profvis: Interactive Visualizations for Profiling R Code

Description

Interactive visualizations for profiling R code.

Author(s)

Maintainer: Hadley Wickham hadley@posit.co

Authors:

• Winston Chang

• Javier Luraschi

• Timothy Mastny

Other contributors:

• Posit Software, PBC [copyright holder, funder]

• jQuery Foundation (jQuery library) [copyright holder]

• jQuery contributors (jQuery library; authors listed in inst/htmlwidgets/lib/jquery/AUTHORS.txt) [contributor, copyright holder]

• Mike Bostock (D3 library) [contributor, copyright holder]

• D3 contributors (D3 library) [contributor]

• Ivan Sagalaev (highlight.js library) [contributor, copyright holder]

See Also

Useful links:

• https://profvis.r-lib.org

• https://github.com/r-lib/profvis

• Report bugs at https://github.com/r-lib/profvis/issues

Parse Rprof output file for use with profvis

Description

Parse Rprof output file for use with profvis

Usage

parse_rprof(path = "Rprof.out", expr_source = NULL)

Arguments

Pause an R process

Description

This function pauses an R process for some amount of time. It differs from Sys.sleep() in that time spent in pause will show up in profiler data. Another difference is that pause uses up 100\ whereas Sys.sleep does not.

Usage

pause(seconds)

Arguments

Examples

# Wait for 0.5 seconds
pause(0.5)

Print a profvis object

Description

Print a profvis object

Usage

## S3 method for class 'profvis'
print(x, ..., width = NULL, height = NULL, split = NULL, aggregate = NULL)

Arguments

Profile an R expression and visualize profiling data

Description

This function will run an R expression with profiling, and then return an htmlwidget for interactively exploring the profiling data.

Usage

profvis(
  expr = NULL,
  interval = 0.01,
  prof_output = NULL,
  prof_input = NULL,
  timing = NULL,
  width = NULL,
  height = NULL,
  split = c("h", "v"),
  torture = 0,
  simplify = TRUE,
  rerun = FALSE
)

Arguments

Details

An alternate way to use profvis is to separately capture the profiling data to a file using Rprof(), and then pass the path to the corresponding data file as the prof_input argument to profvis().

See Also

print.profvis() for printing options.

Rprof() for more information about how the profiling data is collected.

Examples

# Only run these examples in interactive R sessions
if (interactive()) {

# Profile some code
profvis({
  dat <- data.frame(
    x = rnorm(5e4),
    y = rnorm(5e4)
  )

  plot(x ~ y, data = dat)
  m <- lm(x ~ y, data = dat)
  abline(m, col = "red")
})


# Save a profile to an HTML file
p <- profvis({
  dat <- data.frame(
    x = rnorm(5e4),
    y = rnorm(5e4)
  )

  plot(x ~ y, data = dat)
  m <- lm(x ~ y, data = dat)
  abline(m, col = "red")
})
htmlwidgets::saveWidget(p, "profile.html")

# Can open in browser from R
browseURL("profile.html")

}

Widget output and renders functions for use in Shiny

Description

Widget output and renders functions for use in Shiny

Usage

profvisOutput(outputId, width = "100%", height = "600px")

renderProfvis(expr, env = parent.frame(), quoted = FALSE)

Arguments

profvis UI for Shiny Apps

Description

Use this Shiny module to inject profvis controls into your Shiny app. The profvis Shiny module injects UI that can be used to start and stop profiling, and either view the results in the profvis UI or download the raw .Rprof data. It is highly recommended that this be used for testing and debugging only, and not included in production apps!

Usage

profvis_ui(id)

profvis_server(input, output, session, dir = ".")

Arguments

Details

The usual way to use profvis with Shiny is to simply call profvis(shiny::runApp()), but this may not always be possible or desirable: first, if you only want to profile a particular interaction in the Shiny app and not capture all the calculations involved in starting up the app and getting it into the correct state; and second, if you're trying to profile an application that's been deployed to a server.

For more details on how to invoke Shiny modules, see this article.

Examples

# In order to avoid "Hit <Return> to see next plot" prompts,
# run this example with `example(profvis_ui, ask=FALSE)`

if(interactive()) {
  library(shiny)
  shinyApp(
    fluidPage(
      plotOutput("plot"),
      actionButton("new", "New plot"),
      profvis_ui("profiler")
    ),
    function(input, output, session) {
      callModule(profvis_server, "profiler")

      output$plot <- renderPlot({
        input$new
        boxplot(mpg ~ cyl, data = mtcars)
      })
    }
  )
}