lasso (Series)

Least absolute shrinkage and selection operator

Syntax

[B, BStd, Residuals, EStd, Fitted, Range, BCov] = lasso(Y, X, ~Range, ...)

Input arguments marked with a ~ sign may be omitted.

Input arguments

Y [ TimeSubscriptable ]

Time series of left-hand-side (dependent) observations.

X [ TimeSubscriptable ]

Time series of right-hand-side (independent) observations.

~Range [ numeric ]

Date range on which the lasso will be run; if omitted, the entire range available will be used.

Output arguments

B [ numeric ]

Vector of estimated lasso coefficients.

BStd [ numeric ]

Vector of std errors of the estimates.

Residuals [ TimeSubscriptable ]

Time series with the lasso residuals.

EStd [ numeric ]

Estimate of the std deviation of the lasso residuals.

Fitted [ TimeSubscriptable ]

Time series with fitted LHS observations.

Range [ numeric ]

The actually used date range.

bBCov [ numeric ]

Covariance matrix of the coefficient estimates.

Options

Intercept=false [ true | false ]

Include an intercept in the lasso; if true the constant will be placed last in the matrix of lassoors.

Weighting=[ ] [ TimeSubscriptable | numeric | empty ]

Time series with weights on observations in individual periods, or a discount factor for weighting the observations from the most recent to the most distant.

Description

This function calls the built-in lscov function.

Examples