I have some problem with a reactive table in Shiny R - r

Question

I'm trying to create a dashboard using shiny in R, but I'm facing some little problems
I have:
db is my data.frame with:
db$domain:chr,
db$date:chr,
db$value:num.
So I've created:
db_4 <- reactive({ subset(db,db$domain %in% input$domain &
db$date<=input$daterange[2] & db$date>=input$daterange[1]})
the inputs are:
input$domain: selectinput with multiple choices,
input$date: daterangeinput.
I'm trying to create a table that gives me the sum of the db$value, aggregated by db$date. I've tried something like:
output$table2 <- rendertable ({aggregate(db_4()["value"], by=list(db_4()["date"]), sum) })
but I get always an empty table.
Can anybody help me in solving this little issue?
Thx a lot

Answer 1

I would highly recommend you to read this article about debugging.
In Shiny you can use the browser() function within both reactive and render functions. It should help you locate the problem (i.e.: data has the expected structure)
It seems the problem is with the aggregate function: db_4()["date"] returns a data.frame, where you need a vector.
Solution:
library(shiny)
db <- data.frame(
domain = letters[1:3],
date = seq(
from = as.Date("2019-01-01"),
to = as.Date("2019-06-01"),
by = "1 months"
),
value = runif(12)
)
ui <- fluidPage(
sidebarLayout(
sidebarPanel(
selectInput("domain", "Domain", choices = unique(db$domain)),
dateRangeInput("daterange", "Date",
min = min(db$date), max = max(db$date),
start = min(db$date), end = max(db$date))
),
mainPanel(
tableOutput("table2")
)
)
)
server <- function(input, output, session) {
db_4 <- reactive( {
subset(db,
db$domain %in% input$domain &
db$date<=input$daterange[2] &
db$date>=input$daterange[1]
)
})
output$table2 <- renderTable( {
req(db_4()) # Don't render table when db_4() is NULL
# Uncomment next line to check if everything goes as expected
#browser()
aggregate(
data.frame(value = db_4()$value),
by=list(date = as.factor(db_4()$date)),
sum
)
})
}
shinyApp(ui, server)
Also I would highly recommend sharing the code of your minimal example including some dummy data, so that it can be copy-pasted in an instant. It would increase the chances of someone answering.