Scikit-Learn make_swiss_roll() Dataset

The Swiss Roll dataset is a synthetic dataset useful for visualizing and testing manifold learning algorithms.

Key arguments when generating the dataset include n_samples to specify the number of samples, noise to add Gaussian noise, and random_state for reproducibility.

This dataset is suitable for clustering and dimensionality reduction algorithms such as k-means and t-SNE.

from sklearn.datasets import make_swiss_roll
import matplotlib.pyplot as plt
import pandas as pd

# Generate the dataset
n_samples = 1000
noise = 0.2
X, t = make_swiss_roll(n_samples=n_samples, noise=noise, random_state=42)

# Display dataset shape and types
print(f"Dataset shape: {X.shape}")
print(f"Feature types: {type(X)}")

# Show summary statistics
print(f"Summary statistics:\n{pd.DataFrame(X).describe()}")

# Plot the dataset
fig = plt.figure(figsize=(10, 7))
ax = fig.add_subplot(111, projection='3d')
ax.scatter(X[:, 0], X[:, 1], X[:, 2], c=t, cmap=plt.cm.Spectral)
plt.title("Swiss Roll Dataset")
plt.show()

Running the example gives an output like:

Dataset shape: (1000, 3)
Feature types: <class 'numpy.ndarray'>
Summary statistics:
                 0            1            2
count  1000.000000  1000.000000  1000.000000
mean      2.004806    10.647835     0.154515
std       6.571050     6.149676     6.913291
min      -9.919520    -0.144537   -11.416802
25%      -3.386741     5.106111    -4.763130
50%       3.317279    10.890549    -0.511495
75%       6.315301    15.915352     6.430505
max      12.928036    21.176402    14.360618

The steps are as follows:

1. Import the make_swiss_roll function from sklearn.datasets and matplotlib.pyplot for plotting:

   • This function generates the Swiss Roll dataset, and matplotlib is used for visualization.

2. Generate the dataset using make_swiss_roll():

   • Specify the number of samples (n_samples), noise level (noise), and random state (random_state).

3. Print the dataset shape and feature types:

   • Access the shape using X.shape.
   • Show the data type of the features, which is typically a numpy array.

4. Display summary statistics:

   • Use pd.DataFrame(X).describe() to get a statistical summary of the dataset.

5. Plot the dataset:

   • Create a 3D scatter plot using matplotlib to visualize the Swiss Roll structure.
   • Use the c parameter to color the points based on the t values.

This example demonstrates how to quickly generate and visualize the Swiss Roll dataset using scikit-learn’s make_swiss_roll function. This is useful for testing clustering and dimensionality reduction algorithms.