Studying the behaviour of Kendall random walks

Mateusz Staniak

We will approximate the distribution of moments when the random walk changes state through simulations.

First, we simulate many paths of Kendall random walk with normal step distribution.

library(kendallRandomWalks)
library(dplyr)
library(ggplot2)
set.seed(17)
walks <- simulate_kendall_rw(1000, 1000, rnorm, 0.5, T)
walks2 <- simulate_kendall_rw(1000, 1000, rcauchy, 0.5, T)

Example trajectory

plot(walks, max_id = 1)

plot(walks2, max_id = 1)

Number of unique states

ggplot(summarise_kendall_rw(walks, n_distinct), aes(x = aggregated), color = "black") +
  theme_bw() +
  geom_density() +
  geom_density(data = summarise_kendall_rw(walks2, n_distinct), color = "blue") +
  ylab("Estimated density") +
  xlab("Number of unique values") +
  scale_color_discrete(guide = "legend")

Jumps

diffs <- mutate_kendall_rw(walks, function(x) x - lag(x), F)
plot(diffs, max_id = 1)

diffs2 <- mutate_kendall_rw(walks2, function(x) x - lag(x), F)
plot(diffs2, max_id = 1)

Time with no change of state

diffs3 <- mutate_kendall_rw(diffs, function(x) as.numeric(x != 0), F)
lengths <- diffs3$simulation %>%
  group_by(sim_id) %>%
  mutate(id = 1:n()) %>%
  filter(sim != 0) %>%
  mutate(previous = ifelse(is.na(lag(id)), 0, lag(id))) %>%
  mutate(length = id - previous)

diffs4 <- mutate_kendall_rw(diffs2, function(x) as.numeric(x != 0), F)
lengths2 <- diffs4$simulation %>%
  group_by(sim_id) %>%
  mutate(id = 1:n()) %>%
  filter(sim != 0) %>%
  mutate(previous = ifelse(is.na(lag(id)), 0, lag(id))) %>%
  mutate(length = id - previous)


ggplot(subset(lengths, sim_id < 5), 
       aes(x = length, fill = as.factor(sim_id), group = as.factor(sim_id))) +
  geom_density() +
  theme_bw() +
  ggtitle("Distribution of time with no state-change (by simulation)")

ggplot(lengths, aes(x = length)) +
  geom_density() +
  theme_bw() +
  xlab("Jump size") +
  ggtitle("Distribution of time with no state-change (aggregated)")

ggplot(lengths2, aes(x = length)) +
  geom_density() +
  theme_bw() +
  xlab("Jump size") +
  ggtitle("Distribution of time with no state-change (aggregated)")




ggplot(subset(lengths, sim_id < 10), aes(x = id, y = length, color = as.factor(sim_id))) +
  geom_point() +
  theme_bw() +
  geom_line() +
  guides(color = "none") +
  xlab("Time") +
  ylab("Jump size") +
  ggtitle("Time with no state-change in time")