Displaying Distributions of Continuous Data With 5 Different Charts

Histogram, Density, Violin, Boxplot, Ridgeline.

Borsa Istanbul calculates variety of indices to follow the movements of markets, to be an underlying asset for the financial products, and to be used as a benchmark for collective investment instruments. Click here for further information.

You can access the data by downloading post21.xlsx file on my GitHub account. The 12 BIST Indices were chosen by me. You can find all of BIST indices here. Reuters data is used.

library(tidyverse)

df <- readxl::read_excel("data.xlsx") %>% 
  select(-DATE) %>% 
  mutate_all(
    ., function(x) (lead(x)/x-1)*100
  ) %>% 
  na.omit()

head(df)

XU030	XU050	XU100	XUTUM	XBANK	XBLSM	XGIDA	XHOLD	XTRZM	XULAS	XUTEK	XUSIN
0.51	0.53	0.67	0.65	0.85	-0.22	0.03	1.02	-0.01	-1.03	0.29	1.02
0.91	0.88	0.95	0.97	2.58	0.97	1.35	-0.04	2.61	-0.01	0.74	0.98
-0.09	-0.02	0.14	0.25	-0.36	0.68	1.07	0.13	1.93	1.55	-0.20	0.20
0.30	0.33	0.34	0.41	0.48	0.96	0.16	0.94	0.81	-0.01	0.56	0.32
-0.88	-0.75	-0.71	-0.49	-0.55	0.35	0.29	-0.76	-0.66	-0.11	0.09	-0.36
0.36	0.36	0.41	0.47	0.97	0.68	0.63	0.41	0.70	0.78	0.32	0.44

df <- df %>% 
  mutate(t = seq(1,nrow(.),1), .before = 1) %>% 
  pivot_longer(!t, names_to = "vars", values_to = "vals")

Histogram and Density

If there are two variables:

df %>% 
  filter(vars %in% c("XU100","XBANK")) %>% 
  ggplot(aes(x = vals, fill = vars)) +
  geom_histogram(position = "identity", alpha = .5) +
  theme_minimal() +
  theme(axis.title = element_blank(),
        axis.text.y = element_blank(),
        legend.title = element_blank(),
        legend.position = "top") +
  scale_fill_manual(values = c("red","blue"))

df %>% 
  filter(vars %in% c("XU100","XBANK")) %>% 
  ggplot(aes(x = vals, fill = vars)) +
  geom_density(alpha = .5) +
  theme_minimal() +
  theme(axis.title = element_blank(),
        axis.text.y = element_blank(),
        legend.title = element_blank(),
        legend.position = "top") +
  scale_fill_manual(values = c("red","blue"))

df %>% 
  filter(vars %in% c("XU100","XBANK")) %>% 
  pivot_wider(names_from = "vars", values_from = "vals") %>% 
  ggplot(aes(x = vals)) +
  geom_histogram(aes(x = XU100, y = ..density..), fill = "red" ) +
  geom_label(aes(x = 4, y = 0.25, label = "XU100"), color = "red") +
  geom_histogram(aes(x = XBANK, y = -..density..), fill = "blue") +
  geom_label(aes(x = 4, y = -0.25, label = "XBANK"), color = "blue") +
  theme_minimal() +
  theme(axis.title = element_blank(),
        axis.text.y = element_blank())

df %>% 
  filter(vars %in% c("XU100","XBANK")) %>% 
  pivot_wider(names_from = "vars", values_from = "vals") %>% 
  ggplot(aes(x = vals)) +
  geom_density(aes(x = XU100, y = ..density..), fill = "red" ) +
  geom_label(aes(x=4, y=0.25, label = "XU100"), color="red") +
  geom_density(aes(x = XBANK, y = -..density..), fill = "blue") +
  geom_label(aes(x=4, y = -0.25, label = "XBANK"), color="blue") +
  theme_minimal() +
  theme(axis.title = element_blank(),
        axis.text.y = element_blank())

If there are more than two variables:

df %>% 
  filter(vars %in% c("XU100","XBANK","XBLSM")) %>% 
  ggplot(aes(x = vals, color = vars)) +
  geom_density(lwd = 1.5) +
  theme_minimal() +
  theme(axis.title = element_blank(),
        axis.text.y = element_blank(),
        legend.title = element_blank(),
        legend.position = "top")

df %>% 
  ggplot(aes(x = vals, fill = vars)) +
  geom_histogram(position = "identity") +
  theme_minimal() +
  theme(axis.title = element_blank(),
        axis.text.y = element_blank(),
        legend.position = "none") +
  facet_wrap(~vars)

df %>% 
  ggplot(aes(x = vals, fill = vars)) +
  geom_density() +
  theme_minimal() +
  theme(axis.title = element_blank(),
        axis.text.y = element_blank(),
        legend.position = "none") +
  facet_wrap(~vars)

Violin and Boxplot

df %>% 
  ggplot(aes(x = vars, y = vals, fill = vars)) +
  geom_violin() +
  theme_minimal() +
  theme(axis.title = element_blank(),
        legend.position = "none")

df %>% 
  ggplot(aes(x = vars, y = vals, fill = vars)) +
  geom_violin() +
  theme_minimal() +
  theme(axis.title = element_blank(),
        legend.position = "none") +
  coord_flip()

df %>% 
  ggplot(aes(x = vars, y = vals, fill = vars)) +
  geom_boxplot() +
  theme_minimal() +
  theme(axis.title = element_blank(),
        legend.position = "none")

df %>% 
  ggplot(aes(x = vars, y = vals, fill = vars)) +
  geom_boxplot() +
  theme_minimal() +
  theme(axis.title = element_blank(),
        legend.position = "none") +
  coord_flip()

df %>% 
  ggplot(aes(x = vars, y = vals, fill = vars)) +
  geom_violin() +
  geom_boxplot(alpha = .5) +
  theme_minimal() +
  theme(axis.title = element_blank(),
        legend.position = "none")

df %>% 
  ggplot(aes(x = vars, y = vals, fill = vars)) +
  geom_violin() +
  geom_boxplot(alpha = .5) +
  theme_minimal() +
  theme(axis.title = element_blank(),
        legend.position = "none") +
  coord_flip()

Ridgeline

library(ggridges)

df %>% 
  ggplot(aes(x = vals, y = vars, fill = stat(x))) +
  geom_density_ridges_gradient(scale = 3) +
  geom_vline(xintercept = 0, linetype = "dashed") +
  scale_fill_viridis_c(option = "C") +
  theme_minimal() +
  theme(axis.title = element_blank(),
        legend.position = "none")

df %>% 
  ggplot(aes(vals, y = vars, fill = 0.5 - abs(0.5 - stat(ecdf)))) +
  stat_density_ridges(geom = "density_ridges_gradient", calc_ecdf = TRUE) +
  scale_fill_gradient(low = "white", high = "red") +
  theme_minimal() +
  theme(axis.title = element_blank(),
        legend.position = "none")