In aczepielik/ggKNMF: GGplot Extnension for KNMF

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

Package ggKNMF is a ggplot2 extension offering new style designed to be of consistent with the Financial Mathematics Student Association of the Jagiellonian University visual identification. In this introduction we will see how it works and how to use it.

ggKNMF overwrites some of default ggplot style right when loaded. Let us plot some example charts with vasic ggplot and see how the will look like after loading ggKNMF.

library(ggplot2)

scatter_plot <- ggplot(iris, aes(Petal.Length, Petal.Width)) + 
  geom_point() +
  geom_smooth(method = 'lm') +
  labs(title = "Scatter plot", subtitle = "with one regression line")

scatter_plot_many_lines <- ggplot(iris, aes(Petal.Length, Petal.Width, colour = Species)) + 
  geom_point() +
  geom_smooth(method = 'lm') +
  labs(title = "Scatter plot", subtitle = "with line for each category")

density_plot <- ggplot(iris, aes(Petal.Length, fill = Species)) +
  geom_density() +
  labs(title = "Density plot", subtitle = "coloured by category")

scatter_plot
scatter_plot_many_lines
density_plot

Now, let us load ggKNMF and see those plots restyled

library(ggKNMF)
scatter_plot
scatter_plot_many_lines
density_plot

Several things has changed. We will discuss it one by one, showing how to restore default behaviour in each aspect and how to use specyfic features in a standalone manner (without being loaded as default)

theme_knmf()

Themes in ggplots describe look and behaviour of axes, grids, legend, facets and all labels associated with them. Function theme_knmf() set with package loading as a default theme uses thin, solid axis lines, dashed major grid lines, no minor grid lines, white background, and modified string display style.

When ggKNMF is loaded theme_knmf is loaded previous theme is attached to variable old_theme. To set it back as default one can use

Function theme_knmf can be used explicitely to add this theme to plot

# when package ggKNMF is loaded
ggplot(iris, aes(Petal.Length, Petal.Width)) + 
  geom_point() +
  theme_knmf()

#when package ggKNMF is not loaded
ggplot(iris, aes(Petal.Length, Petal.Width)) + 
  geom_point() +
  ggKNMF::theme_knmf()

Colour scheme

Package ggKNMF comes with eight-colour palette which is used to construct colour scales. Their RGB definitions can be seen by prompting knmf_colours. That si how they look like:

Polychrome::swatch(ggKNMF::knmf_colours)

When an discrete-valued aesthetic is mapped onto colour or fill, colours from this palette will be displayed. When continuous (or ordered factor) aesthetic is mappend onto colour and fill one of two oredered, multicolour scales is used: "red" varying from black through red to yellow or "blue" varying from dark grey, through light blue to light yellow:

Polychrome::swatch(ggKNMF:::monochrome_palette("red")(5))
Polychrome::swatch(ggKNMF:::monochrome_palette("red")(15))

Polychrome::swatch(ggKNMF:::monochrome_palette("blue")(5))
Polychrome::swatch(ggKNMF:::monochrome_palette("blue")(15))

Default conitnuous (or ordinal) scale is "red". To change it to "blue" explict usage of scale_colour_continuous_knmf is necessary.

There are two more scales for colour or fill aesthetics in the package. With scale_colour_sequential one can map continuos values to colours varying from colour selected among knmf_colours to light grey. With scale_colour_divergent colours vary from specified midpoint in to directions up to two specified colours.

Let us restore package colour scheme with use_knmf_colours() function nad see colour and fill scales in action

use_knmf_colours()
theme_set(theme_knmf())

ggplot(faithfuld, aes(eruptions, waiting, fill = density)) + 
  geom_tile() +
  labs(title = "Red fill scale (default)")

ggplot(faithfuld, aes(eruptions, waiting, fill = density)) + 
  geom_tile() + scale_fill_continuous_knmf(colour = "blue") +
  labs(title = "Blue fill scale scale")