Using tapply in Shiny to find mean of a column - r

Question

I am running into trouble using the tapply function. I am pulling two vectors from the same data frame which was created from a reactive variable. The first I am calling from a user inputted selection, and the second is one that I have created to keep my code generalisable and to use in my sort function. My sample code is shown below using the r-bloggers example. The data is here.
https://redirect.viglink.com/?format=go&jsonp=vglnk_150821851345614&key=949efb41171ac6ec1bf7f206d57e90b8&libId=j8v6cnh201021u9s000DAhzunvtas&loc=https%3A%2F%2Fwww.r-bloggers.com%2Fbuilding-shiny-apps-an-interactive-tutorial%2F&v=1&out=http%3A%2F%2Fdeanattali.com%2Ffiles%2Fbcl-data.csv&ref=https%3A%2F%2Fduckduckgo.com%2F&title=Building%20Shiny%20apps%20%E2%80%93%20an%20interactive%20tutorial%20%7C%20R-bloggers&txt=here
The error it throws is that they are not the same length, even though their attribute and class print outs are exactly the same.
I know that this is not the best code in the world, but I just threw together a quick example.
library(shiny)
library(tidyverse)
bcl <- read.csv("bcl-data.csv", stringsAsFactors = FALSE)
ui <- fluidPage(titlePanel("Sampling Strategies"),
sidebarLayout(
sidebarPanel(
selectInput("XDATA","xdata",
choices = c(names(bcl))),
selectInput("YDATA","ydata",
choices = c(names(bcl)))
),
mainPanel(
tabsetPanel(
tabPanel("The table",tableOutput("mytable"))
))
))
server <- function(input, output, session) {
filtered <- reactive({
bcl <- bcl %>% mutate(ID = 1:nrow(bcl))
})
output$mytable <- renderTable({
dataset <- filtered() %>% mutate(sampled = "white")
sample.rows <- sample(dataset$ID, 5, replace = FALSE)
dataset$sampled[sample.rows] <- "black"
final <- tapply(dataset[input$XDATA], list(dataset$sampled),mean)[["black"]]
return(final)
})
}
shinyApp(ui = ui, server = server)
Cheers
Edit* Sorry my bad, forgot to change over the drop list codes. All I am interested is one generic xdata vector that can be selected from the loaded data set. I then sample it, and want to find the mean value from the sampled indices.

Answer 1

One of the problems is in the subsetting. the [ still returns a data.frame. So, we need [[. If we look at ?tapply
tapply(X, INDEX, FUN = NULL, ..., default = NA, simplify = TRUE)
where
X is an atomic object, typically a vector
ui <- fluidPage(titlePanel("Sampling Strategies"),
sidebarLayout(
sidebarPanel(
selectInput("XDATA","xdata",
choices = c(names(bcl)[5:7])),
selectInput("YDATA","ydata",
choices = c(names(bcl)))
),
mainPanel(
tabsetPanel(
tabPanel("The table",tableOutput("mytable"))
))
))
server <- function(input, output, session) {
filtered <- reactive({
bcl <- bcl %>% mutate(ID = row_number())
})
output$mytable <- renderTable({
dataset <- filtered() %>% mutate(sampled = "white")
sample.rows <- sample(dataset$ID, 20, replace = FALSE)
dataset$sampled[sample.rows] <- "black"
final <- tapply(dataset[[input$XDATA]], list(dataset$sampled),mean, na.rm = TRUE, simplify = TRUE)
return(final)
})
}
shinyApp(ui = ui, server = server)
-output