I want to create a clickable histogram in shiny but I don't know if it is possible.
Some months ago I saw a clickable volcano plot which gives you a table of what you click.
Source: https://2-bitbio.com/2017/12/clickable-volcano-plots-in-shiny.html
The closest post that I found about creating clickable histograms is this one Click to get coordinates from multiple histogram in shiny
However, I don't want to get the coordinates. I want the rownames of the dataframe.
Having this dataframe, can I get the rownames everytime I click a bar from the histogram?
mtcars <- mtcars %>%
select("hp")
mtcars <- as.matrix(mtcars)
One example (but not clickable) in shiny:
library(shiny)
library(ggplot2)
library(scales)
library(dplyr)
ui <- fluidPage(
titlePanel("Histogram"),
sidebarLayout(
sidebarPanel(
),
mainPanel(
plotOutput("hist"),
)
)
)
mtcars <- mtcars %>%
select("hp")
mtcars <- as.matrix(mtcars)
server <- function(input, output) {
output$hist <- renderPlot({
pp <- qplot(mtcars, geom = "histogram", bins = 10, xlab="values",
ylab="Frequency", main="Histogram",
fill=I("red"), col=I("black"), alpha=I(0.4))
pp + scale_x_continuous(breaks=pretty(mtcars, n=10))
})
}
shinyApp(ui = ui, server = server)
Does anyone know how to do it?
Thanks very much in advance!
Regards
This is a great question, and what makes it challenging is that the qplot/ggplot charts are static images. The below app.r is an example of how I would do it. I'd love to see other approaches.
In essence:
Create a sequence of numbers that will be used both as the breaks in your histogram and as intervals in your dataframe. I based these on user inputs, but you could hardcode them.
Assign a "bin" value to each row in the dataframe based on the interval in which the value falls.
Record the x-coordinate from the user's click event and assign that a "bin" value based on the same set of intervals.
Subset your dataframe and retain only those records where the "bin" value of the data matches the "bin" value of the x-coordinate from the user's click event.
Otherwise, if you're willing to go the d3 route, you could explore something like this posted by R Views.
#Load libraries ----------------------------------------------------
library(shiny)
library(ggplot2)
library(scales)
library(dplyr)
# Prepare data -----------------------------------------------------
df <- mtcars
df <- cbind(model = rownames(df), data.frame(df, row.names = NULL)) # setting the rownames as the first column
dm <- df$hp %>% as.matrix()
# UI function ------------------------------------------------------
ui <- fluidPage(
titlePanel("Histogram"),
sidebarLayout(
sidebarPanel(
tags$h5("I added the below text output only to demonstrate shiny's way for tracking user interaction on static plots. You can click, double-click, or click & drag (i.e. brushing). These functions are AWESOME when exploring scatterplots."),
tags$h3("Chart click and brushing"),
verbatimTextOutput("info"),
tags$h5("Now I'm applying the below UI inputs to the `vec` and `breaks` arguments in `findInterval()` and `qplot()` respectively; I'm using `findInterval()` to bin the values in the dataframe AND to bin the x-value of the user's click event input on the chart. Then we can return the dataframe rows with the same bin values as the x-value of the click input."),
sliderInput("seq_from_to"
, label = h3("Sequence 'From' and 'To'")
, min = 0
, max = 500
, value = c(50, 350)
),
sliderInput("seq_by"
, label = h3("Sequence 'By'")
, min = 25
, max = 200
, value = 50
, step = 5)
),
mainPanel(
plotOutput("hist",
click = "plot_click",
dblclick = "plot_dblclick",
hover = "plot_hover",
brush = "plot_brush"),
dataTableOutput("table")
)
)
)
# Server function --------------------------------------------------
server <- function(input, output) {
# Render Histogram Plot
output$hist <- renderPlot({
# Using the same `qplot` function but inserting the user inputs to set the breaks values in the plot
pp <- qplot(dm
, geom = "histogram"
, breaks = seq(from = input$seq_from_to[1], to = input$seq_from_to[2], by = input$seq_by)
, xlab = "values"
, ylab = "Frequency"
, main = "Histogram"
, fill = I("red")
, col = I("black")
, alpha = I(0.4)
)
# Also using the user inputs to set the breaks values for the x-axis
pp + scale_x_continuous(breaks = seq(from = input$seq_from_to[1], to = input$seq_from_to[2], by = input$seq_by))
})
# This is purely explanatory to help show how shiny can read user interaction on qplot/ggplot objects
# It's taken from the Shiny docs here: https://shiny.rstudio.com/articles/plot-interaction.html
output$info <- renderText({
# Retain the x and y coords of the user click event data
xy_str <- function(e) {
if(is.null(e)) return("NULL\n")
paste0("x=", round(e$x, 1), " y=", round(e$y, 1), "\n")
}
# Retain the x and y range coords of click & drag (brush) data
xy_range_str <- function(e) {
if(is.null(e)) return("NULL\n")
paste0("xmin=", round(e$xmin, 1), " xmax=", round(e$xmax, 1),
" ymin=", round(e$ymin, 1), " ymax=", round(e$ymax, 1))
}
# Paste this together so we can read it in the UI function for demo purposes
paste0(
"click: ", xy_str(input$plot_click),
"dblclick: ", xy_str(input$plot_dblclick),
"hover: ", xy_str(input$plot_hover),
"brush: ", xy_range_str(input$plot_brush)
)
})
# Back to the story. Set a listener to trigger when one of the following is updated:
toListen <- reactive({list(
input$plot_click # user clicks on the plot
, input$seq_from_to # user updates the range slider
, input$seq_by # user updates the number input
)
})
# When one of those events are triggered, update the datatable output
observeEvent(toListen(), {
# Save the user click event data
click_data <- input$plot_click
print(click_data) # during your app preview, you can watch the R Console to see what click data is accessible
# Assign bin values to each row using the intervals that are set by the user input
df$bin <- findInterval(dm, vec = seq(from = input$seq_from_to[1], to = input$seq_from_to[2], by = input$seq_by))
# Similarly assign a bin value to the click event based on what interval the x values falls within
click_data$x_bin <- findInterval(click_data$x, vec = seq(from = input$seq_from_to[1], to = input$seq_from_to[2], by = input$seq_by))
# Lastly, subset the df to only those records within the same interval as the click event x-value
df_results <- subset(df, bin == click_data$x_bin)
# Select what values to view in the table
df_results <- df_results %>% select(model, hp)
# And push these back out to the UI
output$table <- renderDataTable(df_results,
options = list(
pageLength = 5
)
)
})
}
shinyApp(ui = ui, server = server)
Well, someone answered. Since I took the time to put it together, here is another potential solution.
library(shiny)
library(ggplot2)
library(scales)
library(dplyr)
library(DescTools) # added for Closest()
ui <- fluidPage(
titlePanel("Histogram"),
sidebarLayout(
sidebarPanel(
),
mainPanel(
plotOutput("hist", click = 'plot_click'), # added plot_click
verbatimTextOutput("x_value"), # added queues for interactivity
verbatimTextOutput("selected_rows") # added table for bin values
)
)
)
# this can be a dataframe or matrix for qplot or ggplot
# (not sure if there was another reason you had this code?)
# mtcars <- mtcars %>%
# select("hp") # if you only want hp
# mtcars <- as.matrix(mtcars) # I suggest making row names a column
# to keep 2 columns
pp <- ggplot(mtcars) +
geom_histogram(aes(x = hp),
bins = 10,
fill = "red",
color = "black",
alpha = .4) +
labs(x = "values",
y = "Frequency",
title = "Histogram")
# extract data from plot to find where each value falls within the histogram bins
# I kept the pkg name, function in more than one library
bd <- ggplot_build(ggplot2::last_plot())$data[[1]]
# add the assigned bin number to the mtcars frame; used for filtering matches
mtcars$bins <- lapply(mtcars$hp,
function(y) {
which(bd$x == Closest(bd$x, y))
}) %>% unlist()
server <- function(input, output) {
output$hist <- renderPlot({
# moved the plot outside of server, so that global variables could be created
# pp <- qplot(mtcars[,"hp"], geom = "histogram", bins = 10, xlab="values",
# ylab = "Frequency", main = "Histogram",
# fill = I("red"), col = I("black"), alpha = I(0.4))
# scale_x_continuous(breaks=pretty(mtcars, n=10)) # can't use this
pp
})
# # Print the name of the x value # added all that's below with server()
output$x_value <- renderPrint({
if (is.null(input$plot_click$x)) return()
# find the closest bin center to show where the user clicked on the histogram
cBin <- which(bd$x == Closest(bd$x, input$plot_click$x))
paste0("You selected bin ", cBin) # print out selected value based on bin center
})
# Print the rows of the data frame which match the x value
output$selected_rows <- renderPrint({
if (is.null(input$plot_click$x)) return()
# find the closest bin center to show where the user clicked on the histogram
cBin <- which(bd$x == Closest(bd$x, input$plot_click$x))
mtcars %>% filter(bins == cBin)
# mtcars
})
}
shinyApp(ui = ui, server = server)
Just in case someone ends in this post looking a way to include brushedPoints... inspired on this post, I found a way to do it!
Code:
#Load libraries ----------------------------------------------------
library(shiny)
library(ggplot2)
library(scales)
library(dplyr)
# Prepare data -----------------------------------------------------
df <- mtcars
df <- cbind(model = rownames(df), data.frame(df, row.names = NULL)) # setting the rownames as the first column
breaks_data = pretty(mtcars$hp, n=10)
my_breaks = seq(min(breaks_data), to=max(breaks_data), by=30)
# UI function ------------------------------------------------------
ui <- fluidPage(
titlePanel("Histogram"),
sidebarLayout(
sidebarPanel(
actionButton("draw_plot", "Draw the plot")
),
mainPanel(
plotOutput("hist",
brush = brushOpts("plot_brush", resetOnNew = T, direction = "x")),
dataTableOutput("table"),
)
)
)
# Server function --------------------------------------------------
server <- function(input, output) {
observeEvent(input$plot_brush, {
info_plot <- brushedPoints(df, input$plot_brush)
output$table <- renderDataTable(info_plot)
})
# If the user didn't choose to see the plot, it won't appear.
output$hist <- renderPlot({
df %>% ggplot(aes(hp)) +
geom_histogram(alpha=I(0.4), col = I("black"), fill = I("red"), bins=10) +
labs(x = "values",
y = "Frequency",
title = "Histogram") +
scale_x_continuous(breaks = my_breaks)
})
}
shinyApp(ui = ui, server = server)
How to do a scatterplot with hover
library(shiny)
library(tidyverse)
ui <- fluidPage(
titlePanel("hover tooltips demo"),
mainPanel(
plotOutput("plot1", hover = hoverOpts(id = "plot_hover", delay = 100, delayType = "debounce")),
uiOutput("hover_info") # , style = "pointer-events: none")
)
)
server <- function(input, output) {
output$plot1 <- renderPlot({
mtcars %>%
ggplot(aes(mpg, hp)) +
geom_point()
})
output$hover_info <- renderUI({
hover <- input$plot_hover
point <- shiny::nearPoints(mtcars,
coordinfo = hover,
xvar = 'mpg',
yvar = 'hp',
threshold = 20,
maxpoints = 1,
addDist = TRUE)
if (nrow(point) == 0) return(NULL)
style <- paste0("position:absolute; z-index:100; background-color: #3c8dbc; color: #ffffff;",
"font-weight: normal; font-size: 11pt;",
"left:", hover$coords_css$x + 5, "px;",
"top:", hover$coords_css$y + 5, "px;")
wellPanel(
style = style,
p(HTML(paste0("Some info about car: <br/>MPG ", point$mpg, "<br/>HP ", point$hp)))
)
})
}
shinyApp(ui = ui, server = server)
I am setting up a small shiny app where I do not want the plot to change unless the action button is clicked. In the example below, when I first run the app, there is no plot until I click the action button. However, if I then change my menu option in the drop-down from Histogram to Scatter, the scatter plot is automatically displayed even though the value for input$show_plot has not changed because the action button has not been clicked.
Is there a way that I can change my menu selection from Histogram to Scatter, but NOT have the plot change until I click the action button? I've read through several different posts and articles and can't seem to get this worked out.
Thanks for any input!
ui.R
library(shiny)
fluidPage(
tabsetPanel(
tabPanel("Main",
headerPanel(""),
sidebarPanel(
selectInput('plot_type', 'Select plot type', c('Histogram','Scatter'), width = "250px"),
actionButton('show_plot',"Plot", width = "125px"),
width = 2
),
mainPanel(
conditionalPanel(
"input.plot_type == 'Histogram'",
plotOutput('plot_histogram')
),
conditionalPanel(
"input.plot_type == 'Scatter'",
plotOutput('plot_scatter')
)
))
)
)
server.R
library(shiny)
library(ggplot2)
set.seed(10)
function(input, output, session) {
### GENERATE SOME DATA ###
source_data <- reactive({
mydata1 = as.data.frame(rnorm(n = 100))
mydata2 = as.data.frame(rnorm(n = 100))
mydata = cbind(mydata1, mydata2)
colnames(mydata) <- c("value1","value2")
return(mydata)
})
# get a subset of the data for the histogram
hist_data <- reactive({
data_sub = as.data.frame(source_data()[sample(1:nrow(source_data()), 75), "value1"])
colnames(data_sub) <- "value1"
return(data_sub)
})
# get a subset of the data for the scatter plot
scatter_data <- reactive({
data_sub = as.data.frame(source_data()[sample(1:nrow(source_data()), 75),])
return(data_sub)
})
### MAKE SOME PLOTS ###
observeEvent(input$show_plot,{
output$plot_histogram <- renderPlot({
isolate({
plot_data = hist_data()
print(head(plot_data))
p = ggplot(plot_data, aes(x = value1, y = ..count..)) + geom_histogram()
return(p)
})
})
})
observeEvent(input$show_plot,{
output$plot_scatter <- renderPlot({
isolate({
plot_data = scatter_data()
print(head(plot_data))
p = ggplot(plot_data, aes(x = value1, y = value2)) + geom_point()
return(p)
})
})
})
}
Based on your desired behavior I don't see a need for actionButton() at all. If you want to change plots based on user input then the combo of selectinput() and conditionPanel() already does that for you.
On another note, it is not good practice to have output bindings inside any reactives. Here's an improved version of your server code. I think you are good enough to see notice the changes but comment if you have any questions. -
function(input, output, session) {
### GENERATE SOME DATA ###
source_data <- data.frame(value1 = rnorm(n = 100), value2 = rnorm(n = 100))
# get a subset of the data for the histogram
hist_data <- reactive({
# reactive is not needed if no user input is used for creating this data
source_data[sample(1:nrow(source_data), 75), "value1", drop = F]
})
# get a subset of the data for the histogram
scatter_data <- reactive({
# reactive is not needed if no user input is used for creating this data
source_data[sample(1:nrow(source_data), 75), , drop = F]
})
### MAKE SOME PLOTS ###
output$plot_histogram <- renderPlot({
req(hist_data())
print(head(hist_data()))
p = ggplot(hist_data(), aes(x = value1, y = ..count..)) + geom_histogram()
return(p)
})
output$plot_scatter <- renderPlot({
req(scatter_data())
print(head(scatter_data()))
p = ggplot(scatter_data(), aes(x = value1, y = value2)) + geom_point()
return(p)
})
}
I have the dataframe below:
col1<-sample(500, size = 500, replace = TRUE)
col2<-sample(500, size = 500, replace = TRUE)
d<-data.frame(col1,col2)
And I create a histogram of this data frame that has click-event activated. When the user clicks on a bar the rows of the dataframe that have the relative value are displayed in a datatable. The problem is that the app works fine with a few values. If for example my dataframe had 5 rows instead of 500 with :
col1<-sample(5, size = 5, replace = TRUE)
col2<-sample(5, size = 5, replace = TRUE)
d<-data.frame(col1,col2)
But with more values the app does not work since the plotly gives a range of values in every single bar instead of a unique value.
library(plotly)
library(shiny)
library(DT)
ui <- fluidPage(
mainPanel(
plotlyOutput("heat")
),
DT::dataTableOutput('tbl4')
)
server <- function(input, output, session) {
output$heat <- renderPlotly({
render_value(d) # You need function otherwise data.frame NN is not visible
p <- plot_ly(x = d$col2, type = "histogram",source="subset") # set source so
# that you can get values from source using click_event
})
render_value=function(NN){
output$tbl4 <- renderDataTable({
s <- event_data("plotly_click",source = "subset")
print(s)
return(DT::datatable(d[d$col2==s$y,]))
})
}
}
shinyApp(ui, server)
You can try this (added code to capture the count). You need to plot a histogram of count and then you can able to get your original data based on click event.
library(plotly)
library(shiny)
library(DT)
library(dplyr)
ui <- fluidPage(
mainPanel(
plotlyOutput("heat")
),
DT::dataTableOutput('tbl4')
)
server <- function(input, output, session) {
output$heat <- renderPlotly({
col1<-sample(500, size = 500, replace = TRUE)
col2<-sample(500, size = 500, replace = TRUE)
d<-data.frame(col1,col2)
d=d %>%
group_by(col2) %>%
mutate(count = n()) # You can programatically add count for each row
render_value(d) # You need function otherwise data.frame NN is not visible
p <- plot_ly(x = d$count, type = "histogram",source="subset")
# You should histogram of count
# set source so that you can get values from source using click_event
})
render_value=function(d){
output$tbl4 <- renderDataTable({
s <- event_data("plotly_click",source = "subset")
print(s)
return(DT::datatable(d[d$count==s$x,]))
})
}
}
shinyApp(ui, server)
Screenshot from the working prototype:
I would like to prevent reseting of the curve while choosing second variable to display with selectizeInput. For example in the code below we choose one value (mtcars dataset) in selectizeInput of cyl (6), and exclude one point from the curve, then we choose second value of cyl (4) to display, and therefore the previous curve with cyl=6, resets itself (the point which has been excluded, appears again).
Is there any way to prevent this behaviour and while choosing second variable the excluded point stays "excluded"?
Example code:
library(ggplot2)
library(shiny)
ui <- fluidPage(
fluidRow(
column(width = 6,
plotOutput("plot1", height = 350,
click = "plot1_click"),
selectizeInput("valuecyl", "Select value of cyl:", choices=unique(mtcars$cyl), multiple = TRUE))
)
)
server <- function(input, output) {
# For storing which rows have been excluded
vals <- reactiveValues()
data_df <- reactive({
data <- mtcars
data <- data[data$cyl %in% input$valuecyl, ]
vals$keeprows = rep(TRUE, nrow(data))
data
})
output$plot1 <- renderPlot({
data<- data_df()
keep <- data[ vals$keeprows, , drop = FALSE]
exclude <- data[!vals$keeprows, , drop = FALSE]
print(keep)
ggplot(keep, aes(wt,mpg,colour=as.factor(cyl))) + geom_point(data=keep) + geom_line(data=keep) +
geom_point(data = exclude, shape = 21, fill = NA, color = "black", alpha = 0.25)
})
# Toggle points that are clicked
observeEvent(input$plot1_click, {
data <- data_df()
res <- nearPoints(data, input$plot1_click, allRows = TRUE)
vals$keeprows <- xor(vals$keeprows, res$selected_)
})
}
shinyApp(ui, server)
The issue here is that you are overwriting vals$keeprows and replacing it with rep(TRUE, nrow(data)) every time the user selects a value in your selectize.
You need to update the keeprows variable by keeping the rows that were kept by the user and adding in the new rows coming from the extra selection.
To do that, I slightly modified your code:
#added the data in the reactiveValues for convenience
vals <- reactiveValues(keeprows=logical(0),data=mtcars[0,])
#this observes the input and updates the data when the user adds a cyl value
observeEvent(input$valuecyl,{
#get the id (here rownames) of the points excluded by the user
excluded_ids <- rownames(vals$data)[!vals$keeprows]
#make the new data
vals$data=mtcars[mtcars$cyl %in% input$valuecyl,]
#keep the rows that the user had not previously excluded.
vals$keeprows = !(rownames(vals$data) %in% excluded_ids)
})
Since I added the data in the reactiveValues and removed data_df, you need to replace data_df() by vals$data in your code for this to work.
Could anyone can tell me why I get an error when I change a dataset in first selectInput widget? When I change a dataset from diamonds to mtcars I get an error Could not find 'carat' in input$bins and in the plot just for one second and after that everything works fine. Why it happened?
library(shiny)
library(ggplot2)
data(diamonds)
data(mtcars)
ui <- fluidPage(
column(3,
selectInput("data", "", choices = c('mtcars', 'diamonds')),
uiOutput('server_cols'),
uiOutput('server_bins')
),
column(9,
plotOutput("plot")
)
)
server <- function(input, output) {
data <- reactive({
switch(input$data,
diamonds = diamonds,
mtcars = mtcars)
})
output$server_cols <- renderUI({
data <- data()
nam <- colnames(data)
selectInput('cols', "Choose numeric columns:", choices = nam[sapply(data, function(x) is.numeric(x))])
})
output$server_bins <- renderUI({
if (!is.null(input$cols)) {
df <- data()
x <- eval(input$cols)
max_value <- max(df[,x])
sliderInput('bins','Choose number of bins:', min = 0.1,
max = max_value,
value = max_value/2)
}
})
output$plot <- renderPlot({
if (!is.null(input$cols) & !is.null(input$bins)) {
basicData <- data()
var <- eval(input$cols)
ggplot(basicData, aes_string(var)) +
geom_histogram(binwidth = input$bins, color = 'white', fill = 'red')
}
})
}
shinyApp(ui, server)
Your respective output objects respond to any changes of your input variables. Thus, when you change your dataset via input$data, the plot rebuilds itself, although input$cols did not yet adjust. Actually, try inserting some print("a") inside the output$plot to see that it is called up to three times if you change input$data.
The fix is to rethink your reaction logic and let your elements respond only to specific changes, to get some kind of response "thread".
For example, input$data should only trigger output$server_cols. And output$server_bins should only be triggered by input$cols (because this already implies that input$data changed earlier). Ultimately, output$plot just has to listen to changes of input$bins (because changes in input$cols and input$data always result in changes of input$bins since it is at the end of the thread).
Here is my suggestion using isolate.
library(shiny)
library(ggplot2)
data(diamonds)
data(mtcars)
ui <- fluidPage(
column(3,
selectInput("data", "", choices = c('mtcars', 'diamonds')),
uiOutput('server_cols'),
uiOutput('server_bins')
),
column(9,
plotOutput("plot")
)
)
server <- function(input, output) {
data <- reactive({
switch(input$data, diamonds = diamonds, mtcars = mtcars)
})
output$server_cols <- renderUI({
data <- data()
nam <- colnames(data)
selectInput('cols', "Choose numeric columns:", choices = nam[sapply(data, function(x) is.numeric(x))])
})
output$server_bins <- renderUI({
if (!is.null(input$cols)) {
df <- isolate(data())
x <- eval(input$cols)
max_value <- max(df[,x])
sliderInput('bins','Choose number of bins:', min = 0.1, max = max_value, value = max_value/2)
}
})
output$plot <- renderPlot({
if (!is.null(isolate(input$cols)) & !is.null(input$bins)) {
basicData <- isolate(data())
var <- eval(isolate(input$cols))
ggplot(basicData, aes_string(var)) +
geom_histogram(binwidth = input$bins, color = 'white', fill = 'red')
}
})
}
shinyApp(ui, server)
You might also want to look into updateSelectInput and updateSliderInput if you want to alter Input Elements depending on other input.