dataM_shiny

pacman::p_load(tidyverse,shiny,ggvis,maps,reshape2)

Q1

##Q1

ui_0 <- shinyUI(fluidPage(
  h3("Interactive Normal Distribution"),
  sidebarLayout(
    sidebarPanel(
      sliderInput("Mean",label = "Mean of normal distribution",min = -10,max = 10,value = 0),
      sliderInput("Std",label = "Standard deviation of normal distribution",min = 0,max = 2,value = 1,
                  step = 0.1)
    ),
    mainPanel(
      fluidRow(
        plotOutput("out.plot")
      )
    )
  )
))


server_0 <- shinyServer(function(input,output){
  output$out.plot <- renderPlot({
    curve(dnorm(x,mean = input$Mean,sd = input$Std),xlim = c(-10,10),ylim = c(0,5))
    abline(v = input$Mean,lty = 2,col = "red")
    text(7, 4.5, paste0("Mean = ",input$Mean,", SD = ",input$Std))
  })
})

#shinyApp(ui_0,server_0)

Q2

tw <- read.csv("C:/Users/Cheng_wen_sung/Desktop/tw_export_import.csv",header = F)

# unit: 1000 USD

tw$V1 <- as.character(tw$V1)
tw <- data.frame(
  matrix(unlist(strsplit(tw$V1, 'M')), 192, 2, byrow = TRUE),
  tw[, -1])

EI <- rep(c("_E", "_I"), 33)
Country <- rep(c("Total",
                 "Asia", "HK", "ID", "JP", "KR", "CN", "MY", "SG", "THA", "VEN",
                 "Nahost", "KW", "SB",
                 "EU", "FR", "DE", "IT", "EN", "Africa", "NG",
                 "NAmerica", "CA", "US",
                 "MAmerica", "MX", "PA",
                 "SAmerica", "AR", "BR",
                 "Oceania", "AU", "NZ"), each = 2)
names(tw) <- c("Year", "Month", paste0(Country, EI))

# tw reshape + map + static

tw_long <- reshape(tw, varying = names(tw)[3:68],
                   v.names = 'USD', times = names(tw)[3:68],
                   direction = "long")
tmp <- matrix(unlist(strsplit(tw_long$time, "_")),
              12672, 2, byrow = TRUE)
tw_new <- data.frame(tw_long[, c(1, 2, 5)], tmp)
names(tw_new)[4:5] <- c("Country", "Type")


# group by Year and Country

tw_year <- tw_new %>%
  group_by(Year, Country, Type) %>%
  summarise(Sum_of_Year = sum(USD))
tw_year <- as.data.frame(tw_year)

levels(tw_year$Country) <- c("Africa", "Argentina", "Asia",
                             "Australia", "Brazil", "Canada", "China", "Germany", "UK",
                             "Europe", "France", "Hong Kong", "Indonesia", "Italy", "Japan",
                             "South Korea", "Kuwait", "MAmerica", "Mexico", "Malaysia",
                             "Nahost", "NAmerica", "Nigeria", "New Zealand", "Oceania",
                             "Panama", "SAmerica", "Saudi Arabia", "Singapore", "Thailand",
                             "Total", "USA", "Vietnam")
levels(tw_year$Type) <- c("Export", "Import")

#
tmp <- tw_year %>%
  subset(Year == 2015) %>%
  subset(Country != "Africa") %>%
  subset(Country != "Asia") %>%
  subset(Country != "Europe") %>%
  subset(Country != "Hong Kong") %>%
  subset(Country != "Nahost") %>%
  subset(Country != "MAmerica") %>%
  subset(Country != "NAmerica") %>%
  subset(Country != "SAmerica") %>%
  subset(Country != "Oceania") %>%
  subset(Country != "Total")

#
export <- subset(tmp, Type == "Export")
import <- subset(tmp, Type == "Import")
Diff_of_Year <- export$Sum_of_Year - import$Sum_of_Year
Type <- ifelse(Diff_of_Year > 0, "export > import", "import > export")
tw_2015 <- data.frame(export[, 1:2], Diff_of_Year, Type)

#

wrld <- map_data(map = "world")
wrld$cc <- rep("0", 99338)

USD <- vapply(wrld$regi, function(x) {
  if(x %in% as.character(tw_2015$Country)) {
    tw_2015[which(tw_2015$Country == x), 'Diff_of_Year']
  } else {0}
}, numeric(1))

wrld$USD <- USD
wrld$balance <- ifelse(wrld$USD == 0, "",
                       ifelse(wrld$USD < 0, "Surplus", "Deficit"))
#wrld$USD[which(wrld$USD == 0)] <- NA
wrld$USD2 <- ifelse(wrld$USD < 0, -1,
                    ifelse(wrld == 0, 0, 1))

##ggvis

#p <- ggvis(data = wrld,x = ~long,y = ~lat,fill = ~USD)%>%
#  group_by(group)%>%
#  layer_paths()

#p %>% add_tooltip(function(wrld) wrld$USD)

Q3

server.3 <- shinyServer(function(input, output){
  df <- replicate(100, sample(1:6, 100, repl = TRUE))
  tot_m <- mean(df)
  df <- as.data.frame(df)
  df_m <- colMeans(df)
  df_s <- colSums(df)
  df_exp <- data.frame(index = 1:100,
                       expec = cumsum(df_s) / (seq(1:100) * 100))
  
  theme_set(theme_bw())
  output$bar <- renderPlot({
    ggplot(df, aes_string(x = paste0("V", input$roll))) +
      geom_bar() + labs(x = "Roll 100 dices") +
      lims(x = c("1", "2", "3", "4", "5", "6"),y = c(0,35))
  })
  
  output$expectation <- renderPlot({
    ggplot(df_exp[1:input$roll,], aes(x = index, y = expec)) +
      geom_line() +
      geom_hline(yintercept = tot_m, linetype = "dashed") +
      xlim(1, 105) + ylim(2, 5)
  })
})


# ui

ui.3_1 <- shinyUI(fluidPage(
  h3("Rolling dice"),
  actionButton("roll", label = "Roll the dice!"),
  #  actionButton("reset", label = "Reset"),
  fluidRow(
    column(4, plotOutput("bar")),
    column(4, plotOutput("expectation")))
))


ui.3_2 <- shinyUI(fluidPage(
  h3("Rolling 100 dice"),
  sliderInput("roll",label = "Rolling times",min = 1,max = 100,value = 1,step = 1),
  #  actionButton("reset", label = "Reset"),
  fluidRow(
    column(4, plotOutput("bar")),
    column(4, plotOutput("expectation")))
))



#shinyApp(ui.3_1,server.3)
#shinyApp(ui.3_2,server.3)