Shiny input to Plotly does not produce a plot

Question

I'm building an shiny application to show some quality control data to our clients. First i had the application created with GGplot functionalities. Now i am converting all graphs to Plotly output. For one of these plots (a boxplot). I have the problem that i cant pass a shiny input selector to the plot.

In GGplot there is no problem at all and the plot is changed each time i choose a different plotColumn. Here i solved the problem of column parsing with the aes_string function. Basically i am looking for something similar in plotly.

Working GGPLOT example:

ggplot(finalDf, aes_string("runName",input$getBoxplotField),na.rm = T) +
        geom_boxplot(aes_string(fill="runName"), notch = F) +
        geom_jitter() +
        scale_y_continuous(labels = format1) +
        theme_bw()

Not working Plot_ly example

p <- plot_ly(finalDf,x = runName, y = input$getBoxplotField, type = "box")

exampleDf

> dput(head(finalDf))
structure(list(runName = c("Gentrap.1451849446759", "Gentrap.1451849446759", 
"Gentrap.1451849446759", "Gentrap.1451849446759", "Gentrap.1451849446759", 
"Gentrap.1451849446759"), sampleName = c("Hart_FC42b_L5_I2_SRD329", 
"S1", "S2", "S3","S4", "S5"), readGroupName = c(NA, 
NA, NA, NA, NA, NA), maxInsertSize = c(227615351L, 202850798L, 
249001722L, 234388122L, 188295691L, 249009605L), medianCvCoverage = c(0.501303, 
0.494183, 0.574364, 0.487233, 0.495491, 0.483041), medianInsertSize = c(197L, 
203L, 200L, 208L, 200L, 194L), median3PrimeBias = c(0.283437, 
0.263973, 0.372476, 0.266946, 0.296308, 0.292954), median5PrimeBias = c(0.139005, 
0.21233, 0.123449, 0.185168, 0.169128, 0.152902), median5PrimeTo3PrimeBias = c(0.586081, 
0.9234, 0.409042, 0.83276, 0.680496, 0.640518), nBasesAligned = c(1627112497, 
1572782400, 1772774189, 1595461211, 1593529487, 1705441762), 
    nBasesCoding = c(795255442, 778886694, 762223625, 819014623, 
    759061861, 838846117), nBasesIntergenic = c(140893219, 176728812, 
    194156767, 120900630, 137267440, 148815172), nBasesIntron = c(134528982, 
    111795186, 121091943, 96554581, 142587231, 139962698), nBasesRibosomal = c(NA, 
    NA, NA, NA, NA, NA), nBasesUtr = c(556434854, 505371708, 
    695301854, 558991377, 554612955, 577817775), nCorrectStrandReads = c(NA_integer_, 
    NA_integer_, NA_integer_, NA_integer_, NA_integer_, NA_integer_
    ), nIncorrectStrandReads = c(NA_integer_, NA_integer_, NA_integer_, 
    NA_integer_, NA_integer_, NA_integer_), nReadsAligned = c(33157934L, 
    32082625L, 36181227L, 32595741L, 32538544L, 34783342L), nReadsProperPair = c(31935921L, 
    30983730L, 35015854L, 31358224L, 31405592L, 33479007L), nReadsSingleton = c(3919886L, 
    4311016L, 4382092L, 3848808L, 3873270L, 4122759L), nReadsTotal = c(37077604L, 
    36393382L, 40563115L, 36444288L, 36411547L, 38905908L), pctChimeras = c(0.004783, 
    0.003078, 0.003063, 0.004278, 0.002983, 0.00485), rateIndel = c(0.000071, 
    0.000076, 0.000081, 0.000066, 0.000072, 0.00007), rateReadsMismatch = c(0.001438, 
    0.001643, 0.001627, 0.001467, 0.001716, 0.001471), stdevInsertSize = c(120.677992, 
    129.927513, 114.820226, 138.486257, 118.98163, 115.25774), 
    group = c("Gentrap.1451849446759", "Gentrap.1451849446759", 
    "Gentrap.1451849446759", "Gentrap.1451849446759", "Gentrap.1451849446759", 
    "Gentrap.1451849446759")), .Names = c("runName", "sampleName", 
"readGroupName", "maxInsertSize", "medianCvCoverage", "medianInsertSize", 
"median3PrimeBias", "median5PrimeBias", "median5PrimeTo3PrimeBias", 
"nBasesAligned", "nBasesCoding", "nBasesIntergenic", "nBasesIntron", 
"nBasesRibosomal", "nBasesUtr", "nCorrectStrandReads", "nIncorrectStrandReads", 
"nReadsAligned", "nReadsProperPair", "nReadsSingleton", "nReadsTotal", 
"pctChimeras", "rateIndel", "rateReadsMismatch", "stdevInsertSize", 
"group"), row.names = c(NA, 6L), class = "data.frame")

server.R

shinyServer(function(input, output, session) {
  output$selectBoxplotField <- renderUI({
    selectInput("getBoxplotField", label = "Select variable to plot", choices = names(getAllSampleStats()))
  })

output$boxplot <- renderPlotly({
    finalDf #as defined above in the example
     p <- plot_ly(finalDf, x = runName, y = input$getBoxplotField , type = "box")
})
}

GUI.R

shinyUI(navbarPage(
  theme = "bootstrap_sandstone.css",
  "SPIN", fluid = T,
  tabPanel("Gentrap",
           fluidPage(fluidRow(
             sidebarlogin(pipelineName = "gentrap"),
             column(10,
                    tabsetPanel(
                      tabPanel("Metrics distribution",
                               fluidRow(
                                 column(2),
                                 column(8, plotlyOutput("boxplot")),
                                 column(2)
                               ),
                               fluidRow(
                                 column(3, uiOutput("selectBoxplotField")),
                                 column(3, checkboxInput("checkboxplot", label = "Compare to All", value = TRUE))
                               ),
                               fluidRow(
                                 column(9, helpText("If no plot shows up it means this data is not present in the Sentinel QC database"))
                               )),
                    ))
           )))
))

Answer 1

The problem is fixed by passing the DF plus columns directly to the X and Y axes without first passing the DF name as a argument.

Proper plot will be generated when this is done:

    plot_ly(x = finalDf[,'runName'], y = finalDf[,input$getBoxplotField] , type = "box", color = 'red') %>%
    layout(xaxis = list(showticklabels = FALSE, title = ''), yaxis = yName)

This is wrong:

    plot_ly(finalDf, x = runName, y = input$getBoxplotField , type = "box", color = 'red') %>%
    layout(xaxis = list(showticklabels = FALSE, title = ''), yaxis = yName)