我无法处理循环数据。尽管 12 月和 1 月在圆形背景下接近，但箱线图围绕整个圆延伸，无法正确反映它们的连续性。我该如何解决这个问题？
# Necessary libraries
library(ggplot2)
library(dplyr)
library(tidyr)
library(lubridate)
library(geomtextpath)
# Initial dataset with egg laying, hatching, and observation dates
df <- data.frame(
date = as.Date(c("2019-01-08", "2023-01-18", "2023-01-18", "2024-01-21",
"2018-12-30", "2023-03-31", "2023-12-27", "2024-02-24")),
location = c("Obs1", "Obs2", "Obs3", "Obs4", "Obs5", "Obs6", "Obs7", "Obs8")
)
# Calculate hatching and egg laying dates
df <- df %>%
mutate(
hatching_date = date - weeks(5),
egg_laying_date = hatching_date - weeks(4)
)
# Transform data to long format
df_long <- df %>%
pivot_longer(cols = c(date, hatching_date, egg_laying_date),
names_to = "date_type", values_to = "event_date") %>%
mutate(
month_shifted = (month(event_date) - 1) %% 12 + 1, # Shift months so January is the start
day_of_year_shifted = ((yday(event_date) + 0) %% 365) / 365, # Adjust days for circular plot
y_offset = case_when(
date_type == "date" ~ 1,
date_type == "hatching_date" ~ 0.85, # Hatching slightly inward
date_type == "egg_laying_date" ~ 0.75 # Egg laying further outward
)
)
# Create the circular plot with shifted dates
ggplot() +
# Add month labels around the circle, starting from January as month 1
geom_textpath(data = data.frame(x = seq(0.5/13, 12.5/13, length.out = 12),
y = 1.1, label = month.abb),
aes(x, y, label = label), size = 5, color = "darkgrey") +
# Add month dividers
geom_segment(data = data.frame(x = seq(0, 11) / 12, xend = seq(0, 11) / 12,
y = 0.55, yend = 1.15),
aes(x = x, y = y, xend = xend, yend = yend), color = "grey50", linetype = "dashed") +
# Add individual points for each event type
geom_point(data = df_long, aes(x = day_of_year_shifted, y = y_offset, color = date_type), size = 3) +
# Add labels for observation points
geom_text(data = df_long %>% filter(date_type == "date"), aes(x = day_of_year_shifted, y = 1.2, label = location), size = 3) +
# Add box plots based on date types and align them correctly
geom_boxplot(data = df_long, aes(x = day_of_year_shifted, y = y_offset, fill = date_type, group = date_type),
width = 0.05, alpha = 0.3, outlier.shape = NA) +
# Customize colors and legend
scale_color_manual(values = c(date = "red", hatching_date = "blue", egg_laying_date = "green")) +
scale_fill_manual(values = c(date = "red", hatching_date = "blue", egg_laying_date = "green")) +
# Polar coordinates for circular layout
coord_polar() +
# Additional theme adjustments
theme_void() +
labs(title = "Circular Timeline with January at the Top", color = "Date Type", fill = "Date Type") +
theme(legend.position = "right")
我尝试过一些操作，例如从九月开始一年，这解决了我的问题，但如果数据全年分裂，可能没有帮助。我需要更可靠的解决方案。
图的图片链接：IMAGE
可以使用 coord_radial 制作极坐标图，其中元素可以跨越 12 月至 1 月期间。
这是一个例子：
library(lubridate)
library(ggplot2)
# show 1 month segments in a polar(radial) coordinate system
df <- data.frame(
date = as.Date(c("2021-01-01",
"2022-02-27",
"2023-03-12",
"2024-04-16",
"2025-05-20",
"2026-06-02",
"2027-07-30",
"2028-08-15",
"2029-09-20",
"2030-10-25",
"2031-12-30",
"2032-12-15"
)),
label = letters[1:12]
)
# use lubdridate to get the date interval between the min and max of df$date
# add a month, to include the month from that last date
span <- interval(min(df$date), max(df$date)+months(1))
# duration of the interval
dur <- as.duration(span)
# number of years in the interval
n_years <- as.numeric(dur, "years")
# create the breaks for the x axis, one break for every month
# every month since the first of df$date for 1 year()
start = min(df$date)
# end = start + 1 year + 1 month (to include the last month)
end = start + years(1) + months(1)
xbreaks <- seq(start, end, by = "1 month")
# there are only 12 months in a year, so we only need 12 breaks
xbreaks <- xbreaks[1:12]
xlimits = c(min(df$date), max(df$date) + months(1))
ggplot(df) +
geom_segment(
aes(
x = date, xend = date+months(1),
y = year(date), yend = year(date+months(1)),
color = label
),
linewidth = 2
) +
# the panel gris is foobar'd in this plot, so we'll just draw our own
geom_hline(
yintercept = c(2020.5, 2025.5, 2030.5, 2035.5),
color = "grey", linetype = "solid", linewidth = .1
) +
# mark the beginning of each segment
geom_text(
aes(x = date, y = year(date), label = label),
color = "black"
) +
coord_radial(
theta = "x",
start = 0,
# this is to make repating months in a circle
end = 2*pi*n_years,
direction = 1,
clip = "off",
expand = FALSE,
inner.radius = .5
)+
scale_x_continuous(
breaks = as.numeric(xbreaks),
limits = as.numeric(xlimits),
labels = format(xbreaks, "%b")
)+
scale_y_continuous(
expand = c(0, 0),
limits = c(2020, 2035.5)
)+
theme_minimal()+
theme(
# don't draw the panel grid
panel.grid = element_blank()
)