direpack.sudire.sudire.sudire

class sudire(sudiremeth='dcov-sdr', n_components=2, trimming=0, optimizer_options={'max_iter': 1000}, optimizer_constraints=None, optimizer_arguments=None, optimizer_start=None, center_data=True, center='mean', scale_data=True, whiten_data=False, compression=False, n_slices=6, dmetric='euclidean', fit_ols=True, copy=True, response_type='continuous', verbose=True, return_scaling_object=True)[source]

SUDIRE Sufficient Dimension Reduction

The class allows for Sufficient Dimension Reduction using a variety of methods. If the method requires optimization of a function, This optimization is done through the Interior Point Optimizer (IPOPT) algorithm.

Parameters

  • sudiremeth (function or class. sudiremeth in this package can also be used,) –

  • are (but user defined functions can be processed. Built in options) –

    save : Sliced Average Variance Estimation

    sir : Slices Inverse Regression

    dr : Directional Regression

    iht : Iterative Hessian Transformations

    dcov-sdr : SDR via Distance Covariance

    mdd-sdr : SDR via Martingale Difference Divergence.

    bcov-sdr : SDR via ball covariance

  • n_components (int) – dimension of the central subspace.

  • trimming (float) – trimming percentage to be entered as pct/100

  • optimizer_options (dict) – with options to pass on to the optimizer.Includes:

  • max_iter (int) – Maximal number of iterations.

  • tol (float) – relative convergence tolerance

  • constr_viol_tol (float) – Desired threshold for the constraint violation.

  • optimizer_constraints (dict or list of dicts) – further constraints to be passed on to the optimizer function.

  • optimizer_arguments (dict) – extra arguments to be passed to the sudiremeth function during optimization.

  • optimizer_start (numpy array) – starting value for the optimization.

  • center (str) – how to center the data. options accepted are options from sprm.preprocessing

  • center_data (bool) – If True, the data will be centered before the dimension reduction

  • scale_data (bool) – if set to False, convergence to correct optimum is not a given. Will throw a warning.

  • compression (bool) – Use internal data compresion step for flat data.

  • n_slices (int) – The number of slices for SAVE, SIR, DR

  • is_distance_mat (bool) – if the inputed matrices for x and y are distance matrices.

  • dmetric (str) – distance metric used internally. Defaults to ‘euclidean’

  • fit_ols (bool) – if True, an OLS model is fitted after the dimension reduction.

  • copy (bool) – Whether to make a deep copy of the input data or not.

  • verbose (bool) – Set to True prints the iteration number.

  • return_scaling_object (bool.) – If True, the scaling object will be return after the dimension reduction.

Attributes always provided

  • x_loadings_: Estimated basis of the central subspace

  • x_scores_: The projected X data.

  • x_loc_: location estimate for X

  • x_sca_: scale estimate for X

  • ` ols_obj` : fitted OLS objected

  • y_loc_: y location estimate

  • y_sca_: y scale estimate

Attributes created only when corresponding input flags are `True`

  • whitening_: whitened data matrix (usually denoted K)

  • scaling_object_: scaling object from VersatileScaler

__init__(sudiremeth='dcov-sdr', n_components=2, trimming=0, optimizer_options={'max_iter': 1000}, optimizer_constraints=None, optimizer_arguments=None, optimizer_start=None, center_data=True, center='mean', scale_data=True, whiten_data=False, compression=False, n_slices=6, dmetric='euclidean', fit_ols=True, copy=True, response_type='continuous', verbose=True, return_scaling_object=True)[source]

Methods

__init__([sudiremeth, n_components, ...])

fit(X, y, *args, **kwargs)

Fit a Sufficient Dimension Reduction Model.

fit_transform(X[, y])

Fit to data, then transform it.

get_params([deep])

Get parameters for this estimator. :param deep: If True, will return the parameters for this estimator and contained subobjects that are estimators. :type deep: boolean, optional.

predict(Xn[, is_distance_mat])

predicts the response on new data Xn

score(X, y[, sample_weight])

Return the coefficient of determination of the prediction.

set_params(**params)

Set the parameters of this estimator.

transform(Xn[, distance_mat])

Computes the dimension reduction of the data Xn based on the fitted sudire model.

Attributes