Shiny R assign variable based on dropdown using data.frame column and pass this variable to plotly to plot visualization

Question

i have a data.frame from my DB, and i am trying to create a R Shiny Web App in which i could choose from the two dropdown menus (based on values from combination of row and column in my data.frame ) and pass those variables to plotly

Lets say i have this data in my data frame:

# |User|N00   |Percentage|N01 |
  |----|------|----------|----|
  |35  |2229  | 4.46830  |918 |
  |37  |153   | 0.00307  |0   |
  |39  |1557  | 0.03120  |0   |
  |41  |4997  | 0.10014  |0   |
  |42  |1     | 0.00002  |0   |
  |44  |12221 | 0.24490  |0   |
  |46  |5     | 0.00010  |0   |
  |50  |342   | 0.00685  |0   |
  |51  |123   | 0.00246  |0   |
  |56  |2     | 0.00004  |1   |
  |57  |0     | 0.00000  |0   |
  |60  |62275 |1.24796   |2799|

What i would like to do is to have one dropdown with values in Column "User" and second dropdown with Headers, so i could choose "User" and "Header" and get back the appropriate result (data.frame).

So what i would need from the combination of the two dropdowns is to do something like this:

select(data_to_choose_from, User, values)[data_to_choose_from$id==values2, ]

Where values is the result of first dropdown (user), and values2 is the result of the second dropdown (headers), which looks like this:

So when i choose for example User "35" and Columnn "N00", i would like to get this result:

 |User|N00   |
 |----|------|
 |35  |2229  |

After i get this result i would like to pass the number in Column "N00" to some variable and create gauge with this number that would for example look like this (this is just and example, i have to tweak this to make it more "legible"):

I have this:

library(shiny) #Load libraries
library(plotly) #Load libraries

data_to_choose_from <- data.frame(a) #Assign data from data.frame(a), which is data i have from my DB to new variable

ui <- fluidPage(
    sidebarLayout(
        sidebarPanel(
            selectizeInput("User", "User", choices=NULL, selected=NULL), #First dropdown
            selectizeInput("Metrics", "Metrics", choices=NULL, selected=NULL), #Second dropdown
            textOutput("values"), #Show what did i choose in the first dropdown
            textOutput("values2"), #Show what did i choose in the second dropdown
            tableOutput("table") #Output tha table based on my choices in dropdowns
        ),
        mainPanel(
            plotlyOutput('p') #Plot a visualization base on data from output table
        )
    )
)

server <- function(input, output, session) { #First dropdown, reference to my data.frame
    updateSelectizeInput(session, 'User',
                         choices = data_to_choose_from$User,
                         server = TRUE
    )
    updateSelectizeInput(session, 'Metrics',
                         choices = names(data_to_choose_from), #Second dropdown, reference to my data.frame
                         server = TRUE
    )

    output$values <- renderText({(input$User)}) #Output of first dropdown

    output$values2 <- renderText({(input$Metrics)}) #Output of second dropdown

    output$values3 <- as.vector(select(data_to_choose_from, values)[data_to_choose_from$id==values2, ]) #Put this as vector to plot_ly

    output$table <- select(data_to_choose_from, id, values)[data_to_choose_from$id==values2, ] #Filtered data.frame output based on original data

    output$plot <- renderPlotly( #Here i would like to create the plot to render in mainpanel
    p <- plot_ly(
        type = "indicator",
        mode = "gauge+number+delta",
        value = i,
        title = list(text = values2, font = list(size = 24)),
        delta = list(reference = 400, increasing = list(color = "RebeccaPurple")),
        gauge = list(
            axis = list(range = list(NULL, 500), tickwidth = 1, tickcolor = "darkblue"),
            bar = list(color = "darkblue"),
            bgcolor = "white",
            borderwidth = 2,
            bordercolor = "gray",
            steps = list(
                list(range = c(0, values3), color = "cyan"),
                list(range = c(values3, 400), color = "royalblue")),
            threshold = list(
               line = list(color = "red", width = 4),
                thickness = 0.75,
                value = values3 + 100))) %>%
        layout(
            margin = list(l=20,r=30),
            paper_bgcolor = "lavender",
            font = list(color = "darkblue", family = "Arial")))
}
shinyApp(ui, server)

Unfortunately this does not work, i suppose my code is wrong from the ground up, i try to follow the documentation but i am kinda lost..

I am trying to learn R and i am confused as to how variables and Shiny R in general works, any information that would help me to achieve what i want would be greatly appreacited, if there is something unclear please let me know and i will edit my question, thanks

Answer 1

There are multiple problems.

When you do output$values <- renderSomething({......}) , you can't access to ...... by calling values (you even can't call output$values ).

Replace output$values3 with

  values3 <- reactive({
    req(input$User, input$Metrics)
    data_to_choose_from[data_to_choose_from$User == input$User, input$Metrics]
  })

Now you can get the value of values3 by calling values3() . So replace values3 in your renderPlotly with values3() . And replace text = values2 with text = input$Metrics .

The line req(input$User, input$Metrics) means that input$User and input$Metrics must be available. This avoids an error at the launching of the app because these variables are available only after the action of updateSelectizeInput .

If you want to render a table, use renderTable :

  output$table <- renderTable({
    req(input$User, input$Metrics)
    subset(data_to_choose_from, User == input$User, select = c("User", input$Metrics)) 
  })

You do output$plot <- renderPlotly({ p <- plot_ly(.......) }) so in your ui that should be plotlyOutput("plot") , not plotlyOutput("p") . The p <- is useless, you can remove it.

Now in plot_ly you do value = i but i is not defined. I don't know what you intended to do (I'm not familiar with this kind of plots).