Create Stacked Density Plots in ggplot2 to Visualize Sunsplot Activities
In this article, we’ll leverage the ggridges package by Claus O. Wilke to visualize the sunspot activities using a list of density plots stacked in chronological order.
The dataset sunspot.month (time series) used is built-in base R.
library(ggplot2)library(dplyr)library(tidyr)library(TSstudio) # transform time series datalibrary(ggridges) # draw stacked density plots s <- sunspot.month %>%# convert time series to data.framets_reshape() %>%# tidy uppivot_longer(-month, names_to ="year", values_to ="activity") %>%mutate(year =as.integer(year)) %>%filter(year %in%seq(1750, 2010, 2)) head(s, n =3)
p1 <- s %>%ggplot(aes(x = activity, y = year, fill =after_stat(x))) +# create density plots in stacked mannergeom_density_ridges_gradient(aes(group = year), scale =8, color ="white")p1
Adjust the color scale, axis scale, and titles.
p2 <- p1 +scale_fill_viridis_c(option ="C") +scale_y_continuous(breaks =seq(1750, 2010, 10), name =NULL) +coord_cartesian(xlim =c(0, NA), expand =0) +labs(x ="Activity intensity") +# add plot titleannotate(geom ="text", x =220, y =1790,label ="Sunspot\n yearly\n activity\n distribution\n\n1740 ~ 2013", fontface ="bold", size =4, color ="tomato3")p2
library(ggplot2)library(dplyr)library(tidyr)library(TSstudio) # transform time series datalibrary(ggridges) # draw stacked density plots s <- sunspot.month %>%# convert time series to data.framets_reshape() %>%# tidy uppivot_longer(-month, names_to ="year", values_to ="activity") %>%mutate(year =as.integer(year)) %>%filter(year %in%seq(1750, 2010, 2)) head(s, n =3) # ready for visualization! # Create a list of stacked density plots. p1 <- s %>%ggplot(aes(x = activity, y = year, fill =after_stat(x))) +# create density plots in stacked mannergeom_density_ridges_gradient(aes(group = year), scale =8, color ="white")p1 # Adjust the color scale, axis scale, and titles. p2 <- p1 +scale_fill_viridis_c(option ="C") +scale_y_continuous(breaks =seq(1750, 2010, 10), name =NULL) +coord_cartesian(xlim =c(0, NA), expand =0) +labs(x ="Activity intensity") +# add plot titleannotate(geom ="text", x =220, y =1790,label ="Sunspot\n yearly\n activity\n distribution\n\n1740 ~ 2013", fontface ="bold", size =4, color ="tomato3")p2 # A final fine-tune of the theme. p3 <- p2 +theme_classic() +theme(panel.grid =element_blank(),axis.line =element_blank(),legend.position ="none",axis.ticks.y =element_line(color ="snow2"),axis.text.y =element_text(margin =margin(r =10)),axis.title.x =element_text(margin =margin(t =10))) p3
To visualize complicated chronological data like the sunspot activities, in addition to the density plot, heatmaps and line plot, animation is another powerful approach. Check out the following annual population pyramids in animation that vividly unveils the predicted changing dynamics in the population structure in Germany.
