File: summary.l1ce.Rd

package info (click to toggle)
r-cran-lasso2 1.2-20-1
  • links: PTS, VCS
  • area: main
  • in suites: bullseye, sid
  • size: 828 kB
  • sloc: ansic: 779; sh: 13; makefile: 2
file content (121 lines) | stat: -rw-r--r-- 4,841 bytes parent folder | download
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
% Copyright (C) 1998
% Berwin A. Turlach <bturlach@stats.adelaide.edu.au>
% Bill Venables <wvenable@stats.adelaide.edu.au>
% $Id: summary.l1ce.d,v 1.2 1998/09/27 02:02:02 bturlach Exp $
% --> ../COPYRIGHT for more details
\name{summary.l1ce}
\alias{summary.l1ce}
\alias{print.summary.l1ce}
\title{Summary Method for ``l1ce'' Objects (Regression with L1 Constraint)}
\description{
  Returns a summary list for a regression model with an L1 constraint on
  the parameters.  A null value will be returned if printing is invoked.
}
\usage{
\method{summary}{l1ce}(object, correlation = TRUE,
         type = c("OPT", "Tibshirani"),
         gen.inverse.diag = 0, sigma = NULL, \dots)
\method{print}{summary.l1ce}(x, digits = max(3, getOption("digits") - 3), \dots)
}
\arguments{
  \item{object}{fitted model of class \code{"l1ce"}.}
  \item{correlation}{logical indicating if the correlation matrix for
    the coefficients should be included in the summary.}
  \item{type}{character string specifying
    whether to use the covariance formula of Osborne, Presnell and Turlach
    or the formula of Tibshirani.}
  \item{gen.inverse.diag}{
    if Tibshirani's formula for the covariance matrix is used, this value
    is used for the diagonal elements of the generalised inverse that
    appears in the formula that corresponds to parameters estimated to be
    zero.  The default is 0, i.e. use the Moore-Penrose inverse.
    Tibshirani's code uses gen.inverse.diag=1e11.}
  \item{sigma}{
    the residual standard error estimate.  If not provided, then it is
    estimated by the deviance of the model divided by the error degrees of
    freedom.}
  \item{x}{an \R object of class \code{summary.l1ce}.}
  \item{digits}{number of significant digits to use.}
  \item{\dots}{further potential arguments passed to methods.}
}
\value{
  an object of class \code{summary.l1ce} (for which there's a
  \code{print} method).
  It is basically a list with the following components:

  \item{correlation}{the computed correlation coefficient matrix for the
    coefficients in the model.}
  \item{cov.unscaled}{the unscaled covariance matrix; i.e, a matrix such
    that multiplying it by an estimate of the error variance produces an
    estimated covariance matrix for the coefficients.
  }
  \item{df}{the number of degrees of freedom for the model and for residuals.}
  \item{coefficients}{a matrix with three columns, containing the
    coefficients, their standard errors and the corresponding t statistic.}
  \item{residuals}{the model residuals.  These are the weighted
    residuals if weights were given in the model.}
  \item{sigma}{the residual standard error estimate.}
  \item{terms}{the terms object used in fitting this model.}
  \item{call}{the call object used in fitting this model.}
  \item{bound}{the bound used in fitting this model.}
  \item{relative.bound}{the relative bound used in fitting this model
    (may not be present).}
  \item{Lagrangian}{the Lagrangian of the model.}
}
\details{
  This function is a method for the generic function
  \code{\link{summary}()} for class \code{"l1ce"}.
  It can be invoked by calling \code{summary(x)} for an
  object \code{x} of the appropriate class, or directly by
  calling \code{summary.l1ce(x)} regardless of the
  class of the object.
}
\seealso{
  \code{\link{l1ce}}, \code{\link{l1ce.object}}, \code{\link{summary}}.
}
\examples{%%- or just those in ./l1ce.Rd
data(Prostate)
summary(l1ce(lpsa ~ .,Prostate))

# Produces the following output:
\dontrun{
Call:
 l1ce(formula = lpsa ~ ., data = Prostate)
Residuals:
    Min      1Q Median    3Q  Max
 -1.636 -0.4119  0.076 0.452 1.83


Coefficients:
             Value Std. Error Z score Pr(>|Z|)
(Intercept) 0.7285 1.3898     0.5242  0.6002
     lcavol 0.4937 0.0919     5.3711  0.0000
    lweight 0.2682 0.1774     1.5115  0.1307
        age 0.0000 0.0111     0.0000  1.0000
       lbph 0.0093 0.0587     0.1581  0.8744
        svi 0.4551 0.2525     1.8023  0.0715
        lcp 0.0000 0.0947     0.0000  1.0000
    gleason 0.0000 0.1685     0.0000  1.0000
      pgg45 0.0002 0.0046     0.0391  0.9688


Residual standard error: 0.7595 on 88.36 degrees of freedom
The relative L1 bound was      : 0.5
The absolute L1 bound was      : 0.9219925
The Lagrangian for the bound is:  13.05806


Correlation of Coefficients:
        (Intercept)  lcavol lweight     age    lbph     svi     lcp gleason
 lcavol  0.1988
lweight -0.4815     -0.2071
    age -0.3938     -0.0603 -0.0974
   lbph  0.3629     -0.0201 -0.5165 -0.1303
    svi -0.0624     -0.2273 -0.1442  0.0635  0.0648
    lcp  0.0457     -0.4153  0.0598  0.0665  0.0632 -0.3779
gleason -0.7666     -0.2009  0.1163 -0.0774 -0.0617  0.1084 -0.0243
  pgg45  0.4988      0.0956 -0.0380 -0.0630 -0.1111 -0.1921 -0.2935 -0.6526
}
}
\keyword{regression}
% Converted by Sd2Rd version 1.21.