| title | authors | date | bibliography | nocite | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
metricsgraphics: An R htmlwidget interface to the MetricsGraphics.js D3-based visualization library |
|
2016-04-24 |
paper.bib |
@R-base, @R-htmltools, @R-htmlwidgets, @R-jsonlite, @R-yaml, @metrics-graphics.js, @Bostock:2011:DDD:2068462.2068631
|
MetricsGraphics.js is a library built on top of D3 that is optimized for visualizing and laying out time-series data. It provides a simple way to produce common types of graphics in a principled, consistent and responsive way. The library currently supports line charts, scatterplots and histograms as well as features like rug plots and basic linear regression.
The metricsgraphics R package wraps the MetricsGraphics.js library into an htmlwidget, making it easy to use in R Markdown documents and Shiny applications.
Plots are composed first by calling mjs_plot() with the data.frame containing x and y values to plot. The x and y paramters may be bare or quoted column names. This object can be passed or piped to one of the "geom"-like functions to product the desired chart. The package provides support for:
- (multiple) line charts with
mjs_line()andmjs_add_line() - bar charts with
mjs_bar() - histograms with
mjs_histogram() - scatterplots with
mjs_point()
Additional functions for annotations and combining plots are also provided by the package along with the ability to turn simple (e.g. singe Geom) ggplot2 plots automatically into corresponding metricsgraphics objects using as_mjsplot().