ian-curtis / gvsu215

Pretty output for select {mosaic} and {ggformula} functions. Designed for the GVSU Introductory Statistics course (STA 215).

gvsu215

Pretty Plots and Charts for Introductory Statistical Concepts

Note: Due to the fact that some functions return tables (which are essentially HTML code), the README on GitHub does not display properly. As a result, the output is an image of the resulting table rather than the actual HTML table. These images will not include the table caption at this time.

gvsu215 contains a series of wrapper functions around select {mosaic} and {ggformula} functions, with help from {flextable} and {ggplot2}. The goal is to provide introductory statistics students/learners with an easy way to generate nice-looking plots and tables without focusing on the code used to create them. All tables and plots generated from this package have been doctored and have had themes pre-applied. As such, output will not print in “messy” format to the R console and can be used interactively, including in R Markdown documents.

This package was designed and tailored for Grand Valley State University (course code: STA 215) but can be used by anyone.

Installation

This package is not on CRAN. You can install the development version of gvsu215 from GitHub with:

# install.packages("devtools")
devtools::install_github("ian-curtis/gvsu215")

Prerequisites

In order to use this package, you will need a basic understanding of R syntax and how to run a function in your preferred working environment. Functions use formula syntax (e.g., var1 ~ var2) and thus knowledge of formulas is needed as well.

Examples

This package is best used for creating simple plots and charts where extensive customization is not necessary. For instance, you may just want to see summary statistics, including one- or two-way tables.

library(gvsu215)

tbl_2var(mtcars, gear~cyl)

tbl_num_sum(mtcars, ~wt, na_rm = TRUE)

Or maybe you just need percentiles:

tbl_pctile(mtcars, ~wt, probs = c(0, .25, .4, .5, .6, .75, 1))

There is also support for a variety of plots…

plot_bar(mtcars, ~cyl, type = "percent")

plot_box(mtcars, 
         wt~gear, 
         fill = 'orangered4', 
         x = "Gear", 
         y = "Weight", 
         title = "Boxplot of Weight by Gear")

plot_scatter(mtcars, wt~drat)
#> NAs always removed (in pairs) for scatterplots.

… and statistical inference.

infer_1prop(mtcars, ~vs, success = 1)

infer_paired(mtcars, var1 = ~drat, var2 = ~wt)

infer_2mean_int(mtcars, wt~vs)

suppressWarnings(infer_chisq(mtcars, cyl~gear))

suppressWarnings(infer_chisq(mtcars, cyl~gear, type = "expected"))

infer_anova(mtcars, cyl~gear, digits = 2)

Help, Bug Reports, and Feature Requests

Errors happen! If you come across an error you don’t think you should have, feel free to create an issue on GitHub. If you’d like to suggest a change or addition, you can use the issues page for that, too!

gvsu215's Issues

Some plots generated from `plot_*()` give low quality images.

Perhaps I can specify a default width, height, and dpi? Will this take away from users saving their own files with their own specifications?

LaTeX does not like `tbl_corr()` subscript characters

Will want to change these characters into the variable name instead.