Lecture 7: Data Visualization

This lecture covers plotting with Matplotlib and Seaborn. The data are the Fama-French five factors from Lecture 6.

What We'll Cover

Ch 13 (Matplotlib): the anatomy of a plot, the pyplot and object-oriented interfaces, line plots, customization, other plot types, subplots, saving figures, and style sheets.
Ch 14 (Seaborn): distribution, relational, regression, and categorical plots, correlation heatmaps, multi-plot grids, and styling.
Our running data is the Fama-French 5 Factors from Ken French's data library.

Ch 13: Matplotlib

Matplotlib is the foundational plotting library in Python.
Most other Python plotting tools (including pandas.plot() and seaborn) are built on it.
Once you know Matplotlib, you can customize any plot, including ones made by higher-level libraries.

Importing Matplotlib

The conventional alias for the plotting interface is plt.

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

Loading Our Data

We'll use the Fama-French 5 Factors dataset as our running example for plotting.
These are monthly factor returns from Kenneth French's data library: Mkt-RF (market premium), SMB (size), HML (value), RMW (profitability), CMA (investment), and RF (risk-free rate).
The cell below downloads the data fresh. It tries pandas-datareader first and falls back to a direct download if that fails.

import io, zipfile, urllib.request

# Try pandas-datareader first (the cleanest path)
try:
    import pandas_datareader.data as web
    ff = web.DataReader('F-F_Research_Data_5_Factors_2x3',
                        'famafrench',
                        start='1990-01-01',
                        end='2023-12-31')
    ff5 = ff[0]
    ff5.index = ff5.index.to_timestamp()
    print(f"Loaded via pandas-datareader: {ff5.shape}")
except Exception as e:
    # Fallback: download the CSV zip directly from Ken French's website
    print(f"pandas-datareader failed ({type(e).__name__}); falling back to direct download.")
    url = ("https://mba.tuck.dartmouth.edu/pages/faculty/ken.french/"
           "ftp/F-F_Research_Data_5_Factors_2x3_CSV.zip")
    req = urllib.request.Request(url, headers={'User-Agent': 'Mozilla/5.0'})
    with urllib.request.urlopen(req) as r:
        zip_bytes = r.read()
    with zipfile.ZipFile(io.BytesIO(zip_bytes)) as z:
        raw = z.open(z.namelist()[0]).read().decode('latin1')

    # The file has a text header; monthly data starts after a blank-line break.
    # Find the first line that begins with a 6-digit YYYYMM and parse from there.
    lines = raw.splitlines()
    start = next(i for i, ln in enumerate(lines)
                 if len(ln) >= 6 and ln[:6].isdigit())
    end = next((i for i in range(start, len(lines))
                if not (len(lines[i]) >= 6 and lines[i][:6].isdigit())),
               len(lines))
    monthly_csv = "Date,Mkt-RF,SMB,HML,RMW,CMA,RF\n" + "\n".join(lines[start:end])
    ff5 = pd.read_csv(io.StringIO(monthly_csv), index_col='Date')
    ff5.index = pd.to_datetime(ff5.index, format='%Y%m')
    print(f"Loaded via direct download: {ff5.shape}")

ff5.head()

Loaded via pandas-datareader: (408, 6)

	Mkt-RF	SMB	HML	RMW	CMA	RF
Date
1990-01-01	-7.80	-1.21	0.83	-1.12	1.40	0.57
1990-02-01	1.11	1.15	0.65	-0.04	-0.61	0.57
1990-03-01	1.83	1.59	-2.90	1.99	-0.94	0.64
1990-04-01	-3.36	-0.35	-2.57	1.78	-1.00	0.69
1990-05-01	8.44	-2.41	-3.89	1.68	-1.63	0.68

Note on units:

Fama-French factor returns are stored in percent (a value of 1.5 means 1.5%).
For plots like cumulative returns, we'll convert to decimal first (divide by 100).
For histograms of monthly returns, percent is fine but we'll label axes accordingly.

The Anatomy of a Plot

Every Matplotlib plot has the same three-part structure:
- A Figure is the whole image (the page).
- One or more Axes sit inside the Figure (each Axes is one plot).
- Each Axes has Axis objects (x-axis, y-axis) with ticks and labels.

Two Interfaces

The pyplot interface (plt.plot(), plt.title(), ...) is quick for simple plots. It operates on the "current" Axes implicitly.
The object-oriented interface (fig, ax = plt.subplots(), ax.plot(), ...) gives you explicit handles and scales to complex layouts.
We'll show both. Prefer the object-oriented style for anything beyond a one-liner.

Your First Plot

The pyplot Style

# Quick line plot of one factor
plt.plot(ff5.index, ff5['Mkt-RF'])
plt.title('Market Premium (Mkt-RF), Monthly')
plt.xlabel('Date')
plt.ylabel('Return (%)')
plt.show()

The Object-Oriented Style

plt.subplots() creates a Figure and an Axes and returns both.

fig, ax = plt.subplots()
ax.plot(ff5.index, ff5['Mkt-RF'])
ax.set_title('Market Premium (Mkt-RF), Monthly')
ax.set_xlabel('Date')
ax.set_ylabel('Return (%)')
plt.show()

The plots look identical. We'll use the OO style from here on.

Line Plots

Multiple Lines on One Axes

Call ax.plot() multiple times to overlay several series.

fig, ax = plt.subplots(figsize=(10, 4))
ax.plot(ff5.index, ff5['Mkt-RF'], label='Mkt-RF')
ax.plot(ff5.index, ff5['SMB'],    label='SMB')
ax.plot(ff5.index, ff5['HML'],    label='HML')
ax.set_xlabel('Date')
ax.set_ylabel('Monthly return (%)')
ax.set_title('Three Fama-French Factors')
ax.legend()
plt.show()

Cumulative Returns (Converting Percent to Decimal)

Cumulative wealth (1+r).cumprod() requires returns in decimal form.

# Convert from % to decimal, then compute cumulative wealth
cum_wealth = (1 + ff5[['Mkt-RF', 'SMB', 'HML']] / 100).cumprod()

fig, ax = plt.subplots(figsize=(10, 4))
ax.plot(cum_wealth.index, cum_wealth['Mkt-RF'], label='Mkt-RF')
ax.plot(cum_wealth.index, cum_wealth['SMB'],    label='SMB')
ax.plot(cum_wealth.index, cum_wealth['HML'],    label='HML')
ax.set_xlabel('Date')
ax.set_ylabel('Cumulative wealth ($1 invested)')
ax.set_title('Cumulative Returns of Three Factors')
ax.legend()
plt.show()

Mkt-RF compounds to a much larger multiple than SMB or HML over 30+ years.

Plot Customization

Color, Line Style, Width, Markers

fig, ax = plt.subplots(figsize=(10, 4))
ax.plot(ff5.index, ff5['Mkt-RF'],
        color='steelblue',
        linestyle='-',
        linewidth=1.2,
        label='Mkt-RF')
ax.plot(ff5.index, ff5['Mkt-RF'].rolling(12).mean(),
        color='red',
        linestyle='--',
        linewidth=2,
        label='12-month rolling mean')
ax.set_title('Market Premium with 1-Year Rolling Mean')
ax.set_ylabel('Return (%)')
ax.legend()
plt.show()

Reference Lines: `axhline` and `axvline`

fig, ax = plt.subplots(figsize=(10, 4))
ax.plot(ff5.index, ff5['Mkt-RF'], color='steelblue', linewidth=0.8)
ax.axhline(0, color='black', linewidth=0.8)              # zero line
ax.axhline(ff5['Mkt-RF'].mean(), color='red',
           linestyle='--', label='Long-run mean')
ax.set_title('Monthly Market Premium with Zero and Mean Lines')
ax.set_ylabel('Return (%)')
ax.legend()
plt.show()

Grid

fig, ax = plt.subplots(figsize=(8, 3))
ax.plot(ff5.loc['2020', 'Mkt-RF'])
ax.set_title('2020 Market Premium')
ax.set_ylabel('Return (%)')
ax.grid(True, alpha=0.3)
plt.show()

Other Plot Types

Histogram

Use ax.hist() to see the distribution of values.

fig, ax = plt.subplots(figsize=(8, 4))
ax.hist(ff5['Mkt-RF'], bins=40, color='steelblue', edgecolor='white')
ax.set_title('Distribution of Monthly Market Premium')
ax.set_xlabel('Monthly return (%)')
ax.set_ylabel('Count')
plt.show()

Scatter Plot

Use ax.scatter() to plot pairs of points.

# Do small-cap and value premia move together?
fig, ax = plt.subplots(figsize=(6, 6))
ax.scatter(ff5['SMB'], ff5['HML'], alpha=0.4, s=15)
ax.set_xlabel('SMB (%)')
ax.set_ylabel('HML (%)')
ax.set_title('SMB vs HML Monthly Returns')
ax.axhline(0, color='black', linewidth=0.5)
ax.axvline(0, color='black', linewidth=0.5)
plt.show()

Bar Chart

Use ax.bar() for vertical bars (categorical x-axis).

# Average monthly return of each factor
mean_returns = ff5[['Mkt-RF', 'SMB', 'HML', 'RMW', 'CMA']].mean()

fig, ax = plt.subplots(figsize=(7, 4))
ax.bar(mean_returns.index, mean_returns.values, color='steelblue')
ax.axhline(0, color='black', linewidth=0.8)
ax.set_title('Average Monthly Premium by Factor (1990-2023)')
ax.set_ylabel('Average return (%)')
plt.show()

Boxplot

A boxplot summarizes a distribution: median, quartiles, whiskers, and outliers.

fig, ax = plt.subplots(figsize=(7, 4))
ax.boxplot([ff5['Mkt-RF'], ff5['SMB'], ff5['HML'], ff5['RMW'], ff5['CMA']],
           tick_labels=['Mkt-RF', 'SMB', 'HML', 'RMW', 'CMA'])
ax.axhline(0, color='red', linewidth=0.8, linestyle='--')
ax.set_ylabel('Monthly return (%)')
ax.set_title('Distribution of Each Factor')
plt.show()

Subplots

Use plt.subplots(nrows, ncols) to lay out multiple Axes in a grid.
The returned ax is a NumPy array of Axes objects.

factors = ['Mkt-RF', 'SMB', 'HML', 'RMW']

fig, axes = plt.subplots(nrows=2, ncols=2, figsize=(10, 6))

for ax, factor in zip(axes.flat, factors):
    ax.hist(ff5[factor], bins=30, color='steelblue', edgecolor='white')
    ax.set_title(factor)
    ax.set_xlabel('Monthly return (%)')

fig.suptitle('Distributions of Four Fama-French Factors', fontsize=14)
fig.tight_layout()
plt.show()

axes.flat gives a flat iterator over all Axes regardless of grid shape.
fig.tight_layout() automatically adjusts spacing so titles and labels don't overlap.

Saving Figures

Use fig.savefig() to write a figure to disk.
The format is inferred from the file extension (.png, .pdf, .svg, ...).

fig, ax = plt.subplots(figsize=(8, 4))
ax.plot(ff5.index, ff5['Mkt-RF'])
ax.set_title('Market Premium')
ax.set_ylabel('Return (%)')

fig.savefig('mkt_rf.png', dpi=150, bbox_inches='tight')
plt.show()

dpi=150 controls resolution; bbox_inches='tight' trims whitespace.

Style Sheets

plt.style.use(name) applies a preset look-and-feel.
Useful built-ins: 'ggplot', 'seaborn-v0_8', 'fivethirtyeight', 'classic'.

# See what styles are available
print(plt.style.available[:10])

['Solarize_Light2', '_classic_test_patch', '_mpl-gallery', '_mpl-gallery-nogrid', 'bmh', 'classic', 'dark_background', 'fast', 'fivethirtyeight', 'ggplot']

# Apply a style for the rest of the notebook
plt.style.use('ggplot')

fig, ax = plt.subplots(figsize=(8, 3))
ax.plot(ff5.index, ff5['Mkt-RF'])
ax.set_title('Market Premium (ggplot style)')
ax.set_ylabel('Return (%)')
plt.show()

# Reset to default style before continuing
plt.style.use('default')

Note: common pitfall

Forgetting plt.show() at the end of a cell still produces a plot in a notebook, but you'll also see noise like <matplotlib.lines.Line2D at 0x...>.
Calling plt.show() (or assigning to _) keeps the output clean.

# This works in a notebook but prints a stray repr line first.
fig, ax = plt.subplots(figsize=(6, 2))
ax.plot([1, 2, 3], [1, 4, 9])

[<matplotlib.lines.Line2D at 0x7e1da3c65d10>]

In-Class Exercise

13-1. Line Plot:

Plot the RF (risk-free rate) series from ff5 as a line plot using the object-oriented interface.
Add a title 'Risk-Free Rate, Monthly', an x-label 'Date', and a y-label 'Rate (%)'.

13-2. Cumulative Wealth Comparison:

Compute the cumulative wealth of Mkt-RF and RMW (convert from percent to decimal first).
Plot both on the same Axes with distinct colors and a legend.
Set figsize=(10, 4) and give the plot an appropriate title.

13-3. Histogram with Reference Line:

Plot a histogram of HML with 40 bins.
Add a vertical dashed red line at the mean of HML using ax.axvline().
Label the axes appropriately.

13-4. Side-by-Side Subplots:

Create a figure with 1 row and 2 columns using plt.subplots(1, 2, figsize=(12, 4)).
On the left, plot Mkt-RF over time as a line.
On the right, plot a histogram of Mkt-RF.
Add appropriate titles to each subplot.

13-5. Customized Scatter:

Make a scatter plot of Mkt-RF (x) vs SMB (y).
Use alpha=0.3 for transparency and s=10 for point size.
Add horizontal and vertical zero lines using axhline(0) and axvline(0).
Save the figure to 'mkt_vs_smb.png' with dpi=150.

Ch 14: Seaborn

Seaborn is a high-level statistical visualization library built on top of Matplotlib.
It produces attractive plots with less code, and is especially good at:
- distribution plots
- regression plots
- categorical comparisons
- correlation heatmaps
Seaborn understands pandas DataFrames natively. You usually pass data=df and reference columns by name.

Importing Seaborn

The conventional alias for Seaborn is sns.

import seaborn as sns
sns.set_theme(style='whitegrid')

sns.set_theme() applies Seaborn's default styling to all subsequent Matplotlib plots.
Common style choices: 'darkgrid', 'whitegrid', 'dark', 'white', 'ticks'.

Distribution Plots

`histplot()`: Histograms with Density

fig, ax = plt.subplots(figsize=(8, 4))
sns.histplot(data=ff5, x='Mkt-RF', bins=40, kde=True, ax=ax)
ax.set_title('Distribution of Mkt-RF (with KDE)')
ax.set_xlabel('Monthly return (%)')
plt.show()

kde=True overlays a kernel density estimate (a smoothed estimate of the distribution).

`kdeplot()`: Compare Distributions

Plot densities of multiple factors on the same Axes to compare their shapes.

fig, ax = plt.subplots(figsize=(8, 4))
for factor in ['Mkt-RF', 'SMB', 'HML']:
    sns.kdeplot(data=ff5, x=factor, label=factor, ax=ax)
ax.set_title('Distribution Comparison: Mkt-RF, SMB, HML')
ax.set_xlabel('Monthly return (%)')
ax.legend()
plt.show()

`boxplot()` and `violinplot()`

Box and violin plots compare distributions across categories.
They work best with data in long format, so we melt the DataFrame first.

ff5_long = ff5[['Mkt-RF', 'SMB', 'HML', 'RMW', 'CMA']].melt(
    var_name='factor', value_name='return'
)
ff5_long.head()

	factor	return
0	Mkt-RF	-7.80
1	Mkt-RF	1.11
2	Mkt-RF	1.83
3	Mkt-RF	-3.36
4	Mkt-RF	8.44

fig, ax = plt.subplots(figsize=(8, 4))
sns.boxplot(data=ff5_long, x='factor', y='return', ax=ax)
ax.axhline(0, color='red', linestyle='--', linewidth=0.8)
ax.set_title('Monthly Return Distribution by Factor')
ax.set_ylabel('Return (%)')
plt.show()

# A violin plot shows the full KDE on each side
fig, ax = plt.subplots(figsize=(8, 4))
sns.violinplot(data=ff5_long, x='factor', y='return', ax=ax)
ax.axhline(0, color='red', linestyle='--', linewidth=0.8)
ax.set_title('Violin Plot of Monthly Returns by Factor')
ax.set_ylabel('Return (%)')
plt.show()

Relational Plots

`scatterplot()` with a `hue`

hue colors points by a categorical (or numeric) column.

# Add a decade column for color-coding
ff5_demo = ff5.copy()
ff5_demo['decade'] = (ff5_demo.index.year // 10 * 10).astype(str) + 's'
ff5_demo.head()

	Mkt-RF	SMB	HML	RMW	CMA	RF	decade
Date
1990-01-01	-7.80	-1.21	0.83	-1.12	1.40	0.57	1990s
1990-02-01	1.11	1.15	0.65	-0.04	-0.61	0.57	1990s
1990-03-01	1.83	1.59	-2.90	1.99	-0.94	0.64	1990s
1990-04-01	-3.36	-0.35	-2.57	1.78	-1.00	0.69	1990s
1990-05-01	8.44	-2.41	-3.89	1.68	-1.63	0.68	1990s

fig, ax = plt.subplots(figsize=(7, 6))
sns.scatterplot(data=ff5_demo, x='Mkt-RF', y='SMB', hue='decade',
                alpha=0.6, ax=ax)
ax.axhline(0, color='black', linewidth=0.5)
ax.axvline(0, color='black', linewidth=0.5)
ax.set_title('Mkt-RF vs SMB, Colored by Decade')
plt.show()

`lineplot()` with a confidence band

lineplot is convenient when you have repeated observations and want a mean line with a confidence band.

# Mean monthly return by year, with confidence band
ff5_demo['year'] = ff5_demo.index.year

fig, ax = plt.subplots(figsize=(10, 4))
sns.lineplot(data=ff5_demo, x='year', y='Mkt-RF', ax=ax,
             errorbar=('ci', 95))
ax.axhline(0, color='red', linewidth=0.8, linestyle='--')
ax.set_title('Mean Monthly Market Premium by Year (with 95% CI)')
ax.set_ylabel('Monthly return (%)')
plt.show()

Regression Plots

`regplot()`: Scatter with Regression Line

Adds a linear-regression fit and a confidence band on top of a scatter plot.

fig, ax = plt.subplots(figsize=(7, 5))
sns.regplot(data=ff5, x='Mkt-RF', y='HML', scatter_kws={'alpha': 0.4, 's': 15}, ax=ax)
ax.set_title('Value Premium (HML) vs Market Premium (Mkt-RF)')
ax.set_xlabel('Mkt-RF (%)')
ax.set_ylabel('HML (%)')
plt.show()

The shaded band is the 95% confidence interval for the regression line.

`lmplot()`: Regression by Group

lmplot fits a separate regression for each level of a categorical hue variable.

# Split decades into 'early' (1990-2009) and 'late' (2010-2023)
ff5_demo['era'] = np.where(ff5_demo.index.year < 2010, 'Early (1990-2009)', 'Late (2010-2023)')

sns.lmplot(data=ff5_demo, x='Mkt-RF', y='HML', hue='era',
           height=5, aspect=1.2,
           scatter_kws={'alpha': 0.4, 's': 15})
plt.title('HML vs Mkt-RF: Two Eras')
plt.show()

lmplot returns a FacetGrid, not an Axes, so we use plt.title() rather than ax.set_title().

Categorical Plots

`barplot()`: Mean and Confidence Interval

barplot computes the mean of y for each level of x by default, plus a confidence interval.

fig, ax = plt.subplots(figsize=(8, 4))
sns.barplot(data=ff5_demo, x='decade', y='Mkt-RF', ax=ax)
ax.axhline(0, color='black', linewidth=0.8)
ax.set_title('Mean Monthly Market Premium by Decade')
ax.set_ylabel('Mean return (%)')
plt.show()

`countplot()`: Frequency of Each Category

# How many up vs down market months in each decade?
ff5_demo['direction'] = np.where(ff5_demo['Mkt-RF'] > 0, 'up', 'down')

fig, ax = plt.subplots(figsize=(8, 4))
sns.countplot(data=ff5_demo, x='decade', hue='direction', ax=ax)
ax.set_title('Up vs Down Market Months by Decade')
ax.set_ylabel('Number of months')
plt.show()

Matrix Plot: Correlation Heatmap

A correlation heatmap shows how factors move together at a glance.

corr = ff5[['Mkt-RF', 'SMB', 'HML', 'RMW', 'CMA']].corr()
corr

	Mkt-RF	SMB	HML	RMW	CMA
Mkt-RF	1.000000	0.231935	-0.111981	-0.314162	-0.329824
SMB	0.231935	1.000000	0.000160	-0.458621	-0.043114
HML	-0.111981	0.000160	1.000000	0.317990	0.660124
RMW	-0.314162	-0.458621	0.317990	1.000000	0.218510
CMA	-0.329824	-0.043114	0.660124	0.218510	1.000000

fig, ax = plt.subplots(figsize=(6, 5))
sns.heatmap(corr, annot=True, fmt='.2f',
            cmap='RdBu_r', center=0,
            vmin=-1, vmax=1,
            square=True, ax=ax)
ax.set_title('Correlation of Fama-French Factors')
plt.show()

annot=True writes the value in each cell; fmt='.2f' formats to two decimals.
cmap='RdBu_r' is a diverging palette ideal for correlations: red = positive, blue = negative.
center=0 ensures the colormap is symmetric around zero.

Multi-Plot Grids

`pairplot()`: All-Pairs Scatter Matrix

pairplot shows scatter plots for every pair of variables and histograms on the diagonal.

sns.pairplot(ff5[['Mkt-RF', 'SMB', 'HML', 'RMW']],
             plot_kws={'alpha': 0.3, 's': 10},
             diag_kind='hist',
             height=2)
plt.suptitle('Pairwise Relationships Among Factors', y=1.02)
plt.show()

Faceting with `catplot`

catplot and relplot create grids of plots split by one or two categorical variables.

# Build a long-format frame that carries decade through the melt
ff5_long_dec = (ff5_demo[['Mkt-RF', 'SMB', 'HML', 'RMW', 'CMA', 'decade']]
                .melt(id_vars='decade', var_name='factor', value_name='return'))

sns.catplot(data=ff5_long_dec, x='factor', y='return', col='decade',
            kind='box', col_wrap=2, height=3.5, aspect=1.3, sharey=True)
plt.show()

The col_wrap=2 argument wraps the facets into a 2-column grid.
We melted with id_vars='decade' so decade survives the wide-to-long reshape.

Styling and Palettes

Themes

sns.set_theme(style=..., palette=...) configures the global look.

# Show a few palette options
for palette in ['deep', 'pastel', 'colorblind']:
    print(palette)
    sns.palplot(sns.color_palette(palette))

deep
pastel
colorblind

Per-Plot Palette

You can also pass palette= to most Seaborn functions for one-off control.

fig, ax = plt.subplots(figsize=(8, 4))
sns.boxplot(data=ff5_long, x='factor', y='return',
            hue='factor', palette='Set2', legend=False, ax=ax)
ax.axhline(0, color='red', linestyle='--', linewidth=0.8)
ax.set_title('Boxplot with Set2 Palette')
ax.set_ylabel('Return (%)')
plt.show()

Matplotlib + Seaborn Together

Seaborn returns Matplotlib Axes (for most functions), so you can mix the two freely.
Use Seaborn for the statistical layer, then add Matplotlib annotations on top.

fig, ax = plt.subplots(figsize=(8, 5))

# Seaborn regression layer
sns.regplot(data=ff5, x='Mkt-RF', y='SMB',
            scatter_kws={'alpha': 0.3, 's': 12},
            ax=ax)

# Matplotlib annotations on the same Axes
ax.axhline(0, color='black', linewidth=0.5)
ax.axvline(0, color='black', linewidth=0.5)
ax.annotate('Tech-bust cluster',
            xy=(-15, -10),
            xytext=(-25, -15),
            arrowprops={'arrowstyle': '->'})

ax.set_title('SMB vs Mkt-RF (Seaborn + Matplotlib)')
plt.show()

In-Class Exercise

14-1. Distribution Plot:

Use sns.histplot() with kde=True to plot the distribution of RMW.
Set bins=40 and give the plot a title and x-label.

14-2. Comparing Distributions:

Use sns.kdeplot() to overlay the densities of Mkt-RF, SMB, and HML on a single Axes.
Add a legend and an appropriate title.

14-3. Regression Plot:

Use sns.regplot() to plot Mkt-RF (x) against RMW (y).
Use semi-transparent markers (alpha=0.4) and a reasonable point size.
What sign does the fitted line have? (Brief written answer in a markdown comment is fine.)

14-4. Correlation Heatmap:

Compute the correlation matrix of the five factors ['Mkt-RF', 'SMB', 'HML', 'RMW', 'CMA'].
Plot it with sns.heatmap() using annot=True, cmap='RdBu_r', center=0, and vmin=-1, vmax=1.
Which two factors are most negatively correlated?

14-5. Categorical Comparison by Decade:

Create a decade column (e.g., '1990s', '2000s', ...) on a copy of ff5.
Use sns.barplot() to show the mean monthly Mkt-RF by decade with confidence intervals.
Add a zero-line with ax.axhline(0) for reference.

14-6. Pairplot:

Use sns.pairplot() to plot all pairwise relationships among Mkt-RF, SMB, HML, and CMA.
Use semi-transparent markers and histograms on the diagonal.

What We'll Cover

Ch 13: Matplotlib

Importing Matplotlib

Loading Our Data

The Anatomy of a Plot

Two Interfaces

Your First Plot

The pyplot Style

The Object-Oriented Style

Line Plots

Multiple Lines on One Axes

Cumulative Returns (Converting Percent to Decimal)

Plot Customization

Color, Line Style, Width, Markers

Reference Lines: axhline and axvline

Grid

Other Plot Types

Histogram

Scatter Plot

Bar Chart

Boxplot

Subplots

Saving Figures

Style Sheets

In-Class Exercise

Ch 14: Seaborn

Importing Seaborn

Distribution Plots

histplot(): Histograms with Density

kdeplot(): Compare Distributions

boxplot() and violinplot()

Relational Plots

scatterplot() with a hue

lineplot() with a confidence band

Regression Plots

regplot(): Scatter with Regression Line

lmplot(): Regression by Group

Categorical Plots

barplot(): Mean and Confidence Interval

countplot(): Frequency of Each Category

Matrix Plot: Correlation Heatmap

Multi-Plot Grids

pairplot(): All-Pairs Scatter Matrix

Faceting with catplot

Styling and Palettes

Themes

Per-Plot Palette

Matplotlib + Seaborn Together

In-Class Exercise

Reference Lines: `axhline` and `axvline`

`histplot()`: Histograms with Density

`kdeplot()`: Compare Distributions

`boxplot()` and `violinplot()`

`scatterplot()` with a `hue`

`lineplot()` with a confidence band

`regplot()`: Scatter with Regression Line

`lmplot()`: Regression by Group

`barplot()`: Mean and Confidence Interval

`countplot()`: Frequency of Each Category

`pairplot()`: All-Pairs Scatter Matrix

Faceting with `catplot`