Multiplot on the same graph with checkboxGroupInput in a Shiny App - r

Question

I want to plot on the same graph one or more plots, depending on the boxes checked in my "checkboxGroupInput".
After checking a few topics on SO, I have found that a easy-to-use solution would be to add "add=TRUE" in the second, third...plots.
Here is a simplified copy of my code :
Server.R
function(input, output) {
dataInput <- reactive({
#I use getMyPlotValues with two parameters.
#idFcast is the one which doesn't work.
m_PARAM$idSite <- input$country
m_PARAM$idFcast <- input$model[i]
getMyPlotValues(m_PARAM)
})
output$plot1 <- renderPlot({
## We plot all the models on the same graph.
if(length(input$model) > 0)
{
firstplot<- TRUE
#To put add = TRUE for the second, third... plots
for(i in 1:length(input$model))
{
valueshere <- dataInput()
#We pick up the vectors (ts,obs) with the function
#And plot them.
if(firstplot)
{
plot(valueshere$ts,valueshere$obs)
}
else
{
plot(valueshere$ts,valueshere$obs,add=TRUE)
}
firstplot <- FALSE
}
}
})
}
And now the UI.R simplified :
shinyUI(fluidPage(
sidebarLayout(
sidebarPanel(
fluidRow(
column((4),
selectInput("country", label = h3("Pays"),
choices = list("First" = 2,"Second" = 1), selected = 1)
)
fluidRow(
#checkboxGroupInput
column(6,
checkboxGroupInput("model",
choices = list("First" = 1,"Second"=2,"Third"=3),
selected = 1)
)
)
),
mainPanel
(
plotOutput("plot1")
)
)
)
)
I found on this topic (How can I pass data between functions in a Shiny app) that I could try to pass the variable "i" by writing valueshere <- dataInput()$i... Because I have the feeling this is where the error comes from. I try do display input$model and it works, input$model[i] doesn't.
There must be a problem in my code, as I am a beginner in Shiny, but I couldn't fix, even with all the topics about Shiny on SO.
Thanks for your help and have a nice day !

Answer 1

Seems that add argument isn't work. I finished with the solution that consists in creating plots for all your models and only displays them when it's value is selected. So here is:
I have tried to make your example reproduceble so you will observe some changes
UPDATE: Supposing your SQL statement will give you a vector with your model names (models_list) you could generate Shiny objects using lapply (like here)
library(shiny)
# Simulate data
set.seed(189)
df <- data.frame("obs" = 1:30, "ts" = (rnorm(30)+100)
, "country" = rep(c("First", "Second"), 15)
, "models" = rep(c("First", "Second", "Third"), 10),
stringsAsFactors = F)
# Suppose your SQL query returns this kind of output
models_list <- c("First", "Second", "Third")
models_num <- length(models_list)
# Run App
shiny::runApp(list(
ui = fluidPage(
sidebarLayout(
sidebarPanel(
selectInput("country", label = "Countries",
choices = c("First", "Second"), selected = "First"),
checkboxGroupInput("model", label = "Models",
choices = c("First", "Second", "Third"),
selected = "First")
),
mainPanel(
lapply(1:models_num, function(i) {
plotOutput(paste0('plot', i))
})
)
)
),
server = function(input, output) {
dataInput <- reactive({
df_out <- df[(df$country == input$country), ]
return(df_out)
})
lapply(1:models_num, function(i) {
output[[paste0('plot', i)]] <- renderPlot({
if(any(input$model %in% models_list[i])){
valueshere <- dataInput()[df$models == models_list[i],]
plot(valueshere$ts,valueshere$obs)
}
})
})
}
))