Error When Running Shiny App: the condition has length > 1

[Paul07]

This is the first time I write on here, so please excuse me if I do not use the proper format. I was trying to create an interactive forecasting application using Shiny.
Whenever I try to run the app, I get the following error: the condition has length > 1.

I tried to include the dataset I am using for this app but it seems that I am not allowed to attach it.
What could be causing this error?

Here is my ui.R code:
#install.packages("plyr")
#install.packages("rintrojs")
#install.packages("DT")
#install.packages("forecast")

#loading required packages
library(readxl)
library(plotly)
library(shinycssloaders)
library(shinythemes)
library(shinydashboard)
library(plyr)
library(rintrojs)
library(DT)
library(forecast)

#defining working directory
setwd("C:/Users/paulbernal/Desktop/Project 04")

#importing dataset for interactive visualization
df1 <- read.csv("dataset01.csv")

shinyUI(fluidPage(theme = shinytheme("cerulean"),

tabPanel("Forecasting",
fluidRow(
sidebarLayout(
sidebarPanel(
tabsetPanel(id = "tabs",
tabPanel("Metric Forecast", value = "Tab1",
selectInput("Metric", "Metric:",
choices = colnames(df1[2:49]),
hr(),
helpText("choose metric above to forecast")))),
mainPanel(
plotlyOutput("Plot1")
)

))))))

This is my server.R code:
#installing required R packages for shiny app
#install.packages("readxl")
#install.packages("plotly")
#install.packages("shiny")
#install.packages("shinycssloaders")
#install.packages("shinythemes")
#install.packages("shinydashboard")
#install.packages("plyr")
#install.packages("rintrojs")
#install.packages("DT")
#install.packages("forecast")

#loading required packages
library(readxl)
library(plotly)
library(shinycssloaders)
library(shinythemes)
library(shinydashboard)
library(plyr)
library(rintrojs)
library(DT)
library(forecast)

#defining working directory
setwd("C:/Users/paulbernal/Desktop/Project 04")

#importing dataset for interactive visualization
df1 <- read.csv("dataset01.csv")

#create forecast dates
fcast_dates <- (2021:2027)

server <- function(input, output) {

  observeEvent(input$Metric, {
    
      output$plot1 <- renderPlotly({
        plot_ly(x = df1$Year, y = df1[,colnames(df1) == input$Metric], name = 'Historical',
                type = 'scatter', mode = 'lines', line = list(color = 'rgb(205, 12, 24', width = 4))%>%
          add_trace(x = fcast_dates, y = forecast(auto.arima(df1[,colnames(df1) == input$Metric]), h = 7)$mean,
                    name = 'Forecast', line = list(color = 'rgb(22, 96, 167)', diwth = 4, dash = 'dot'))%>%
          layout(title = "Non-Seasonal Forecast)",
                 xaxis = list(title = "Years"),
                 yaxis = list(title = "Selected Metric"))
                 
      })})}

[scottyd22]

To better help troubleshoot, can you please share what you get when executing the following? This will provide a look at the data structure you're working with.

dplyr::glimpse(df1)

[Paul07]

Hi Scott,

This is what I get when I run dplyr::glimpse(df1)

[scottyd22]

I noticed a couple parentheses out of place in the UI section, so I reformatted below and added comments on closing parentheses. After making these updates, the error was resolved.

shinyUI(
  fluidPage(
    theme = shinytheme("cerulean"),
    tabPanel("Forecasting",
             fluidRow(
               sidebarLayout(
                 sidebarPanel(
                   tabsetPanel(id = "tabs",
                               tabPanel("Metric Forecast", 
                                        value = "Tab1",
                                        selectInput("Metric", 
                                                    "Metric:",
                                                    choices = colnames(df1[2:4])
                                                    ), 
                                        hr(),
                                        helpText("choose metric above to forecast")
                                        ) # closes tabPanel
                               ) # closes tabsetPanel
                   ), # closes sidebarPanel
                 mainPanel(
                   plotlyOutput("Plot1")
                   )
                 ) # closes sidebarLayout
               ) # closes fluidRow
             ) # closes tabPanel
    ) # closes fluidPage
  ) # closes shinyUI

[Paul07]

Hi Scott,

Thank you so much for your valuable feedback. For some reason, once the page pops up the Shiny application stops, why could this be happening?

[scottyd22]

Does an error show up in the console? If so, can you share it?

[Paul07]

Hi Scott,

I can now see the dashboard, however, the forecasts do not appear. Which is odd.

This is my ui.R code:
#loading required packages
library(readxl)
library(plotly)
library(shinycssloaders)
library(shinythemes)
library(shinydashboard)
library(plyr)
library(rintrojs)
library(DT)
library(forecast)

#defining working directory
setwd("C:/Users/paulbernal/Desktop/Project 04")

#importing dataset for interactive visualization
df1 <- read.csv("dataset01.csv")

shinyUI(
fluidPage(
theme = shinytheme("cerulean"),
tabPanel("Forecasting",
fluidRow(
sidebarLayout(
sidebarPanel(
tabsetPanel(id = "tabs",
tabPanel("Metric Forecast",
value = "Tab1",
selectInput("Metric",
"Metric:",
choices = colnames(df1[2:49]),
),
hr(),
helpText("choose metric above to forecast")
) # closes tabPanel
) # closes tabsetPanel
), # closes sidebarPanel
mainPanel(
plotlyOutput("Plot1")
)
) # closes sidebarLayout
) # closes fluidRow
) # closes tabPanel
) # closes fluidPage
) # closes shinyUI

And this is my server.R code:
#loading required packages
library(readxl)
library(plotly)
library(shinycssloaders)
library(shinythemes)
library(shinydashboard)
library(plyr)
library(rintrojs)
library(DT)
library(forecast)

#defining working directory
setwd("C:/Users/paulbernal/Desktop/Project 04")

#importing dataset for interactive visualization
df1 <- read.csv("dataset01.csv")

#create forecast dates
fcast_dates <- (2021:2027)

server <- function(input, output) {

  observeEvent(input$Metric, {
    
      output$plot1 <- renderPlotly({
        plot_ly(x = df1$Year, y = df1[,colnames(df1) == input$Metric], name = 'Historical',
                type = 'scatter', mode = 'lines', line = list(color = 'rgb(205, 12, 24', width = 4))%>%
          add_trace(x = fcast_dates, y = forecast(auto.arima(df1[,colnames(df1) == input$Metric]), h = 7)$mean,
                    name = 'Forecast', line = list(color = 'rgb(22, 96, 167)', diwth = 4, dash = 'dot'))%>%
          layout(title = "Non-Seasonal Forecast)",
                 xaxis = list(title = "Years"),
                 yaxis = list(title = "Selected Metric"))
                 
      })})}

You can see the selection menu, but no forecast is shown.

Best,
Paul

[scottyd22]

In the UI section, change Plot1 to lowercase to match the output in the server.

plotlyOutput("plot1")

Also, since input$Metric is within your plotly code, the plot will update each time a new metric is selected. Therefore, you do not need to wrap the output in an observeEvent().