Line Charts for Time Series

When the question is "how did this change over time," the line chart is the answer. Time flows continuously, and the line connecting your data points expresses that flow — it tells the reader that the value moved smoothly from one moment to the next. That single property, continuity, is what makes line charts right for trends and wrong for unordered categories.

Trend analysis is daily work for an analyst: revenue over quarters, signups per week, error rates by day. A clear line chart lets a stakeholder see momentum, seasonality and inflection points at a glance, so mastering it is time well spent.

Time belongs on the x-axis, in order

The first rule is mechanical but often broken: the x-axis must be time, in true chronological order, evenly spaced by the real interval between points. If your dates are stored as text, sort and parse them to real dates first, or the line will connect points in the wrong order and produce a meaningless zig-zag.

import matplotlib.pyplot as plt
import matplotlib.dates as mdates
from datetime import date

months = [date(2026, m, 1) for m in range(1, 9)]
signups = [320, 410, 480, 460, 610, 750, 720, 880]

fig, ax = plt.subplots(figsize=(8, 4))
ax.plot(months, signups, color="#2f6fdb", linewidth=2, marker="o")

ax.set_ylabel("New signups")
ax.set_title("Signups trend upward through H1 2026",
             loc="left", fontweight="bold")
ax.xaxis.set_major_formatter(mdates.DateFormatter("%b"))
for spine in ["top", "right"]:
    ax.spines[spine].set_visible(False)
ax.grid(axis="y", color="#eee")
ax.set_axisbelow(True)

# Label the final point directly
ax.annotate(f"{signups[-1]}", xy=(months[-1], signups[-1]),
            xytext=(6, 0), textcoords="offset points", va="center")

plt.tight_layout()
plt.show()

What this renders: a single blue line rising from about 320 signups in
January to 880 in August, with a small marker at each month and month
abbreviations on the x-axis. The upward slope is obvious, a dip in
April is visible, and the final value is labeled at the end of the line
so the reader sees the latest number without reading the axis.

Multiple series without the spaghetti

The most common line-chart failure is too many lines. Five or six series crossing over each other become an unreadable tangle. You have three good fixes.

First, highlight one, mute the rest: draw the series you are discussing in a strong color and the others in light gray, so the chart has a clear subject. Second, small multiples: draw one small chart per series in a grid, all sharing the same axes, so each trend is clean and comparison is by position. Third, direct labels: place each series' name at the end of its line instead of in a legend, removing the constant lookup.

Spaghetti (avoid):            Highlight one (prefer):
   /\  /\   /\                    ___________ Product A (bold)
  /  \/  \ / X crossing      ....../....... Product B (gray)
 / crossing \/               ....\......... Product C (gray)
 Five lines, unreadable      One clear subject, context muted

Choose the right granularity

Daily data shows every spike but can look like static; monthly data smooths noise and reveals the underlying trend. The right choice depends on the signal you want. If a stakeholder cares about the yearly growth story, aggregate to months. If they are debugging a specific outage, daily or hourly is essential. Aggregating is a deliberate analytical decision, not a default.

The y-axis and honesty

Unlike bar charts, line charts do not have to start at zero, because they encode change rather than length. A framed non-zero axis can make a real trend readable. The danger is the opposite of the bar-chart problem: zooming the axis so tightly that ordinary fluctuation looks like a crisis. Label the axis clearly and keep the vertical scale proportionate to the story. This tension between framing and exaggeration is explored in misleading charts.

Practical usage

Analysts pair line charts with a few standard techniques. A rolling average line laid over noisy daily data separates trend from noise. Reference lines — a target, a launch date, last year's level — give the reader context for whether a value is good. Annotations on key events (a price change, a campaign) explain why the line moved. And when comparing to a prior period, plotting this year and last year as two lines is clearer than two separate charts.

Common mistakes

• Lines for categories. Connecting unordered categories implies a trend that does not exist. Use bars.
• Unsorted or unevenly spaced time. Points connected out of order, or with equal spacing for unequal intervals, distort the trend.
• Too many lines. More than about five series becomes spaghetti; highlight or split into small multiples.
• Silently dropping missing dates. Connecting across a gap hides missing data and fakes a smooth trend; show the gap instead.
• Over-tight y-axis. Zooming so noise looks dramatic is a subtle way to mislead. Keep the scale honest.

In interviews

A common prompt is "you're given daily revenue for the last two years — how would you present the trend?" A strong answer covers granularity (aggregate to weekly or monthly to reduce noise), a rolling average for smoothing, comparison to the prior year as a second line, and direct labeling. If asked why not a bar chart, explain that time is continuous and the line communicates flow and momentum that separated bars cannot.

Where this fits in your learning path

Line charts follow naturally from the choosing the right chart framework as the answer to trend questions, and they pair with bar charts best practices as the two most-used analyst charts. Watch for the axis-scaling pitfalls detailed in misleading charts. Time-series literacy is a core skill on the data analyst roadmap and across the data analytics hub.