Data-visualization-in-R

Some elements of data visualization with R 'ggplot2' mostly

Examples include scatterplots, histograms, kernel density estimators (KDE), time series plots, bar plots, diverging bar plots, pareto plots... It also includes code for multiple plots on the same window. Here is an excerpt of some of the visualizations taken as examples.

Examples

1. Plots of time data

Time series plot (numerical x time variable)

ggplot(dataset, aes(x = dates, y = value)) +
  geom_line() + 
  customization

Diverging bar plot (eg. numerical x time variable)

ggplot(dataset, aes(x = dates, y = value)) +
  geom_bar() + 
  customization

Plot with values above a threshold (eg. numerical x time variable)

ggplot(dataset, aes(x = dates, y = value)) +
  geom_point() +                                                                         
  geom_text(aes(label = ID), dataset %>% filter(y>mean(y)), show_guide  = FALSE) +                           
  customization

2. Multiple plots

Scatterplot with marginal densities and regression lines by group (eg. numerical x numerical, by group)

library(cowplot)

Scatterplot with different pointsize

ggplot(data=dataset, aes(x=variable1, y=variable2, colour = group))+
  geom_point(aes(size = variable3)) +
  customization

Plot multiples subplots (eg. mixed of numerical and categorical/nominal variables

p1 = ggplot(data=dataset) +
  customization
  
p2 = ggplot(data=dataset) +
  customization

grid.arrange(p1, p2)

Plot of histogram with kernel density estimator (eg. numerical by group)

ggplot(dataset) +
  geom_histogram() +
  geom_density() +
  customization

Ridge plots (eg. numerical by group)

ggplot(iris, aes(x = value, y = group, fill = group)) +
  geom_density_ridges()

Plot joint distribution of two categorical or nominal variages (eg. nominal x categorical)

ggplot(dataset, aes(x = variable1, y = prop, width = cut.count, fill = variable2)) +
  geom_bar() +
  facet_grid(~ variable1, scales = "free_x", space = "free_x") +
  customization

3. Statistics

Plot a distribution with colored or shaded area

ggplot(aes(dataset)) +
  as_reference(geom_density(), id = "density") +
  with_blend(annotate("), bg_layer = "density", blend_type = "atop") +
  customization

Plot multiple theoretical distributions on the same plot (eg. three Normal densities)

ggplot(data=dataset) +
  stat_function(fun=dnorm, args=list(mean = 0, sd = 1)) +
  customization

Convergence of MC estimators (eg. numerical x integer sequence)

ggplot(dataset, aes( x = seq)) +
  geom_line( aes(y = cumsum(x1)/(1:length(x1)))) +
  customization

And much more.