Time Series Decomposition Plot — lets-plot

A time series decomposition plot displays a time series broken down into its constituent components: the original series, trend, seasonal pattern, and residual noise. Each component is shown as a separate subplot stacked vertically, sharing a common time axis. This visualization is essential for understanding the underlying structure of time series data and identifying patterns that may not be visible in the raw series.

Time Series Decomposition Plot rendered with lets-plot

Python source (lets-plot)

""" anyplot.ai
timeseries-decomposition: Time Series Decomposition Plot
Library: letsplot 4.9.0 | Python 3.13.13
Quality: 88/100 | Updated: 2026-05-14
"""

import os
import shutil

import numpy as np
import pandas as pd
from lets_plot import (
    LetsPlot,
    aes,
    element_blank,
    element_line,
    element_rect,
    element_text,
    geom_line,
    gggrid,
    ggplot,
    ggsave,
    ggsize,
    ggtitle,
    labs,
    theme,
)
from statsmodels.tsa.seasonal import seasonal_decompose


LetsPlot.setup_html()

# Theme tokens
THEME = os.getenv("ANYPLOT_THEME", "light")
PAGE_BG = "#FAF8F1" if THEME == "light" else "#1A1A17"
INK = "#1A1A17" if THEME == "light" else "#F0EFE8"
INK_SOFT = "#4A4A44" if THEME == "light" else "#B8B7B0"

# Okabe-Ito palette for components
IMPRINT = ["#009E73", "#C475FD", "#4467A3", "#BD8233"]

# Data: Monthly temperature readings over 5 years (60 months)
np.random.seed(42)
n_months = 60
dates = pd.date_range("2019-01-01", periods=n_months, freq="MS")

# Create realistic temperature data with trend, seasonality, and noise
trend = np.linspace(15, 18, n_months)  # Gradual warming trend
seasonal = 12 * np.sin(2 * np.pi * np.arange(n_months) / 12)  # Annual cycle
noise = np.random.normal(0, 1.5, n_months)
values = trend + seasonal + noise

# Create DataFrame for decomposition
df_ts = pd.DataFrame({"date": dates, "value": values})
df_ts = df_ts.set_index("date")

# Perform seasonal decomposition (additive model)
decomposition = seasonal_decompose(df_ts["value"], model="additive", period=12)

# Extract components and create plotting DataFrames
df_original = pd.DataFrame({"date": dates, "value": values, "component": "Original"})
df_trend = pd.DataFrame({"date": dates, "value": decomposition.trend, "component": "Trend"})
df_seasonal = pd.DataFrame({"date": dates, "value": decomposition.seasonal, "component": "Seasonal"})
df_residual = pd.DataFrame({"date": dates, "value": decomposition.resid, "component": "Residual"})

# Create individual plots for each component
component_colors = {"Original": IMPRINT[0], "Trend": IMPRINT[1], "Seasonal": IMPRINT[2], "Residual": IMPRINT[3]}

anyplot_theme = theme(
    plot_background=element_rect(fill=PAGE_BG, color=PAGE_BG),
    panel_background=element_rect(fill=PAGE_BG),
    panel_grid_major=element_line(color=INK_SOFT, size=0.5),
    axis_title=element_text(color=INK, size=16),
    axis_text=element_text(color=INK_SOFT, size=14),
    axis_line=element_line(color=INK_SOFT),
    plot_title=element_text(color=INK, size=20, face="bold"),
)

# Plot 1: Original Series
p1 = (
    ggplot(df_original, aes(x="date", y="value"))
    + geom_line(color=component_colors["Original"], size=1.2)
    + labs(x="", y="Temperature (°C)", title="Original Series")
    + anyplot_theme
    + theme(axis_text_x=element_blank())
    + ggsize(1600, 200)
)

# Plot 2: Trend Component
p2 = (
    ggplot(df_trend.dropna(), aes(x="date", y="value"))
    + geom_line(color=component_colors["Trend"], size=1.2)
    + labs(x="", y="Temperature (°C)", title="Trend")
    + anyplot_theme
    + theme(axis_text_x=element_blank())
    + ggsize(1600, 200)
)

# Plot 3: Seasonal Component
p3 = (
    ggplot(df_seasonal, aes(x="date", y="value"))
    + geom_line(color=component_colors["Seasonal"], size=1.2)
    + labs(x="", y="Temperature (°C)", title="Seasonal")
    + anyplot_theme
    + theme(axis_text_x=element_blank())
    + ggsize(1600, 200)
)

# Plot 4: Residual Component
p4 = (
    ggplot(df_residual.dropna(), aes(x="date", y="value"))
    + geom_line(color=component_colors["Residual"], size=1.2)
    + labs(x="Date", y="Temperature (°C)", title="Residual")
    + anyplot_theme
    + theme(axis_text_x=element_text(angle=45))
    + ggsize(1600, 200)
)

# Create combined plot using gggrid
combined = gggrid([p1, p2, p3, p4], ncol=1)

# Add overall title
final_plot = (
    combined
    + ggsize(1600, 900)
    + ggtitle("timeseries-decomposition · letsplot · anyplot.ai")
    + theme(plot_title=element_text(color=INK, size=24, face="bold"))
)

# Save as PNG with scale for 4800x2700 resolution
ggsave(final_plot, f"plot-{THEME}.png", scale=3)

# Save HTML for interactive version
ggsave(final_plot, f"plot-{THEME}.html")

# Move files from lets-plot subdirectory to current directory if needed
lp_dir = "lets-plot-images"
if os.path.exists(lp_dir):
    for fname in [f"plot-{THEME}.png", f"plot-{THEME}.html"]:
        src = os.path.join(lp_dir, fname)
        if os.path.exists(src):
            shutil.move(src, fname)
    if not os.listdir(lp_dir):
        os.rmdir(lp_dir)

Part of Time Series Decomposition Plot on anyplot.ai.

Other implementations