Learn R

GGPlot2

• ggplot2 is universally considered the best package for general graphing and is part of the tidyverse (see Data Management).
• The function name is ggplot() - without the 2.
• There is some graphing in base r that can be quick and easy for exploration, but few people use lattice any more.

Useful ggplot-based packages:

• GGally - Good for bivariate graphing with functions like ggpairs()
• ggformula - uses formula notation and ggplot for ease of teaching
• SmartEDA - functions for all types of exploratory data analysis
• DataExplorer - easy functions for data cleaning and exploration
• ggpubr - ‘ggplot2’ Based Publication Ready Plots

Point-and-Click Options

• esquiesse - Best to create a single graph from cleaned/prepared data and/or to learn ggplot syntax.
  • Tableau-like workflow helps build and refine a single graph using drag and drop. Automatically chooses chart types based on data types.
  • Produces nice, clean code.
  • Allows for specifying variable types on import (e.g., create factors), transforming values (e.g., log), and filtering by value.
• ggquickeda - Best if you have categorical variables and would like to modify them before creating visuals.
  • Only imports csv files within the interface, or can specify dataset when starting.
  • Includes many tools to modify categorical data, like creating factors and merging groups.
  • Many options to create plots, though the interface is cluttered and not always intuitive. Can also create Descriptive Statistics tables.
• ExPanDar - Best for exploratory use.
  • Originally for time series, but datasets with any unique numeric variable will work. Also works with SPSS files.
  • Automatically produces a variety of EDA charts and tables (e.g., regression) which look nice, but the appearance is not customizable.
  • You can easily change variables, filter cases, and group by factor levels.
  • Can remove some outliers and create new variables through formulas.
  • Can save and restore settings, and can right click to save images (they will be png). No code is produced.

Interactive Graphics

• ggvis – make interactive graphics using ggplot2
• shiny – make interactive graphical applications using RStudio
  • Shiny Tutorials for Statistics Instructors (Show Me Shiny)