Plot Concentration Prior

Introduction

This lab demonstrates how to use the scikit-learn BayesianGaussianMixture class to fit a toy dataset containing a mixture of three Gaussians. The class can adapt its number of mixture components automatically using a concentration prior, which is specified using the weight_concentration_prior_type parameter. This lab shows the difference between using a Dirichlet distribution prior and a Dirichlet process prior to select the number of components with non-zero weights.

VM Tips

After the VM startup is done, click the top left corner to switch to the Notebook tab to access Jupyter Notebook for practice.

Sometimes, you may need to wait a few seconds for Jupyter Notebook to finish loading. The validation of operations cannot be automated because of limitations in Jupyter Notebook.

If you face issues during learning, feel free to ask Labby. Provide feedback after the session, and we will promptly resolve the problem for you.

Skills Graph

Import Libraries

In this step, we will import the necessary libraries, which are numpy, matplotlib, gridspec, and BayesianGaussianMixture from sklearn.mixture.

import numpy as np
import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec
from sklearn.mixture import BayesianGaussianMixture

Define Functions

In this step, we define two functions. The first function plots ellipsoids obtained from the toy dataset fitted by the BayesianGaussianMixture class models. The second function plots the results for three different values of the weight concentration prior.

def plot_ellipses(ax, weights, means, covars):
    for n in range(means.shape[0]):
        eig_vals, eig_vecs = np.linalg.eigh(covars[n])
        unit_eig_vec = eig_vecs[0] / np.linalg.norm(eig_vecs[0])
        angle = np.arctan2(unit_eig_vec[1], unit_eig_vec[0])
        angle = 180 * angle / np.pi
        eig_vals = 2 * np.sqrt(2) * np.sqrt(eig_vals)
        ell = mpl.patches.Ellipse(
            means[n], eig_vals[0], eig_vals[1], angle=180 + angle, edgecolor="black"
        )
        ell.set_clip_box(ax.bbox)
        ell.set_alpha(weights[n])
        ell.set_facecolor("#56B4E9")
        ax.add_artist(ell)

def plot_results(ax1, ax2, estimator, X, y, title, plot_title=False):
    ax1.set_title(title)
    ax1.scatter(X[:, 0], X[:, 1], s=5, marker="o", color=colors[y], alpha=0.8)
    ax1.set_xlim(-2.0, 2.0)
    ax1.set_ylim(-3.0, 3.0)
    ax1.set_xticks(())
    ax1.set_yticks(())
    plot_ellipses(ax1, estimator.weights_, estimator.means_, estimator.covariances_)

    ax2.get_xaxis().set_tick_params(direction="out")
    ax2.yaxis.grid(True, alpha=0.7)
    for k, w in enumerate(estimator.weights_):
        ax2.bar(
            k,
            w,
            width=0.9,
            color="#56B4E9",
            zorder=3,
            align="center",
            edgecolor="black",
        )
        ax2.text(k, w + 0.007, "%.1f%%" % (w * 100.0), horizontalalignment="center")
    ax2.set_xlim(-0.6, 2 * n_components - 0.4)
    ax2.set_ylim(0.0, 1.1)
    ax2.tick_params(axis="y", which="both", left=False, right=False, labelleft=False)
    ax2.tick_params(axis="x", which="both", top=False)
    if plot_title:
        ax1.set_ylabel("Estimated Mixtures")
        ax2.set_ylabel("Weight of each component")

Set Parameters for the Toy Dataset

In this step, we set the parameters for the toy dataset, which include the random state, number of components, number of features, colors, covariances, samples, and means.

random_state, n_components, n_features = 2, 3, 2
colors = np.array(["#0072B2", "#F0E442", "#D55E00"])
covars = np.array(
    [[[0.7, 0.0], [0.0, 0.1]], [[0.5, 0.0], [0.0, 0.1]], [[0.5, 0.0], [0.0, 0.1]]]
)
samples = np.array([200, 500, 200])
means = np.array([[0.0, -0.70], [0.0, 0.0], [0.0, 0.70]])

Define Estimators

In this step, we define two estimators. The first estimator uses a Dirichlet distribution prior to set the number of components with non-zero weights. The second estimator uses a Dirichlet process prior to select the number of components.

estimators = [
    (
        "Finite mixture with a Dirichlet distribution\nprior and " r"$\gamma_0=$",
        BayesianGaussianMixture(
            weight_concentration_prior_type="dirichlet_distribution",
            n_components=2 * n_components,
            reg_covar=0,
            init_params="random",
            max_iter=1500,
            mean_precision_prior=0.8,
            random_state=random_state,
        ),
        [0.001, 1, 1000],
    ),
    (
        "Infinite mixture with a Dirichlet process\n prior and" r"$\gamma_0=$",
        BayesianGaussianMixture(
            weight_concentration_prior_type="dirichlet_process",
            n_components=2 * n_components,
            reg_covar=0,
            init_params="random",
            max_iter=1500,
            mean_precision_prior=0.8,
            random_state=random_state,
        ),
        [1, 1000, 100000],
    ),
]

Generate Data

In this step, we generate data using the numpy.random.RandomState function and the parameters defined in Step 3.

rng = np.random.RandomState(random_state)
X = np.vstack(
    [
        rng.multivariate_normal(means[j], covars[j], samples[j])
        for j in range(n_components)
    ]
)
y = np.concatenate([np.full(samples[j], j, dtype=int) for j in range(n_components)])

Plot Results

In this step, we plot the results for each estimator using the plot_results function defined in Step 2.

for title, estimator, concentrations_prior in estimators:
    plt.figure(figsize=(4.7 * 3, 8))
    plt.subplots_adjust(
        bottom=0.04, top=0.90, hspace=0.05, wspace=0.05, left=0.03, right=0.99
    )

    gs = gridspec.GridSpec(3, len(concentrations_prior))
    for k, concentration in enumerate(concentrations_prior):
        estimator.weight_concentration_prior = concentration
        estimator.fit(X)
        plot_results(
            plt.subplot(gs[0:2, k]),
            plt.subplot(gs[2, k]),
            estimator,
            X,
            y,
            r"%s$%.1e$" % (title, concentration),
            plot_title=k == 0,
        )

plt.show()

Summary

This lab demonstrated how to use the BayesianGaussianMixture class in scikit-learn to fit a toy dataset containing a mixture of three Gaussians. The class can adapt its number of mixture components automatically using a concentration prior, which is specified using the weight_concentration_prior_type parameter. This lab showed the difference between using a Dirichlet distribution prior and a Dirichlet process prior to select the number of components with non-zero weights.