A cycle plot, also known as a seasonal subseries plot (introduced by William Cleveland), separates the seasonal pattern of a time series from the trend within each season. The series is split by seasonal period (e.g. one group per month or weekday); within each group the values are drawn in chronological order as a small line, and a horizontal reference line marks that group's mean. Comparing the mean lines across groups reveals the seasonal effect, while the slope of each subseries reveals the trend within that season — two patterns that are entangled in a standard time series plot.
""" anyplot.ai
line-cycle-seasonal: Cycle Plot (Seasonal Subseries)
Library: pygal 3.1.0 | Python 3.13.13
Quality: 82/100 | Created: 2026-06-15
"""
import os
import sys
import numpy as np
# Prevent the local pygal.py from shadowing the installed pygal package
_script_dir = os.path.dirname(os.path.abspath(__file__))
if _script_dir in sys.path:
sys.path.remove(_script_dir)
import pygal
from pygal.style import Style
# Theme tokens
THEME = os.getenv("ANYPLOT_THEME", "light")
PAGE_BG = "#FAF8F1" if THEME == "light" else "#1A1A17"
INK = "#1A1A17" if THEME == "light" else "#F0EFE8"
INK_MUTED = "#6B6A63" if THEME == "light" else "#A8A79F"
TREND_COLOR = "#009E73" # Imprint palette position 1 — monthly subseries
MEAN_COLOR = "#AE3030" # Imprint palette position 5 — seasonal mean reference
# Data: Monthly average temperatures (°C), temperate Northern European city, 2015–2024
np.random.seed(42)
years = list(range(2015, 2025)) # 10 years
n_years = len(years)
months = ["Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul", "Aug", "Sep", "Oct", "Nov", "Dec"]
n_months = 12
# Base monthly temperatures inspired by Berlin climate
month_base = np.array([1.3, 2.7, 6.5, 11.5, 16.0, 19.2, 21.3, 21.0, 16.2, 10.8, 5.6, 2.2])
warming_rate = 0.08 # °C per year — slight but visible warming over 10 years
temps = np.zeros((n_years, n_months))
for y in range(n_years):
for m in range(n_months):
temps[y, m] = month_base[m] + y * warming_rate + np.random.normal(0, 1.1)
month_means = temps.mean(axis=0)
# XY approach: month m occupies x-range [m, m+1).
# Within each month: 10 year points at x = m + y/n_years (0.0 to 0.9 within group).
# This gives only 12 x-label positions → slot width ~267px → labels fit at large font.
# Title with font size scaled to length (floor: 44)
title = "Monthly Temperature Cycle · line-cycle-seasonal · python · pygal · anyplot.ai"
n_ch = len(title)
title_font = max(44, round(66 * 67 / n_ch)) if n_ch > 67 else 66
# Palette: 12 × TREND_COLOR first, then 12 × MEAN_COLOR
palette_colors = tuple([TREND_COLOR] * n_months + [MEAN_COLOR] * n_months)
custom_style = Style(
background=PAGE_BG,
plot_background=PAGE_BG,
foreground=INK,
foreground_strong=INK,
foreground_subtle=INK_MUTED,
colors=palette_colors,
title_font_size=title_font,
label_font_size=56,
major_label_font_size=44,
legend_font_size=44,
value_font_size=36,
stroke_width=2.5,
)
# XY chart: x-axis is numeric, groups are [m, m+1) for month m
chart = pygal.XY(
style=custom_style,
width=3200,
height=1800,
title=title,
x_title="Month",
y_title="Average Temperature (°C)",
show_dots=True,
dots_size=3,
show_legend=False,
show_x_guides=False,
show_y_guides=True,
)
# x_labels: place month names at equal-interval positions across x range
chart.x_labels = months
# Trend line series — one per month, brand green (Imprint position 1)
# Within-group x: 0.0, 0.1, ..., 0.9 → 10 year points connected chronologically
for m in range(n_months):
data = [(m + y / n_years, round(float(temps[y, m]), 2)) for y in range(n_years)]
chart.add(months[m], data)
# Seasonal mean reference lines — one per month, matte red (Imprint position 5)
# dasharray "12, 6" gives strong CVD-robust distinction from solid trend lines
for m in range(n_months):
mv = round(float(month_means[m]), 2)
chart.add(f"Mean {months[m]}", [(m, mv), (m + 0.9, mv)], stroke_style={"width": 5, "dasharray": "12, 6"})
# Save
chart.render_to_png(f"plot-{THEME}.png")
with open(f"plot-{THEME}.html", "wb") as f:
f.write(chart.render())
Part of Cycle Plot (Seasonal Subseries) on anyplot.ai.