Use R shiny to simulate CLT by using different distribution - r

Question

library(shiny)
library(ggplot2)
ui <- shinyUI(fluidPage(
titlePanel("Central Limit Theorem Simulation"),
sidebarLayout(
sidebarPanel(
numericInput("sample.size", "Size of each random sample",
value = 30, min = 1, max = 100, step = 1),
sliderInput("simulation", "THe number of simulation",
value = 100, min = 100, max = 1000, step = 1),
selectInput("sample.dist", "Population Distribution where each sample is from",
choices = c("Binomial","Poisson", "Normal", "Uniform") ),
numericInput("bins", "Number of bins in the histogram",
value = 20, min = 1, max = 50, step = 1),
submitButton(text = "Submit")
),
mainPanel(
h3('Illustrating outputs'),
h4('mean of random sample mean'),
textOutput(outputId = "output_mean" ),
h4('variance of random sample mean'),
textOutput(outputId = "output_var"),
h4("Table"),
tableOutput(outputId = "output_table"),
h4('histogram of random normal sample'),
plotOutput(outputId = "output_hist")
)
)
))
server <- shinyServer(function(input, output) {
# Return the random sample
rsample <- ifelse(
input$sample.dist == "Binomial", rbinom(input$sample.size * input$simulation, 1, 0.5),
ifelse(input$sample.dist == "Poisson", rpois(input$sample.size * input$simulation, 1),
ifelse(input$sample.dist == "Normal", rnorm(input$sample.size * input$simulation),
runif(input$sample.size * input$simulation) ) ) )
# Return the random sample matrix
rsample.matrix <- matrix(rsample, nrow = input$simulation)
# output mean of sample mean
output$output_mean <- renderText({
sample.mean <- rowMeans(rsample.matrix)
mean(sample.mean)
})
# output variance of sample mean
output$output_var <- renderText({
sample.mean <- rowMeans(rsample.matrix)
var(sample.mean)
})
# output histogram of sample mean
output$output_hist <- renderPlot({
sample.mean <- rowMeans(rsample.matrix)
ggplot(data.frame(sample.mean), aes(x = sample.mean)) +
geom_histogram(bins = input$bins)
})
})
shinyApp(ui = ui, server = server)
(1) The above codes are to create a shiny application to simulate random sample and verify Central Limit Theorem. However， since I just learned shiny, I have no idea where it is wrong.
(2) Also, if I want to change the parameter of a specific distribution based on the distribution selected by the user, what should I do?
The error returned from Rstudio is as follows:
Warning: Error in .getReactiveEnvironment()$currentContext: Operation not allowed without an active reactive context. (You tried to do something that can only be done from inside a reactive expression or observer.)
Stack trace (innermost first):
49: .getReactiveEnvironment()$currentContext
48: .subset2(x, "impl")$get
47: $.reactivevalues
46: $ [#4]
45: server [#4]
4: <Anonymous>
3: do.call
2: print.shiny.appobj
1: <Promise>
Error in .getReactiveEnvironment()$currentContext() :
Operation not allowed without an active reactive context. (You tried to do something that can only be done from inside a reactive expression or observer.)

Answer 1

The error says that code should be within a reactive or an observe statement. Have a look how I wrapped and used the variables
library(shiny)
library(ggplot2)
ui <- shinyUI(fluidPage(
titlePanel("Central Limit Theorem Simulation"),
sidebarLayout(
sidebarPanel(
numericInput("sample.size", "Size of each random sample",
value = 30, min = 1, max = 100, step = 1),
sliderInput("simulation", "THe number of simulation",
value = 100, min = 100, max = 1000, step = 1),
selectInput("sample.dist", "Population Distribution where each sample is from",
choices = c("Binomial","Poisson", "Normal", "Uniform") ),
numericInput("bins", "Number of bins in the histogram",
value = 20, min = 1, max = 50, step = 1),
submitButton(text = "Submit")
),
mainPanel(
h3('Illustrating outputs'),
h4('mean of random sample mean'),
textOutput(outputId = "output_mean" ),
h4('variance of random sample mean'),
textOutput(outputId = "output_var"),
h4("Table"),
tableOutput(outputId = "output_table"),
h4('histogram of random normal sample'),
plotOutput(outputId = "output_hist")
)
)
))
server <- shinyServer(function(input, output) {
# Return the random sample
rsample <- reactive({
ifelse(
input$sample.dist == "Binomial", rbinom(input$sample.size * input$simulation, 1, 0.5),
ifelse(input$sample.dist == "Poisson", rpois(input$sample.size * input$simulation, 1),
ifelse(input$sample.dist == "Normal", rnorm(input$sample.size * input$simulation),
runif(input$sample.size * input$simulation) ) ) )
})
# Return the random sample matrix
rsamplematrix <- reactive({
matrix(rsample(), nrow = input$simulation)
})
# output mean of sample mean
output$output_mean <- renderText({
sample.mean <- rowMeans(rsamplematrix())
mean(sample.mean)
})
# output variance of sample mean
output$output_var <- renderText({
sample.mean <- rowMeans(rsamplematrix())
var(sample.mean)
})
# output histogram of sample mean
output$output_hist <- renderPlot({
sample.mean <- rowMeans(rsamplematrix())
ggplot(data.frame(sample.mean), aes(x = sample.mean)) +
geom_histogram(bins = input$bins)
})
})
shinyApp(ui = ui, server = server)