#
library(dplyr)
library(shiny)
library(ggplot2)
master<-read.csv(url("http://www.studenti.famnit.upr.si/~89161011/masterBig.csv"))
master<-masterBig

master2<-master %>% group_by(country,year,sex) %>% summarise(GDP = mean(GDP),suicides_no=sum(suicides_no),population=sum(population))






# Define UI for application that draws a histogram
ui <- fluidPage(

   # Application title
   titlePanel("Suicid rate compared to gender, GDP, age and population"),
   HTML("<p>In this project I am trying to investigate suicidal rates in the world and whether things like gender, age,
        GDP of certain country or population size plays role in predicting future rates. By comparing results of each country we can see which one
        was the most succesful in tackling this issue</p>"),

   # Sidebar with a slider input for number of bins
   sidebarLayout(
      sidebarPanel(
        selectInput(inputId = "country",
                    label = "Select country:",
                    choices = master2$country,
                    selected = "Slovenia",
                    multiple = TRUE,
                    selectize = TRUE),

        radioButtons("dist", "Choose criteria:",
                     c("Suicid rate according to gender" = "sexbutton",
                       "Suicid rate according to GDP" = "GDPbutton",
                       "Suicid rate according to age" = "agebutton",
                       "Suicid-population growth relation" = "populationbutton")),


        wellPanel(
          # Show data table
          checkboxInput(inputId = "show_data",
                        label = "Show data table",
                        value = TRUE)
        ),

        HTML("<h3>Data interpretation</h3>
             <p>As we switch between radio buttons in order to see how GDP,gender,age and population affect suicidal rates accros different countries.
             By doing data modeling and visualization we get a chance to draw a conclusion that gender plays huge role. Males commit suicide more often in
             all observations. Another conlcusion we could make is that suicidal rate is reciprocal to GDP growth in most observations. Most observations showed
             that population grows despite high level of suicids. People of age 35-54 are more likely commit suicide than any other age group. Kids aged from 5 to 14
             are not at high risk of self-harm according to our plot.Interesting thing is that in Slovenia critical age group is one from 55-74 and fact that
             35-54 population is decreasing drastically which is amazing considering that this might be the only case in the whole world.</p><h4>Elmir SUT</h4>")


          ), #closing sidebarPanel
      # Show a plot of the generated distribution
      mainPanel(
        plotOutput(outputId = "scatterplot"),
        conditionalPanel(condition = "input.show_data == true", h3("Data table")),     # Third level header: Data table
        DT::dataTableOutput(outputId = "countrystable"),

        width = 7
      )
      ) #closing sidebarLayout
  )#closing fluidPage

# Define server logic required to draw a histogram
server <- function(input, output) {

   output$scatterplot <- renderPlot({
     req(input$country)
     if(input$dist=="sexbutton"){
       master2$year<-as.factor(master2$year)
     filter(master2,country==input$country) %>%
       ggplot(., aes(x=year, y=suicides_no, group=sex, color=sex)) +
         geom_line() + geom_point()+
         scale_color_brewer(palette="Dark2")+
         theme_minimal()

     }else if (input$dist=="GDPbutton"){
       filter(master2,country==input$country)%>%group_by(., country,year)%>%summarize(GDP=mean(GDP),suicides_no=sum(suicides_no))%>%mutate(gdpsuicid=min(suicides_no)/min(GDP)*GDP)%>%
       ggplot(., aes(x=year,y=(max(suicides_no)))) + geom_line(aes(y=suicides_no, color="suicides")) +
          geom_line(aes(y=gdpsuicid, color="gdp_ratio")) +

         theme_minimal()
     }else if(input$dist=="agebutton"){
       master2$year<-as.factor(master2$year)
       filter(master,country==input$country) %>% group_by(.,age,year)%>%summarize(suicides_no=sum(suicides_no))%>%
         ggplot(., aes(x=year, y=suicides_no, group=age, color=age)) +
         geom_line(size=1.5) + geom_point(size=1.5)+
         scale_color_brewer(palette="Set2")+
         theme_dark()
     } else{
       filter(master,country==input$country) %>% group_by(.,year)%>%summarize(.,suicides_no=sum(suicides_no),population=mean(population))%>%
         ggplot(., aes(x=year,y=min(population/500))) +
         geom_line(aes(y=suicides_no),col="red", size=1.1,stat="identity")+geom_line(aes(y=(population/500),col="yellow", size=1.1,stat="identity")) +
         scale_color_brewer(palette="Set2")+     scale_color_discrete(name = "Population/Suicid", labels = c("Population", "Suicid rate"))+
         theme_dark()
     }


   })

   output$countrystable <- DT::renderDataTable(
     if(input$show_data){
       DT::datatable(data = master2[, 1:6],
                     options = list(pageLength = 10),
                     rownames = FALSE)
     }
   )
}

# Run the application
shinyApp(ui = ui, server = server)