but succeeds when a date conversion has happened ( lubridate makes date conversion easy).

#11

I am assuming you are editing the csv in excel. Could you try opening the csv file with notepad (or a text editor)? Save the changes and try re-running.
Not sure about why you are facing this issue, but csv does behave weird sometimes when opening with excel.