Master Python Matplotlib: Scatter Plots, Polynomial Fit, and Multi‑Panel Visualizations

Recently I have been reading the first two chapters of Machine Learning Systems Design and learned several useful Matplotlib data‑visualization methods, which I summarize here.

Disclaimer: Most code is adapted from the book examples; only illustrative snippets are provided.

1. Scatter Plot

Use plt.scatter(x, y) to draw a scatter plot; a continuous curve can be drawn with plt.plot(x, y). Customize x‑axis ticks with plt.xticks(loc, label) and automatically adjust the figure with plt.autoscale(tight=True).

2. Polynomial Fitting and Curve Plotting

Fit a polynomial to the data and plot the fitted curve (example image shown).

3. Multiple Subplots with the Iris Dataset

The Iris dataset contains four features: sepal length, sepal width, petal length, and petal width. By generating all pairwise combinations of these features, we can plot each pair in a separate subplot.

Example code for generating the combinations and plotting:

for i, k in enumerate(feature_names_2[1]):
    index1 = feature_names.index(k[0])
    plt.subplot(2, 3, 1 + i)
    for t, marker, c in zip(range(3), ">ox", "rgb"):
        # plotting logic here
        pass

4. Drawing Horizontal and Vertical Reference Lines

Use plt.vlines(x, y_min, y_max) for vertical lines and plt.hlines(y, x_min, x_max) for horizontal lines to separate classes or highlight thresholds.

5. Dynamic (Interactive) Plotting

Enable interactive mode with plt.ion(); plt.show() will no longer block execution. Remember to adjust the axes with plt.axis() as needed.