randomized.slope

Module: randomized.slope

Inheritance diagram for selectinf.randomized.slope:

digraph inheritance7910e6f37d { rankdir=LR; size="8.0, 12.0"; "randomized.query.gaussian_query" [URL="selectinf.randomized.query.html#selectinf.randomized.query.gaussian_query",fontname="Vera Sans, DejaVu Sans, Liberation Sans, Arial, Helvetica, sans",fontsize=10,height=0.25,shape=box,style="setlinewidth(0.5)",target="_top"]; "randomized.query.query" -> "randomized.query.gaussian_query" [arrowsize=0.5,style="setlinewidth(0.5)"]; "randomized.query.query" [URL="selectinf.randomized.query.html#selectinf.randomized.query.query",fontname="Vera Sans, DejaVu Sans, Liberation Sans, Arial, Helvetica, sans",fontsize=10,height=0.25,shape=box,style="setlinewidth(0.5)",target="_top",tooltip="This class is the base of randomized selective inference"]; "randomized.slope.slope" [URL="#selectinf.randomized.slope.slope",fontname="Vera Sans, DejaVu Sans, Liberation Sans, Arial, Helvetica, sans",fontsize=10,height=0.25,shape=box,style="setlinewidth(0.5)",target="_top"]; "randomized.query.gaussian_query" -> "randomized.slope.slope" [arrowsize=0.5,style="setlinewidth(0.5)"]; }

slope

class selectinf.randomized.slope.slope(loglike, slope_weights, ridge_term, randomizer, perturb=None)

Bases: selectinf.randomized.query.gaussian_query

__init__(loglike, slope_weights, ridge_term, randomizer, perturb=None)

Create a new post-selection object for the SLOPE problem

Parameters

loglike : regreg.smooth.glm.glm

A (negative) log-likelihood as implemented in regreg.

slope_weights : np.ndarray

SLOPE weights for L-1 penalty. If a float, it is broadcast to all features.

ridge_term : float

How big a ridge term to add?

randomizer : object

Randomizer – contains representation of randomization density.

perturb : np.ndarray

Random perturbation subtracted as a linear term in the objective function.

fit(solve_args={'min_its': 50, 'tol': 1e-12}, perturb=None)

static gaussian(X, Y, slope_weights, sigma=1.0, quadratic=None, ridge_term=0.0, randomizer_scale=None)

get_sampler()

randomize(perturb=None)

The actual randomization step.

Parameters

perturb : ndarray, optional

Value of randomization vector, an instance of \(\omega\).

property sampler

Sampler of optimization (augmented) variables.

selective_MLE(observed_target, target_cov, target_score_cov, level=0.9, solve_args={'tol': 1e-12})

Parameters

observed_target : ndarray

Observed estimate of target.

target_cov : ndarray

Estimated covaraince of target.

target_score_cov : ndarray

Estimated covariance of target and score of randomized query.

level : float, optional

Confidence level.

solve_args : dict, optional

Arguments passed to solver.

set_sampler(sampler)

setup_sampler()

Setup query to prepare for sampling. Should set a few key attributes:

• observed_score_state

• observed_opt_state

• opt_transform

solve()

summary(observed_target, target_cov, target_score_cov, alternatives, opt_sample=None, target_sample=None, parameter=None, level=0.9, ndraw=10000, burnin=2000, compute_intervals=False)

Produce p-values and confidence intervals for targets of model including selected features

Parameters

target : one of [‘selected’, ‘full’]

features : np.bool

Binary encoding of which features to use in final model and targets.

parameter : np.array

Hypothesized value for parameter – defaults to 0.

level : float

Confidence level.

ndraw : int (optional)

Defaults to 1000.

burnin : int (optional)

Defaults to 1000.

compute_intervals : bool

Compute confidence intervals?

dispersion : float (optional)

Use a known value for dispersion, or Pearson’s X^2?

useC = True