Class PLSRegression

source code

     object --+        
              |        
  BaseEstimator --+    
                  |    
RegressionEstimator --+
                      |
                     PLSRegression

Estimator for PLS regression

    f(beta, X, Y) = -Cov(X.beta, Y),

where Cov(., .) is the covariance.

Parameters
----------
K : Positive integer. The number of components to compute.

algorithm : OR(ImplicitAlgorithm, ExplicitAlgorithm). The algorithm that
        should be used. Should be one of:
            1. PLSR()
            2. MultiblockFISTA(...)

        Default is PLSR(...).

algorithm_params : A dict. The dictionary algorithm_params contains
        parameters that should be set in the algorithm. Passing
        algorithm=MyAlgorithm(**params) is equivalent to passing
        algorithm=MyAlgorithm() and algorithm_params=params. Default
        is an empty dictionary.

start_vector : BaseStartVector. Generates the start vector that will be
        used.

mean : Boolean. Whether or not to compute the means squared error or the
        squared error. Default is True, compute the means squared error.

Examples
--------
>>> import parsimony.estimators as estimators
>>> import parsimony.algorithms.nipals as nipals
>>> import parsimony.algorithms.multiblock as multiblock
>>> import numpy as np
>>> np.random.seed(42)
>>>
>>> n, p = 16, 10
>>> X = np.random.rand(n, p)
>>> y = np.random.rand(n, 1)
>>> plsr = estimators.PLSRegression(K=4, algorithm=nipals.PLSR(),
...                                 algorithm_params=dict(max_iter=100))
>>> error = plsr.fit(X, y).score(X, y)
>>> print "error = ", error
error =  0.0222345224457