Probabilistic Models

Variational CCA

class cca_zoo.probabilisticmodels.probabilisticcca.ProbabilisticCCA(latent_dims=1, copy_data=True, random_state=0, num_samples=100, num_warmup=100)

Bases: cca_zoo.models._cca_base._CCA_Base

A class used to fit a Probabilistic CCA. Not quite the same due to using VI methods rather than EM

Citation:

Bach, Francis R., and Michael I. Jordan. “A probabilistic interpretation of canonical correlation analysis.” (2005). Wang, Chong. “Variational Bayesian approach to canonical correlation analysis.” IEEE Transactions on Neural Networks 18.3 (2007): 905-910.

fit(views, y=None, **kwargs)

Infer the parameters (mu: mean, psi: within view variance) and latent variables (z) of the generative CCA model

Parameters:views (Iterable[ndarray]) – list/tuple of numpy arrays or array likes with the same number of rows (samples)

transform(views, y=None, **kwargs)

Predict the latent variables that generate the data in views using the sampled model parameters

Parameters:views (Iterable[ndarray]) – list/tuple of numpy arrays or array likes with the same number of rows (samples)

correlations(views, **kwargs)

Predicts the correlations between each view for each dimension for the given data using the fit model

Parameters:

• views (Iterable[ndarray]) – list/tuple of numpy arrays or array likes with the same number of rows (samples)
• kwargs – any additional keyword arguments required by the given model

Returns:

all_corrs: an array of the pairwise correlations (k,k,self.latent_dims) where k is the number of views

fit_transform(views, **kwargs)

Fits and then transforms the training data

Parameters:

• views (Iterable[ndarray]) – list/tuple of numpy arrays or array likes with the same number of rows (samples)
• kwargs – any additional keyword arguments required by the given model

get_loadings(views, normalize=False, **kwargs)

Returns the model loadings for each view for the given data

Parameters:

• views (Iterable[ndarray]) – list/tuple of numpy arrays or array likes with the same number of rows (samples)
• kwargs – any additional keyword arguments required by the given model
• normalize – scales loadings to ensure that they represent correlations between features and scores

score(views, y=None, **kwargs)

Returns average correlation in each dimension (averages over all pairs for multiview)

Parameters:

• views (Iterable[ndarray]) – list/tuple of numpy arrays or array likes with the same number of rows (samples)
• y – unused but needed to integrate with scikit-learn