Chapter 6: Matplotlib

Explicit control over figures, axes, and chart styling

Sections: Matplotlib basics · Figures and axes · Styling and plot types

Use arrow keys or Space to navigate · Press F for fullscreen

Chapter Roadmap

Part Main Question Core Vocabulary
Matplotlib basics How do I draw and label a first chart? pyplot, stateful interface, line plot, legend
Figures and axes How do I build and arrange charts on purpose? figure, axes, plt.subplots(), inset, DPI
Styling and plot types How do I style a chart and pick the right type? rcParams, axis range, alpha, histogram, boxplot

pandas plotting (Chapter 5) is built on Matplotlib. This chapter shows the layer underneath, for when a business chart needs more control.

Chapter Goals

By the end of this chapter, you will be able to:

  1. Distinguish a figure from its axes and build single- and multi-chart figures with plt.subplots().
  2. Create line, bar, scatter, histogram, and box plots that match a management question about trend, comparison, relationship, or distribution.
  3. Label and style charts with titles, axis labels, legends, colors, line styles, markers, transparency, and axis ranges.
  4. Arrange charts for comparison with subplots, shared axes, figure titles, and add_axes() insets.
  5. Set figure size and resolution, save figures for reports, and finish pandas plots with Matplotlib methods.

Matplotlib Basics

First plots, labels, and the pyplot interface

Two Ways to Use Matplotlib

Stateful (pyplot) Object-oriented
Looks like plt.plot(), plt.title() fig, ax = plt.subplots(), ax.set_title()
Acts on The current figure and axes The figure and axes you name
Best for Quick, one-off exploration Most real charts, especially several at once 
import matplotlib.pyplot as plt   # the plotting module, always imported as plt

"Stateful" means pyplot quietly tracks a current figure and axes; each plt.* call changes whatever is current.

A First Line Plot

month = np.arange(1, 7)
revenue = np.array([110, 104, 118,
                    121, 125, 119])

plt.figure(figsize=(4, 3))
plt.plot(month, revenue, c='red',
         label='Revenue')
plt.xlabel('Month')
plt.ylabel('Revenue ($K)')
plt.title('Monthly Revenue')
plt.legend()
  • plt.figure(figsize=...) sets the canvas size in inches.
  • plt.plot(x, y) draws a line: good for a trend.
  • label= names the line; plt.legend() shows it.
  • Every chart needs a title and axis labels with units.

pandas does the same in one call: df.plot(x='month', y='revenue', title='Monthly Revenue').

Scatter Plots and plt.subplot()

Relationship between two measures

ad_spend = np.array([5, 10, 15, 20, 25, 30])
sales = np.array([60, 85, 110, 120, 150, 165])

plt.scatter(ad_spend, sales)

Two charts in a grid, stateful style

plt.figure(figsize=(8, 3))
plt.subplot(1, 2, 1)     # 1 row, 2 cols, chart 1
plt.plot(month, revenue, 'r--')
plt.subplot(1, 2, 2)     # chart 2
plt.plot(month, units, 'g*-')

plt.subplot() (singular) is the older way. For new code, use plt.subplots() (plural), next section.

Figures and Axes

Building, arranging, sizing, and saving charts

Figure vs Axes

Figure Axes
What it is The whole canvas One chart on the canvas
Contains One or more axes Data, ticks, labels, title, legend
Made by plt.figure(), plt.subplots() plt.subplots(), fig.add_axes()

Most formatting is done with axes methods: ax.set_title(), ax.set_xlabel(), ax.set_ylabel(), ax.legend().

Figure
└── Axes (one chart)
    ├── XAxis
    ├── YAxis
    ├── Title
    ├── Lines / Bars
    └── Legend

plt.subplots(): The Everyday Workflow

One chart

fig, ax = plt.subplots(figsize=(4, 3))
ax.plot(month, revenue, 'g--')
ax.set_xlabel('Month')
ax.set_ylabel('Revenue ($K)')
ax.set_title('Monthly Revenue')

Several charts: axes is an array

fig, axes = plt.subplots(1, 2, figsize=(6, 2))
axes[0].plot(month, store_a, 'r--')
axes[0].set_title('Store A')
axes[1].plot(month, store_b, 'g*-')
axes[1].set_title('Store B')

plt.subplots() returns the figure and its axes together; unpack them with fig, ax = ... or fig, axes = ....

Loop, Tidy, and Title the Figure

fig, axes = plt.subplots(1, 2, figsize=(6, 2))
for ax, sales, name in zip(axes,
                           [store_a, store_b],
                           ['Store A', 'Store B']):
    ax.plot(month, sales)
    ax.set_xlabel('Month')
    ax.set_ylabel('Sales ($K)')
    ax.set_title(name)
plt.tight_layout()

fig, axes = plt.subplots(1, 2, figsize=(8, 3),
                         sharey=True)
axes[0].plot(quarters, north, marker='o')
axes[1].plot(quarters, south, marker='o')
fig.suptitle('Quarterly Revenue by Region ($K)')
  • zip() pairs each axes with its data and title.
  • tight_layout() fixes overlapping labels.
  • sharey=True: same scale, fair comparison.
  • fig.suptitle() titles the whole figure.

Insets with add_axes()

fig = plt.figure(figsize=(4, 3))
main = fig.add_axes([0.1, 0.1, 0.8, 0.8])
inset = fig.add_axes([0.2, 0.5, 0.25, 0.3])

main.plot(day, daily_sales)
main.set_title('Daily Sales in June')

inset.plot(day[14:21], daily_sales[14:21], 'r')
inset.set_title('Promotion Week')

[left, bottom, width, height] are fractions of the figure:

Value Meaning
0.1, 0.1 Start 10% from the left and bottom
0.8, 0.8 Take 80% of the width and height

Use plt.subplots() for grids; use add_axes() only when you need exact placement, such as a zoomed inset.

Size, Resolution, and Saving

fig, ax = plt.subplots(figsize=(6, 3))
ax.plot(quarters, revenue, marker='o')
ax.set_title('Quarterly Revenue')

fig.savefig('quarterly_revenue.png', dpi=150)
Setting Controls Typical value
figsize=(w, h) Size in inches (6, 3) for a slide
dpi= Dots per inch 100 screen, 200–300 print
File extension Format .png, .pdf, .svg

A chart that looks sharp on screen can print blurry. Save report figures with dpi= of 200 or more.

Styling and Plot Types

Styling charts and matching the chart to the question

Figure Methods vs Axes Methods

Figure: the whole canvas

Method Sets
fig.suptitle() Title over all charts
fig.tight_layout() Spacing
fig.savefig() Saved file
fig.set_size_inches() Size

Axes: one chart

Method Sets
ax.set_title() Chart title
ax.set_xlabel(), ax.set_ylabel() Axis labels
ax.set_xlim(), ax.set_ylim() Axis ranges
ax.legend(), ax.grid() Legend, grid

Defaults, Legends, and Axis Ranges

Change defaults for every later chart

plt.style.use('ggplot')             # a style sheet
plt.rcParams['figure.dpi'] = 150    # one setting

Label each series, then place the legend

ax.plot(month, laptops, label='Laptops')
ax.plot(month, phones, label='Phones')
ax.legend(loc='lower right')

Zoom in on the promotion week

ax.plot(day, daily_sales)
ax.set_xlim(14, 22)
ax.set_ylim(30, 80)

Axis ranges change what the reader sees, not the data. Starting a bar chart's y-axis above zero can exaggerate small differences.

Colors, Line Styles, and Markers

ax.plot(month, actual, 'b.-')     # format string
ax.plot(month, budget, color='green',
        ls='--', lw=1, marker='s')
ax.plot(month, north, color='#8B008B',
        alpha=0.5)                # hex color, half transparent
Parameter Short form Example
color c 'red', '#FF8C00'
linestyle ls '-', '--', ':'
linewidth lw 2.5
marker 'o', 's', '^'
alpha 0.5 (0 to 1)

A format string such as 'r--' sets color, line style, and marker in one argument.

Choosing a Plot Type

Management question Plot type Matplotlib
How did revenue change over the year? (trend) Line plot ax.plot(x, y)
Which department sold the most? (comparison) Bar plot ax.bar(categories, values)
Does a higher price mean fewer units sold? (relationship) Scatter plot ax.scatter(x, y)
What order values are typical? (distribution) Histogram ax.hist(values, bins=10)
Which warehouse delivers most consistently? (distribution by group) Boxplot ax.boxplot([a, b, c])

Start from the question, then pick the chart. The same data can answer different questions with different charts.

Plot Types in Code

fig, ax = plt.subplots()
ax.bar(['Apparel', 'Electronics', 'Home', 'Toys'],
       [240, 410, 185, 120])
ax.set_title('Sales by Department ($K)')

fig, ax = plt.subplots()
ax.hist(order_values, bins=10)
ax.set_xlabel('Order Value ($)')

fig, ax = plt.subplots()
ax.boxplot(delivery_days)
ax.set_xticks([1, 2, 3],
              ['Warehouse A', 'Warehouse B',
               'Warehouse C'])
ax.set_ylabel('Delivery Time (days)')

A boxplot shows the median (line), the middle 50% (box), the typical range (whiskers), and outliers (circles).

Three Workflows

Workflow Use it when Start with
Stateful pyplot Quick, one-off exploration plt.plot(x, y)
plt.subplots() Most charts, one or many fig, ax = plt.subplots()
fig.add_axes() Exact placement or an inset fig.add_axes([l, b, w, h])

Default to plt.subplots(). Every chart then has a named ax you can format, and every figure a named fig you can save.

Chapter 6 Vocabulary

# Term Short Meaning
1 Figure The whole canvas that holds one or more axes.
2 Axes One chart inside a figure.
3 Stateful interface pyplot functions that act on the current figure and axes.
4 Object-oriented interface Figure and axes stored in variables; methods called on them.
5 Figure title One title above all subplots (fig.suptitle()).
6 Shared axis One axis range for every subplot (sharey=True).
7 Inset axes A small chart placed on a larger one (fig.add_axes()).
8 DPI Dots per inch: the resolution of a drawn or saved figure.
9 Alpha Transparency from 0 (invisible) to 1 (opaque).
10 Boxplot Median, quartiles, whiskers, and outliers of a distribution.

Practice

Assignment What it checks
Preview Chapter vocabulary, before class
Lab Five charts for one retailer: trend, actual vs forecast, dashboard, scatter, saved report figure
Homework Scenario true/false questions and five charts: targets, overlapping histograms, an inset, shared axes, pandas plus Matplotlib

The lab and homework are graded on the chart itself (axes, titles, labels, styles, saved files), so either plotting style earns credit.

What To Carry Forward

After this chapter, students should be ready to:

  • Build any chart with fig, ax = plt.subplots() and finish it with axes methods.
  • Choose a chart type from the business question it has to answer.
  • Arrange charts so comparisons are fair and the message is clear.
  • Save report-ready figures, and fine-tune pandas and Seaborn charts through their Matplotlib axes.

End of Chapter 6

Next: Chapter 7: Seaborn