BayesianGaussianMixture
Variational Bayesian estimation of a Gaussian mixture.
This class allows to infer an approximate posterior distribution over the parameters of a Gaussian mixture distribution. The effective number of components can be inferred from the data.
This class implements two types of prior for the weights distribution: a finite mixture model with Dirichlet distribution and an infinite mixture model with the Dirichlet Process. In practice Dirichlet Process inference algorithm is approximated and uses a truncated distribution with a fixed maximum number of components (called the Stick-breaking representation). The number of components actually used almost always depends on the data.
Python Reference (opens in a new tab)
Constructors
constructor()
Signature
new BayesianGaussianMixture(opts?: object): BayesianGaussianMixture;Parameters
| Name | Type | Description |
|---|---|---|
opts? | object | - |
opts.covariance_prior? | number | ArrayLike | The prior on the covariance distribution (Wishart). If it is undefined, the emiprical covariance prior is initialized using the covariance of X. The shape depends on covariance\_type: |
opts.covariance_type? | "full" | "tied" | "diag" | "spherical" | String describing the type of covariance parameters to use. Must be one of: Default Value 'full' |
opts.degrees_of_freedom_prior? | number | The prior of the number of degrees of freedom on the covariance distributions (Wishart). If it is undefined, it’s set to n\_features. |
opts.init_params? | "k-means++" | "random" | "kmeans" | "random_from_data" | The method used to initialize the weights, the means and the covariances. String must be one of: Default Value 'kmeans' |
opts.max_iter? | number | The number of EM iterations to perform. Default Value 100 |
opts.mean_precision_prior? | number | The precision prior on the mean distribution (Gaussian). Controls the extent of where means can be placed. Larger values concentrate the cluster means around mean\_prior. The value of the parameter must be greater than 0. If it is undefined, it is set to 1. |
opts.mean_prior? | ArrayLike | The prior on the mean distribution (Gaussian). If it is undefined, it is set to the mean of X. |
opts.n_components? | number | The number of mixture components. Depending on the data and the value of the weight\_concentration\_prior the model can decide to not use all the components by setting some component weights\_ to values very close to zero. The number of effective components is therefore smaller than n_components. Default Value 1 |
opts.n_init? | number | The number of initializations to perform. The result with the highest lower bound value on the likelihood is kept. Default Value 1 |
opts.random_state? | number | Controls the random seed given to the method chosen to initialize the parameters (see init\_params). In addition, it controls the generation of random samples from the fitted distribution (see the method sample). Pass an int for reproducible output across multiple function calls. See Glossary. |
opts.reg_covar? | number | Non-negative regularization added to the diagonal of covariance. Allows to assure that the covariance matrices are all positive. Default Value 0.000001 |
opts.tol? | number | The convergence threshold. EM iterations will stop when the lower bound average gain on the likelihood (of the training data with respect to the model) is below this threshold. Default Value 0.001 |
opts.verbose? | number | Enable verbose output. If 1 then it prints the current initialization and each iteration step. If greater than 1 then it prints also the log probability and the time needed for each step. Default Value 0 |
opts.verbose_interval? | number | Number of iteration done before the next print. Default Value 10 |
opts.warm_start? | boolean | If ‘warm_start’ is true, the solution of the last fitting is used as initialization for the next call of fit(). This can speed up convergence when fit is called several times on similar problems. See the Glossary. Default Value false |
opts.weight_concentration_prior? | number | The dirichlet concentration of each component on the weight distribution (Dirichlet). This is commonly called gamma in the literature. The higher concentration puts more mass in the center and will lead to more components being active, while a lower concentration parameter will lead to more mass at the edge of the mixture weights simplex. The value of the parameter must be greater than 0. If it is undefined, it’s set to 1. / n\_components. |
opts.weight_concentration_prior_type? | "dirichlet_process" | "dirichlet_distribution" | String describing the type of the weight concentration prior. Default Value 'dirichlet_process' |
Returns
Defined in: generated/mixture/BayesianGaussianMixture.ts:25 (opens in a new tab)
Properties
_isDisposed
boolean=false
Defined in: generated/mixture/BayesianGaussianMixture.ts:23 (opens in a new tab)
_isInitialized
boolean=false
Defined in: generated/mixture/BayesianGaussianMixture.ts:22 (opens in a new tab)
_py
PythonBridge
Defined in: generated/mixture/BayesianGaussianMixture.ts:21 (opens in a new tab)
id
string
Defined in: generated/mixture/BayesianGaussianMixture.ts:18 (opens in a new tab)
opts
any
Defined in: generated/mixture/BayesianGaussianMixture.ts:19 (opens in a new tab)
Accessors
converged_
True when convergence was reached in fit(), false otherwise.
Signature
converged_(): Promise<boolean>;Returns
Promise<boolean>
Defined in: generated/mixture/BayesianGaussianMixture.ts:651 (opens in a new tab)
covariance_prior_
The prior on the covariance distribution (Wishart). The shape depends on covariance\_type:
Signature
covariance_prior_(): Promise<number | ArrayLike>;Returns
Promise<number | ArrayLike>
Defined in: generated/mixture/BayesianGaussianMixture.ts:921 (opens in a new tab)
covariances_
The covariance of each mixture component. The shape depends on covariance\_type:
Signature
covariances_(): Promise<ArrayLike>;Returns
Promise<ArrayLike>
Defined in: generated/mixture/BayesianGaussianMixture.ts:570 (opens in a new tab)
degrees_of_freedom_
The number of degrees of freedom of each components in the model.
Signature
degrees_of_freedom_(): Promise<ArrayLike>;Returns
Promise<ArrayLike>
Defined in: generated/mixture/BayesianGaussianMixture.ts:894 (opens in a new tab)
degrees_of_freedom_prior_
The prior of the number of degrees of freedom on the covariance distributions (Wishart).
Signature
degrees_of_freedom_prior_(): Promise<number>;Returns
Promise<number>
Defined in: generated/mixture/BayesianGaussianMixture.ts:867 (opens in a new tab)
feature_names_in_
Names of features seen during fit. Defined only when X has feature names that are all strings.
Signature
feature_names_in_(): Promise<ArrayLike>;Returns
Promise<ArrayLike>
Defined in: generated/mixture/BayesianGaussianMixture.ts:975 (opens in a new tab)
lower_bound_
Lower bound value on the model evidence (of the training data) of the best fit of inference.
Signature
lower_bound_(): Promise<number>;Returns
Promise<number>
Defined in: generated/mixture/BayesianGaussianMixture.ts:705 (opens in a new tab)
mean_precision_
The precision of each components on the mean distribution (Gaussian).
Signature
mean_precision_(): Promise<ArrayLike>;Returns
Promise<ArrayLike>
Defined in: generated/mixture/BayesianGaussianMixture.ts:813 (opens in a new tab)
mean_precision_prior_
The precision prior on the mean distribution (Gaussian). Controls the extent of where means can be placed. Larger values concentrate the cluster means around mean\_prior. If mean_precision_prior is set to undefined, mean\_precision\_prior\_ is set to 1.
Signature
mean_precision_prior_(): Promise<number>;Returns
Promise<number>
Defined in: generated/mixture/BayesianGaussianMixture.ts:786 (opens in a new tab)
mean_prior_
The prior on the mean distribution (Gaussian).
Signature
mean_prior_(): Promise<ArrayLike>;Returns
Promise<ArrayLike>
Defined in: generated/mixture/BayesianGaussianMixture.ts:840 (opens in a new tab)
means_
The mean of each mixture component.
Signature
means_(): Promise<ArrayLike[]>;Returns
Promise<ArrayLike[]>
Defined in: generated/mixture/BayesianGaussianMixture.ts:543 (opens in a new tab)
n_features_in_
Number of features seen during fit.
Signature
n_features_in_(): Promise<number>;Returns
Promise<number>
Defined in: generated/mixture/BayesianGaussianMixture.ts:948 (opens in a new tab)
n_iter_
Number of step used by the best fit of inference to reach the convergence.
Signature
n_iter_(): Promise<number>;Returns
Promise<number>
Defined in: generated/mixture/BayesianGaussianMixture.ts:678 (opens in a new tab)
precisions_
The precision matrices for each component in the mixture. A precision matrix is the inverse of a covariance matrix. A covariance matrix is symmetric positive definite so the mixture of Gaussian can be equivalently parameterized by the precision matrices. Storing the precision matrices instead of the covariance matrices makes it more efficient to compute the log-likelihood of new samples at test time. The shape depends on covariance\_type:
Signature
precisions_(): Promise<ArrayLike>;Returns
Promise<ArrayLike>
Defined in: generated/mixture/BayesianGaussianMixture.ts:597 (opens in a new tab)
precisions_cholesky_
The cholesky decomposition of the precision matrices of each mixture component. A precision matrix is the inverse of a covariance matrix. A covariance matrix is symmetric positive definite so the mixture of Gaussian can be equivalently parameterized by the precision matrices. Storing the precision matrices instead of the covariance matrices makes it more efficient to compute the log-likelihood of new samples at test time. The shape depends on covariance\_type:
Signature
precisions_cholesky_(): Promise<ArrayLike>;Returns
Promise<ArrayLike>
Defined in: generated/mixture/BayesianGaussianMixture.ts:624 (opens in a new tab)
py
Signature
py(): PythonBridge;Returns
PythonBridge
Defined in: generated/mixture/BayesianGaussianMixture.ts:139 (opens in a new tab)
Signature
py(pythonBridge: PythonBridge): void;Parameters
| Name | Type |
|---|---|
pythonBridge | PythonBridge |
Returns
void
Defined in: generated/mixture/BayesianGaussianMixture.ts:143 (opens in a new tab)
weight_concentration_
The dirichlet concentration of each component on the weight distribution (Dirichlet).
Signature
weight_concentration_(): Promise<ArrayLike>;Returns
Promise<ArrayLike>
Defined in: generated/mixture/BayesianGaussianMixture.ts:759 (opens in a new tab)
weight_concentration_prior_
The dirichlet concentration of each component on the weight distribution (Dirichlet). The type depends on weight\_concentration\_prior\_type:
Signature
weight_concentration_prior_(): Promise<number>;Returns
Promise<number>
Defined in: generated/mixture/BayesianGaussianMixture.ts:732 (opens in a new tab)
weights_
The weights of each mixture components.
Signature
weights_(): Promise<ArrayLike>;Returns
Promise<ArrayLike>
Defined in: generated/mixture/BayesianGaussianMixture.ts:516 (opens in a new tab)
Methods
dispose()
Disposes of the underlying Python resources.
Once dispose() is called, the instance is no longer usable.
Signature
dispose(): Promise<void>;Returns
Promise<void>
Defined in: generated/mixture/BayesianGaussianMixture.ts:220 (opens in a new tab)
fit()
Estimate model parameters with the EM algorithm.
The method fits the model n\_init times and sets the parameters with which the model has the largest likelihood or lower bound. Within each trial, the method iterates between E-step and M-step for max\_iter times until the change of likelihood or lower bound is less than tol, otherwise, a ConvergenceWarning is raised. If warm\_start is true, then n\_init is ignored and a single initialization is performed upon the first call. Upon consecutive calls, training starts where it left off.
Signature
fit(opts: object): Promise<any>;Parameters
| Name | Type | Description |
|---|---|---|
opts | object | - |
opts.X? | ArrayLike[] | List of n_features-dimensional data points. Each row corresponds to a single data point. |
opts.y? | any | Not used, present for API consistency by convention. |
Returns
Promise<any>
Defined in: generated/mixture/BayesianGaussianMixture.ts:239 (opens in a new tab)
fit_predict()
Estimate model parameters using X and predict the labels for X.
The method fits the model n_init times and sets the parameters with which the model has the largest likelihood or lower bound. Within each trial, the method iterates between E-step and M-step for max\_iter times until the change of likelihood or lower bound is less than tol, otherwise, a ConvergenceWarning is raised. After fitting, it predicts the most probable label for the input data points.
Signature
fit_predict(opts: object): Promise<any>;Parameters
| Name | Type | Description |
|---|---|---|
opts | object | - |
opts.X? | ArrayLike[] | List of n_features-dimensional data points. Each row corresponds to a single data point. |
opts.y? | any | Not used, present for API consistency by convention. |
Returns
Promise<any>
Defined in: generated/mixture/BayesianGaussianMixture.ts:281 (opens in a new tab)
init()
Initializes the underlying Python resources.
This instance is not usable until the Promise returned by init() resolves.
Signature
init(py: PythonBridge): Promise<void>;Parameters
| Name | Type |
|---|---|
py | PythonBridge |
Returns
Promise<void>
Defined in: generated/mixture/BayesianGaussianMixture.ts:152 (opens in a new tab)
predict()
Predict the labels for the data samples in X using trained model.
Signature
predict(opts: object): Promise<any>;Parameters
| Name | Type | Description |
|---|---|---|
opts | object | - |
opts.X? | ArrayLike[] | List of n_features-dimensional data points. Each row corresponds to a single data point. |
Returns
Promise<any>
Defined in: generated/mixture/BayesianGaussianMixture.ts:324 (opens in a new tab)
predict_proba()
Evaluate the components’ density for each sample.
Signature
predict_proba(opts: object): Promise<any>;Parameters
| Name | Type | Description |
|---|---|---|
opts | object | - |
opts.X? | ArrayLike[] | List of n_features-dimensional data points. Each row corresponds to a single data point. |
Returns
Promise<any>
Defined in: generated/mixture/BayesianGaussianMixture.ts:361 (opens in a new tab)
sample()
Generate random samples from the fitted Gaussian distribution.
Signature
sample(opts: object): Promise<any>;Parameters
| Name | Type | Description |
|---|---|---|
opts | object | - |
opts.n_samples? | number | Number of samples to generate. Default Value 1 |
Returns
Promise<any>
Defined in: generated/mixture/BayesianGaussianMixture.ts:399 (opens in a new tab)
score()
Compute the per-sample average log-likelihood of the given data X.
Signature
score(opts: object): Promise<number>;Parameters
| Name | Type | Description |
|---|---|---|
opts | object | - |
opts.X? | ArrayLike[] | List of n_features-dimensional data points. Each row corresponds to a single data point. |
opts.y? | any | Not used, present for API consistency by convention. |
Returns
Promise<number>
Defined in: generated/mixture/BayesianGaussianMixture.ts:438 (opens in a new tab)
score_samples()
Compute the log-likelihood of each sample.
Signature
score_samples(opts: object): Promise<any>;Parameters
| Name | Type | Description |
|---|---|---|
opts | object | - |
opts.X? | ArrayLike[] | List of n_features-dimensional data points. Each row corresponds to a single data point. |
Returns
Promise<any>
Defined in: generated/mixture/BayesianGaussianMixture.ts:478 (opens in a new tab)