The goal of tsspiral is to create spiral version of time series plots.
Installation
You can install the development version of tsspiral from GitHub with:
# install.packages("pak")
pak::pak("thiyangt/tsspiralplot")Applications
Monthly data
library(tsspiral)
library(ggplot2)
library(viridis)
#> Loading required package: viridisLite
monthly_data <- data.frame(
date = seq.Date(
as.Date("2020-01-01"),
as.Date("2024-12-01"),
by = "month"
)
)
monthly_data$value <- rnorm(
nrow(monthly_data),
mean = 100,
sd = 20
)
ggplot(
monthly_data,
aes(
x = date,
y = value,
fill = value
)
) +
geom_tsspiral(
ring_spacing = 2
) +
scale_fill_viridis_c() +
theme_void()
ggplot(
monthly_data,
aes(
x = date,
y = value,
fill = value
)
) +
geom_tsspiral(
ring_spacing = 1
) +
scale_fill_viridis_c() +
theme_void()
monthly_data <- data.frame(
date = seq.Date(
as.Date("2020-02-01"),
as.Date("2024-12-01"),
by = "month"
)
)
monthly_data$value <- rnorm(
nrow(monthly_data),
mean = 100,
sd = 20
)
ggplot(
monthly_data,
aes(
x = date,
y = value,
fill = value
)
) +
geom_tsspiral(
ring_spacing = 2
) +
scale_fill_viridis_c() +
theme_void()
Quarterly
quarterly_data <- data.frame(
date = seq.Date(
as.Date("2020-01-01"),
as.Date("2025-10-01"),
by = "quarter"
)
)
quarterly_data$value <- rnorm(
nrow(quarterly_data),
100,
20
)
ggplot(
quarterly_data,
aes(
x = date,
y = value,
fill = value
)
) +
geom_tsspiral(
ring_spacing = 2
) +
scale_fill_viridis_c() +
theme_void()
Weekly
weekly_data <- data.frame(
date = seq.Date(
as.Date("2023-01-01"),
as.Date("2025-12-28"),
by = "week"
)
)
weekly_data$value <- rnorm(
nrow(weekly_data),
100,
20
)
ggplot(
weekly_data,
aes(
x = date,
y = value,
fill = value
)
) +
geom_tsspiral(
ring_spacing = 2
) +
scale_fill_viridis_c() +
theme_void()
Daily
set.seed(123)
dat <- data.frame(
date = seq.Date(
as.Date("2020-01-01"),
as.Date("2023-12-31"),
by = "day"
)
)
dat$value <- 100 +
20 * sin(
2 * pi *
lubridate::yday(dat$date) / 365
) +
rnorm(
nrow(dat),
sd = 5
)
ggplot(
dat,
aes(
x = date,
y = value,
fill = value
) ) +
geom_tsspiral(
ring_spacing = 2
) +
scale_fill_viridis_c() +
theme_void()
Spiral Barchart
set.seed(123)
dat <- data.frame(
date = seq.Date(
as.Date("2020-01-01"),
as.Date("2023-12-31"),
by = "day"
)
)
doy <- lubridate::yday(dat$date)
dat$value <-
20 +
15 * sin(2 * pi * doy / 365) +
35 * exp(-((doy - 100) / 20)^2) +
25 * exp(-((doy - 280) / 30)^2) +
rnorm(nrow(dat), 0, 3)
ggplot(
dat,
aes(
x = date,
y = value,
fill = value
)
) +
geom_tsspiralbar(
ring_spacing = 2
) +
scale_fill_viridis_c() +
theme_void()
library(ggplot2)
set.seed(123)
dat <- data.frame(
date = seq.Date(
as.Date("2020-01-01"),
as.Date("2023-12-31"),
by = "day"
)
)
dat$value <- 100 +
30 * sin(
2 * pi * lubridate::yday(dat$date) / 365
) +
rnorm(
nrow(dat),
sd = 8
)
ggplot(
dat,
aes(
x = date,
y = value,
fill = value
)
) +
geom_tsspiralbar(
ring_spacing = 2
) +
scale_fill_viridis_c() +
theme_void()
Missing Value Display Plot
library(ggplot2)
set.seed(123)
# Generate a daily time series
spiral_dat <- data.frame(
date = seq.Date(
from = as.Date("2020-01-01"),
to = as.Date("2023-12-31"),
by = "day"
)
)
# Generate a seasonal time series
spiral_dat$value <-
100 +
25 * sin(
2 * pi * lubridate::yday(spiral_dat$date) / 365
) +
rnorm(
nrow(spiral_dat),
sd = 5
)
# Create missing periods
missing_dates <- c(
seq.Date(
as.Date("2020-03-01"),
as.Date("2020-03-20"),
by = "day"
),
seq.Date(
as.Date("2021-07-10"),
as.Date("2021-08-05"),
by = "day"
),
seq.Date(
as.Date("2023-01-15"),
as.Date("2023-02-10"),
by = "day"
)
)
# Remove missing observations
spiral_dat <- spiral_dat[
!spiral_dat$date %in% missing_dates,
]
# Plot missingness
ggplot(
spiral_dat,
aes(x = date)
) +
geom_tsspiralmiss(
ring_spacing = 2,
observed_colour = "grey85",
missing_colour = "black"
) +
theme_void()
Calendar plot for hourly time series
# =========================================================
# Example: Two months of hourly data
# =========================================================
set.seed(123)
hourly_data <- tibble::tibble(
datetime = seq(
from = as.POSIXct("2026-01-01 00:00"),
to = as.POSIXct("2026-02-28 23:00"),
by = "hour"
)
) |>
dplyr::mutate(
hour = lubridate::hour(datetime),
day = lubridate::day(datetime),
temperature =
25 +
4 * sin(2 * pi * hour / 24) +
0.05 * day +
stats::rnorm(
length(datetime),
0,
0.8
)
)
# January
calendar_hourly_polar(
hourly_data,
datetime,
temperature,
month = "2026-01-01"
)
# February
calendar_hourly_polar(
hourly_data,
datetime,
temperature,
month = "2026-02-01"
)
Temporal Hierachy
set.seed(123)
dat <- data.frame(
date = seq.Date(
from = as.Date("2020-01-01"),
to = as.Date("2023-12-31"),
by = "day" ))
dat$value <- 20 +
5 * sin(
2 * pi * lubridate::yday(dat$date) / 365.25
) +
rnorm(nrow(dat), 0, 1)
temporal_hierarchical(
data = dat,
date = date,
value = value)