layout detail

mse=function(y,yy){
  return(mean((y-yy)^2))
}

update.mse = function(old.mse, old.n, y.new, yy.new){
  return((old.n/(old.n + 1))*old.mse+(1/(old.n + 1))*(y.new-yy.new)^2)
}

current.mse=mse(y,fitted(reg1))
current.n=12
current.sales=y
current.periods=seq(1,12)
mytol = 0.1
ui = fluidPage(
  titlePanel("Battery Sales"),
  hr(),
  hr(),
  sliderInput("MonthSales",label="Select Monthly Sales",min=0,max=100,value=20,step=1),
  actionButton(inputId = "mybutton1",label="Update"),
  actionButton(inputId = "mybutton2",label="Predict"),
  hr(),
  verbatimTextOutput("myresults1")
)
server = function(input, output) {
  observeEvent(input$mybutton1,{
    output$myresults2 = renderText({""})
    y.hat=coef(reg1)[1]+coef(reg1)[2]*(current.n + 1)
    new.mse=update.mse(current.mse,current.n,input$MonthSales,y.hat)
    current.n <<- current.n + 1
    current.sales <<- c(current.sales,input$MonthSales)
    current.periods <<- seq(1,current.n)
    if (new.mse/current.mse > mytol + 1){
      output$myresults1 = renderText({
        paste0("Input #", current.n, " MSE ratio exceeds tolerance: We need to update the model!")
      })
      reg1 <<- lm(current.sales~current.periods)
      current.mse <<- mse(current.sales,fitted(reg1))
    }
    else{
      output$myresults1 = renderText({
        paste0("Input #", current.n, " MSE ratio is within tolerance: We keep the model!")
      })
      current.mse <<- new.mse
    }
  })
}
shinyApp(ui, server)

# 
y = BatterySales$`Battery Sales`
x = seq(1,12)
reg5 = lm(y~x)

current.mse=mse(y,fitted(reg5))
current.n=12
current.sales=y
current.periods=seq(1,12)
LB = -qnorm(0.975)*sqrt(current.mse)
UB = qnorm(0.975)*sqrt(current.mse)
ui = fluidPage(
  titlePanel("Battery Sales"),
  hr(),
  hr(),
  sliderInput("MonthSales",label="Select Monthly Sales",min=0,max=100,value=20,step=1),
  actionButton(inputId = "mybutton",label="Update"),
  hr(),
  verbatimTextOutput("myresults1")
)
server = function(input, output) {
  observeEvent(input$mybutton,{
    y.hat=coef(reg5)[1]+coef(reg5)[2]*(current.n + 1)
    current.n <<- current.n + 1
    current.sales <<- c(current.sales,input$MonthSales)
    current.periods <<- seq(1,current.n)
    e = input$MonthSales - y.hat
    if ((e>UB)||(e<LB)){
      output$myresults1 = renderText({
        paste0("Input #", current.n, " Error outside tolerance interval: We need to update the model!")
      })
      reg5 <<- lm(current.sales~current.periods)
      current.mse <<- mse(current.sales,fitted(reg5))
      LB <<- -qnorm(0.975)*sqrt(current.mse)
      UB <<- qnorm(0.975)*sqrt(current.mse)
    }
    else{
      output$myresults1 = renderText({
        paste0("Input #", current.n, " MSE ratio is within tolerance: We keep the model!")
      })
    }
  })
}
shinyApp(ui, server)

#web scraping
install.packages("XML")
library(XML)
uconn.url="https://uconn.edu/"
uconn.source=readLines(uconn.url,encoding="UTF-8")
View(uconn.source)
uconn.parsed=htmlParse(uconn.source,encoding="UTF-8")
getHTMLLinks(uconn.source)
getHTMLExternalFiles(uconn.source)
x=readHTMLList(uconn.source)
x[19]
y=unlist(x[19])
y[1]


myurl="https://flightaware.com/live/airport/KBDL"
mysource=readLines(myurl,encoding = "UTF-8")
mytables=readHTMLTable(mysource)
arrivals=mytables[2]$`NULL`
class(arrivals)
View(arrivals)
View(mytables[2]$`NULL`)
arrivals = arrivals[c(-1,-2),-5]
View(arrivals)

Extract.Table = function(myurl){
  mysource=readLines(myurl,encoding = "UTF-8")
  mytables=readHTMLTable(mysource)
  arrivals=mytables[2]$`NULL`
  arrivals=arrivals[3:13,-5]
  return(arrivals)
}

View(Extract.Table("https://flightaware.com/live/airport/KJFK"))

ui = fluidPage(
  tableOutput("mytable")
)
server = function(input, output, session) {
  autoInvalidate <- reactiveTimer(30000)
  output$mytable=renderTable({
    autoInvalidate()
    Extract.Table("https://flightaware.com/live/airport/KJFK")
  })
}
shinyApp(ui, server)

library(openxlsx)
ui = fluidPage(
  radioButtons("airport","Select Airport", c("BDL"="bdl","JFK"="jfk","LAX"="lax")),
  plotOutput("myplot"),
  tableOutput("mytable")
)
server = function(input, output, session) {
  autoInvalidate <- reactiveTimer(1000)
  arrivals=reactive({
    airportname = switch(input$airport,
                         bdl = "BDL",
                         jfk = "JFK",
                         lax = "LAX"
                         
    )
    autoInvalidate()
    Extract.Table(paste0("https://flightaware.com/live/airport/K",airportname))
  })
  output$mytable=renderTable({
    autoInvalidate()
    arrivals()
  })
  output$myplot=renderPlot({
    df = read.xlsx("/Users/YH/Desktop/data.xlsx")
    df = rbind(df,arrivals())
    write.xlsx(arrivals(),"/Users/YH/Desktop/data.xlsx")
    autoInvalidate()
    barplot(prop.table(table(arrivals()$Type)))
  })
}
shinyApp(ui, server)




mse=function(y,yy){
  return(mean((y-yy)^2))
}

update.mse = function(old.mse, old.n, y.new, yy.new){
  return((old.n/(old.n + 1))*old.mse+(1/(old.n + 1))*(y.new-yy.new)^2)
}

current.mse=mse(y,fitted(reg1))
current.n=12
current.sales=y
current.periods=seq(1,12)
mytol = 0.1
ui = fluidPage(
  titlePanel("Battery Sales"),
  hr(),
  hr(),
  sliderInput("MonthSales",label="Select Monthly Sales",min=0,max=100,value=20,step=1),
  actionButton(inputId = "mybutton1",label="Update"),
  actionButton(inputId = "mybutton2",label="Predict"),
  hr(),
  verbatimTextOutput("myresults1")
)
server = function(input, output) {
  observeEvent(input$mybutton1,{
    output$myresults2 = renderText({""})
    y.hat=coef(reg1)[1]+coef(reg1)[2]*(current.n + 1)
    new.mse=update.mse(current.mse,current.n,input$MonthSales,y.hat)
    current.n <<- current.n + 1
    current.sales <<- c(current.sales,input$MonthSales)
    current.periods <<- seq(1,current.n)
    if (new.mse/current.mse > mytol + 1){
      output$myresults1 = renderText({
        paste0("Input #", current.n, " MSE ratio exceeds tolerance: We need to update the model!")
      })
      reg1 <<- lm(current.sales~current.periods)
      current.mse <<- mse(current.sales,fitted(reg1))
    }
    else{
      output$myresults1 = renderText({
        paste0("Input #", current.n, " MSE ratio is within tolerance: We keep the model!")
      })
      current.mse <<- new.mse
    }
  })
}
shinyApp(ui, server)

# 
y = BatterySales$`Battery Sales`
x = seq(1,12)
reg5 = lm(y~x)

current.mse=mse(y,fitted(reg5))
current.n=12
current.sales=y
current.periods=seq(1,12)
LB = -qnorm(0.975)*sqrt(current.mse)
UB = qnorm(0.975)*sqrt(current.mse)
ui = fluidPage(
  titlePanel("Battery Sales"),
  hr(),
  hr(),
  sliderInput("MonthSales",label="Select Monthly Sales",min=0,max=100,value=20,step=1),
  actionButton(inputId = "mybutton",label="Update"),
  hr(),
  verbatimTextOutput("myresults1")
)
server = function(input, output) {
  observeEvent(input$mybutton,{
    y.hat=coef(reg5)[1]+coef(reg5)[2]*(current.n + 1)
    current.n <<- current.n + 1
    current.sales <<- c(current.sales,input$MonthSales)
    current.periods <<- seq(1,current.n)
    e = input$MonthSales - y.hat
    if ((e>UB)||(e<LB)){
      output$myresults1 = renderText({
        paste0("Input #", current.n, " Error outside tolerance interval: We need to update the model!")
      })
      reg5 <<- lm(current.sales~current.periods)
      current.mse <<- mse(current.sales,fitted(reg5))
      LB <<- -qnorm(0.975)*sqrt(current.mse)
      UB <<- qnorm(0.975)*sqrt(current.mse)
    }
    else{
      output$myresults1 = renderText({
        paste0("Input #", current.n, " MSE ratio is within tolerance: We keep the model!")
      })
    }
  })
}
shinyApp(ui, server)

#web scraping
install.packages("XML")
library(XML)
uconn.url="https://uconn.edu/"
uconn.source=readLines(uconn.url,encoding="UTF-8")
View(uconn.source)
uconn.parsed=htmlParse(uconn.source,encoding="UTF-8")
getHTMLLinks(uconn.source)
getHTMLExternalFiles(uconn.source)
x=readHTMLList(uconn.source)
x[19]
y=unlist(x[19])
y[1]


myurl="https://flightaware.com/live/airport/KBDL"
mysource=readLines(myurl,encoding = "UTF-8")
mytables=readHTMLTable(mysource)
arrivals=mytables[2]$`NULL`
class(arrivals)
View(arrivals)
View(mytables[2]$`NULL`)
arrivals = arrivals[c(-1,-2),-5]
View(arrivals)

Extract.Table = function(myurl){
  mysource=readLines(myurl,encoding = "UTF-8")
  mytables=readHTMLTable(mysource)
  arrivals=mytables[2]$`NULL`
  arrivals=arrivals[3:13,-5]
  return(arrivals)
}

View(Extract.Table("https://flightaware.com/live/airport/KJFK"))

ui = fluidPage(
  tableOutput("mytable")
)
server = function(input, output, session) {
  autoInvalidate <- reactiveTimer(30000)
  output$mytable=renderTable({
    autoInvalidate()
    Extract.Table("https://flightaware.com/live/airport/KJFK")
  })
}
shinyApp(ui, server)

library(openxlsx)
ui = fluidPage(
  radioButtons("airport","Select Airport", c("BDL"="bdl","JFK"="jfk","LAX"="lax")),
  plotOutput("myplot"),
  tableOutput("mytable")
)
server = function(input, output, session) {
  autoInvalidate <- reactiveTimer(1000)
  arrivals=reactive({
    airportname = switch(input$airport,
                         bdl = "BDL",
                         jfk = "JFK",
                         lax = "LAX"
                         
    )
    autoInvalidate()
    Extract.Table(paste0("https://flightaware.com/live/airport/K",airportname))
  })
  output$mytable=renderTable({
    autoInvalidate()
    arrivals()
  })
  output$myplot=renderPlot({
    df = read.xlsx("/Users/YH/Desktop/data.xlsx")
    df = rbind(df,arrivals())
    write.xlsx(arrivals(),"/Users/YH/Desktop/data.xlsx")
    autoInvalidate()
    barplot(prop.table(table(arrivals()$Type)))
  })
}
shinyApp(ui, server)




mse=function(y,yy){
  return(mean((y-yy)^2))
}

update.mse = function(old.mse, old.n, y.new, yy.new){
  return((old.n/(old.n + 1))*old.mse+(1/(old.n + 1))*(y.new-yy.new)^2)
}

current.mse=mse(y,fitted(reg1))
current.n=12
current.sales=y
current.periods=seq(1,12)
mytol = 0.1
ui = fluidPage(
  titlePanel("Battery Sales"),
  hr(),
  hr(),
  sliderInput("MonthSales",label="Select Monthly Sales",min=0,max=100,value=20,step=1),
  actionButton(inputId = "mybutton1",label="Update"),
  actionButton(inputId = "mybutton2",label="Predict"),
  hr(),
  verbatimTextOutput("myresults1")
)
server = function(input, output) {
  observeEvent(input$mybutton1,{
    output$myresults2 = renderText({""})
    y.hat=coef(reg1)[1]+coef(reg1)[2]*(current.n + 1)
    new.mse=update.mse(current.mse,current.n,input$MonthSales,y.hat)
    current.n <<- current.n + 1
    current.sales <<- c(current.sales,input$MonthSales)
    current.periods <<- seq(1,current.n)
    if (new.mse/current.mse > mytol + 1){
      output$myresults1 = renderText({
        paste0("Input #", current.n, " MSE ratio exceeds tolerance: We need to update the model!")
      })
      reg1 <<- lm(current.sales~current.periods)
      current.mse <<- mse(current.sales,fitted(reg1))
    }
    else{
      output$myresults1 = renderText({
        paste0("Input #", current.n, " MSE ratio is within tolerance: We keep the model!")
      })
      current.mse <<- new.mse
    }
  })
}
shinyApp(ui, server)

# 
y = BatterySales$`Battery Sales`
x = seq(1,12)
reg5 = lm(y~x)

current.mse=mse(y,fitted(reg5))
current.n=12
current.sales=y
current.periods=seq(1,12)
LB = -qnorm(0.975)*sqrt(current.mse)
UB = qnorm(0.975)*sqrt(current.mse)
ui = fluidPage(
  titlePanel("Battery Sales"),
  hr(),
  hr(),
  sliderInput("MonthSales",label="Select Monthly Sales",min=0,max=100,value=20,step=1),
  actionButton(inputId = "mybutton",label="Update"),
  hr(),
  verbatimTextOutput("myresults1")
)
server = function(input, output) {
  observeEvent(input$mybutton,{
    y.hat=coef(reg5)[1]+coef(reg5)[2]*(current.n + 1)
    current.n <<- current.n + 1
    current.sales <<- c(current.sales,input$MonthSales)
    current.periods <<- seq(1,current.n)
    e = input$MonthSales - y.hat
    if ((e>UB)||(e<LB)){
      output$myresults1 = renderText({
        paste0("Input #", current.n, " Error outside tolerance interval: We need to update the model!")
      })
      reg5 <<- lm(current.sales~current.periods)
      current.mse <<- mse(current.sales,fitted(reg5))
      LB <<- -qnorm(0.975)*sqrt(current.mse)
      UB <<- qnorm(0.975)*sqrt(current.mse)
    }
    else{
      output$myresults1 = renderText({
        paste0("Input #", current.n, " MSE ratio is within tolerance: We keep the model!")
      })
    }
  })
}
shinyApp(ui, server)

#web scraping
install.packages("XML")
library(XML)
uconn.url="https://uconn.edu/"
uconn.source=readLines(uconn.url,encoding="UTF-8")
View(uconn.source)
uconn.parsed=htmlParse(uconn.source,encoding="UTF-8")
getHTMLLinks(uconn.source)
getHTMLExternalFiles(uconn.source)
x=readHTMLList(uconn.source)
x[19]
y=unlist(x[19])
y[1]


myurl="https://flightaware.com/live/airport/KBDL"
mysource=readLines(myurl,encoding = "UTF-8")
mytables=readHTMLTable(mysource)
arrivals=mytables[2]$`NULL`
class(arrivals)
View(arrivals)
View(mytables[2]$`NULL`)
arrivals = arrivals[c(-1,-2),-5]
View(arrivals)

Extract.Table = function(myurl){
  mysource=readLines(myurl,encoding = "UTF-8")
  mytables=readHTMLTable(mysource)
  arrivals=mytables[2]$`NULL`
  arrivals=arrivals[3:13,-5]
  return(arrivals)
}

View(Extract.Table("https://flightaware.com/live/airport/KJFK"))

ui = fluidPage(
  tableOutput("mytable")
)
server = function(input, output, session) {
  autoInvalidate <- reactiveTimer(30000)
  output$mytable=renderTable({
    autoInvalidate()
    Extract.Table("https://flightaware.com/live/airport/KJFK")
  })
}
shinyApp(ui, server)

library(openxlsx)
ui = fluidPage(
  radioButtons("airport","Select Airport", c("BDL"="bdl","JFK"="jfk","LAX"="lax")),
  plotOutput("myplot"),
  tableOutput("mytable")
)
server = function(input, output, session) {
  autoInvalidate <- reactiveTimer(1000)
  arrivals=reactive({
    airportname = switch(input$airport,
                         bdl = "BDL",
                         jfk = "JFK",
                         lax = "LAX"
                         
    )
    autoInvalidate()
    Extract.Table(paste0("https://flightaware.com/live/airport/K",airportname))
  })
  output$mytable=renderTable({
    autoInvalidate()
    arrivals()
  })
  output$myplot=renderPlot({
    df = read.xlsx("/Users/YH/Desktop/data.xlsx")
    df = rbind(df,arrivals())
    write.xlsx(arrivals(),"/Users/YH/Desktop/data.xlsx")
    autoInvalidate()
    barplot(prop.table(table(arrivals()$Type)))
  })
}
shinyApp(ui, server)