io_app.Rmd

---
title: "More Interactive Applications"
author: "Daniel N. Albohn"
date: "12/06/2017"
output: 
  html_document:
    theme: cosmo
    css: style.css
editor_options: 
  chunk_output_type: console
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE, eval = FALSE)
```

# Stepping Up the Problem
Obviously, we are all shiny experts by now. The previous walk through gave us a crash
course on how to layout an app, how to provide static elements like text and images, and
finally how to deal with reactive events by setting up an observer.

A logical next step might be how to construct an app that not only responds to
user-defined information, but different types of user information.

Let's build an app that takes a `csv` file, allows the user to preview the data file that
was read in, select two numeric variables, and plot those against each other. This seems
like quite an easy task, given what we just accomplished before, but when we break down
the step-by-step issues it is quite complex.

**Disclosure**: Most of this app was scavenged from parts of the
[File Upload](http://shiny.rstudio.com/gallery/file-upload.html) tutorial,
and this [stackoverflow](https://stackoverflow.com/questions/36949769/how-to-plot-uploaded-dataset-using-shiny)
problem on reactive data sets.

# The Problem(s)
An in-depth walk through of this issue is beyond the scope of this tutorial, but let
me highlight some of the new issues that this problem brings to a coder's attention.

**1). Upload an external `.csv` file**
  
  - Allow the user to search machine for file and upload it
    + `fileInput()`
  - Cause non-perfect `.csv` files to not break the app
    + User selected file attributes

**2). Preview uploaded file as data frame**
    
  - If the file can be uploaded successfully, need to make the data frame
  reactive, not static
    + Can create a `reactive()` data object
    + Grabbing the first few rows of our reactive data + `tableOutput()`

**3). Select two numeric variables**
    
  - How are we going to fit all of this on the same page?
    + Scroll?
    + `tabsetPanel()` with one for uploading the file, another for selecting and
    plotting
  - Allow the user to select two variables that can constantly change
    + Drop down list with variable names from data set
    + `selectInput()`
  - Allow for constant updating of user-defined variable combinations
    + `updateSelectInput()`
  - Isolate those variables somehow for further use
    + Subsetting a reactive data frame?

**4). Plot user-selected variables**
  
  - `renderPlot()` + `plotOutput()`

# The Solution
You can see this app running on [shinyapps.io](https://d-bohn.shinyapps.io/io_example_app/)

```{r}
library(shiny)
library(ggplot2)

ui <- fluidPage(
  titlePanel("File Upload and Scatter Plot"),
  tabsetPanel(
    
    ## First tab
    tabPanel("Upload File",
             
             ## A sidebar layout within the first tab
             sidebarLayout(
               
               ## left sidebar gets this:
               sidebarPanel(
                 fileInput('file1', 'Choose CSV File',
                           accept=c('text/csv', 
                                    'text/comma-separated-values,text/plain', 
                                    '.csv')),
                 ## -----
                 tags$br(),
                 ## -----
                 
                 ## Some values for how to read in the file
                 checkboxInput('header', 'Header', TRUE),
                 radioButtons('sep', 'Separator',
                              c(Comma=',',
                                Semicolon=';',
                                Tab='\t'),
                              ',')
               ),
               
               ## Main panel gets this:
               mainPanel(
                 tableOutput('contents')
               )
             )
    ),
    
    ## Second tab
    tabPanel("Scatter Plot",
             sidebarLayout(
               sidebarPanel(
                 
                 selectInput('xcol', 'X Variable', ""),
                 selectInput('ycol', 'Y Variable', "", selected = "")
                 
               ),
               mainPanel(
                 plotOutput('MyPlot')
               )
             )
    )
    
  )
)

server <- function(input, output, session) {
  # added "session" because updateSelectInput requires it
  
  data <- reactive({ 
    req(input$file1) 
    
    inFile <- input$file1 

    df <- read.csv(inFile$datapath, header = input$header, sep = input$sep)
  
    updateSelectInput(session, inputId = 'xcol', label = 'X Variable', choices = names(df),
                      selected = names(df)[sapply(df, is.numeric)])
    updateSelectInput(session, inputId = 'ycol', label = 'Y Variable', choices = names(df),
                      selected = names(df)[sapply(df, is.numeric)])
    
    return(df)
  })
  
  output$contents <- renderTable({
    head(data(), 10)
  })
  
  output$MyPlot <- renderPlot({
    
    x <- data()[, input$xcol]
    y <- data()[, input$ycol]
    newdf <- as.data.frame(cbind(x,y))
    
    ggplot(newdf, aes(x, y)) +
      geom_point() +
      xlab(as.character(input$xcol)) +
      ylab(as.character(input$ycol)) +
      stat_smooth(method = 'lm') +
      theme_bw(base_size = 15)
    
  })
}

shinyApp(ui, server)
```


If your itch for shiny tutorials still isn't complete, check out how to use the `shiny`
package for collecting surveys with this simple [survey app](simple_shiny_survey.html).
You can also check out the [resource page](beyond_hello.html)
for more tutorials and how to extend shiny.