Description Usage Arguments Value References Examples
Returns an object of class "l1ce"
or "licelist"
that represents
fit(s) of linear models while imposing L1 constraint(s) on the parameters.
1 2 3 4 5 |
formula |
a formula object, with the response on the left of a
|
data |
a |
weights |
vector of observation weights. The length of |
subset |
expression saying which subset of the rows of the data should be used in the fit. This can be a logical vector (which is replicated to have length equal to the number of observations), or a numeric vector indicating which observation numbers are to be included, or a character vector of the row names to be included. All observations are included by default. |
na.action |
a function to filter missing data. This is applied to the
|
sweep.out |
a formula object, variables whose parameters are not put under the
constraint are swept out first. The variables should appear on the
right of a |
x |
logical indicating if the model matrix should be returned in
component |
y |
logical indicating if the response should be returned in
component |
contrasts |
a list giving contrasts for some or all of the factors appearing in the model formula. The elements of the list should have the same name as the variable and should be either a contrast matrix (specifically, any full-rank matrix with as many rows as there are levels in the factor), or else a function to compute such a matrix given the number of levels. |
standardize |
logical flag: if |
trace |
logical flag: if |
guess.constrained.coefficients |
initial guess for the parameters that are constrained. |
bound |
numeric, either a single number or a vector: the constraint(s) that is/are put onto the L1 norm of the parameters. |
absolute.t |
logical flag: if |
an object of class l1ce
(if bound
was a single value) or
l1celist
(if bound
was a vector of values) is returned.
See l1ce.object
and l1celist.object
for details.
Osborne, M.R., Presnell, B. and Turlach, B.A. (2000) On the LASSO and its Dual, Journal of Computational and Graphical Statistics 9(2), 319–337.
Tibshirani, R. (1996) Regression shrinkage and selection via the lasso, Journal of the Royal Statistical Society, Series B 58(1), 267–288.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 | data(Iowa)
l1c.I <- l1ce(Yield ~ ., Iowa, bound = 10, absolute.t=TRUE)
l1c.I
## The same, printing information in each step:
l1ce(Yield ~ ., Iowa, bound = 10, trace = TRUE, absolute.t=TRUE)
data(Prostate)
l1c.P <- l1ce(lpsa ~ ., Prostate, bound=(1:30)/30)
length(l1c.P)# 30 l1ce models
l1c.P # -- MM: too large; should do this in summary(.)!
plot(resid(l1c.I) ~ fitted(l1c.I))
abline(h = 0, lty = 3, lwd = .2)
|
R Package to solve regression problems while imposing
an L1 constraint on the parameters. Based on S-plus Release 2.1
Copyright (C) 1998, 1999
Justin Lokhorst <jlokhors@stats.adelaide.edu.au>
Berwin A. Turlach <bturlach@stats.adelaide.edu.au>
Bill Venables <wvenable@stats.adelaide.edu.au>
Copyright (C) 2002
Martin Maechler <maechler@stat.math.ethz.ch>
Call:
l1ce(formula = Yield ~ ., data = Iowa, bound = 10, absolute.t = TRUE)
Coefficients:
(Intercept) Year Rain0 Temp1 Rain1
-1.223687e+03 6.665384e-01 8.895341e-03 0.000000e+00 0.000000e+00
Temp2 Rain2 Temp3 Rain3 Temp4
0.000000e+00 1.700162e+00 -1.612996e-01 0.000000e+00 -2.385012e-01
The absolute L1 bound was : 10
The Lagrangian for the bound is: 69.63196
******************************
--> Adding variable: 1
******************************
Iteration number: 0
Value of primal object function : 2782.530000
Value of dual object function : -384.942354
L1 norm of current beta : 0.000000 <= 10.000000
Maximal absolute value in t(X)%*%r : 3.167472e+02 attained 1 time(s)
Number of parameters allowed to vary : 1
******************************
Iteration number: 1
Value of primal object function : 1214.892326
Value of dual object function : 25.579582
L1 norm of current beta : 9.898351 <= 10.000000
Maximal absolute value in t(X)%*%r : 1.189313e+02 attained 1 time(s)
Number of parameters allowed to vary : 1
--> Adding variable: 9
--> Stepping onto the border of the L1 ball.
******************************
Iteration number: 2
Value of primal object function : 1088.391622
Value of dual object function : 509.867040
L1 norm of current beta : 10.000000 <= 10.000000
Maximal absolute value in t(X)%*%r : 1.155562e+02 attained 1 time(s)
Number of parameters allowed to vary : 2
--> Adding variable: 6
******************************
Iteration number: 3
Value of primal object function : 1015.808773
Value of dual object function : 925.285484
L1 norm of current beta : 10.000000 <= 10.000000
Maximal absolute value in t(X)%*%r : 7.728661e+01 attained 1 time(s)
Number of parameters allowed to vary : 3
--> Adding variable: 7
******************************
Iteration number: 4
Value of primal object function : 1013.514687
Value of dual object function : 997.935649
L1 norm of current beta : 10.000000 <= 10.000000
Maximal absolute value in t(X)%*%r : 7.058775e+01 attained 1 time(s)
Number of parameters allowed to vary : 4
--> Adding variable: 2
******************************
Iteration number: 5
Value of primal object function : 1013.485899
Value of dual object function : 1013.485899
L1 norm of current beta : 10.000000 <= 10.000000
Maximal absolute value in t(X)%*%r : 6.963196e+01 attained 5 time(s)
Number of parameters allowed to vary : 5
Call:
l1ce(formula = Yield ~ ., data = Iowa, trace = TRUE, bound = 10,
absolute.t = TRUE)
Coefficients:
(Intercept) Year Rain0 Temp1 Rain1
-1.223687e+03 6.665384e-01 8.895341e-03 0.000000e+00 0.000000e+00
Temp2 Rain2 Temp3 Rain3 Temp4
0.000000e+00 1.700162e+00 -1.612996e-01 0.000000e+00 -2.385012e-01
The absolute L1 bound was : 10
The Lagrangian for the bound is: 69.63196
[1] 30
Call:
l1ce(formula = lpsa ~ ., data = Prostate, bound = (1:30)/30)
Coefficients:
(Intercept) lcavol lweight age lbph svi
[1,] 2.4079829 0.05215074 0.00000000 0.000000000 0.000000000 0.00000000
[2,] 2.3375789 0.10430149 0.00000000 0.000000000 0.000000000 0.00000000
[3,] 2.2671749 0.15645223 0.00000000 0.000000000 0.000000000 0.00000000
[4,] 2.1967709 0.20860298 0.00000000 0.000000000 0.000000000 0.00000000
[5,] 2.1263669 0.26075372 0.00000000 0.000000000 0.000000000 0.00000000
[6,] 2.0559629 0.31290446 0.00000000 0.000000000 0.000000000 0.00000000
[7,] 1.9884010 0.36118122 0.00000000 0.000000000 0.000000000 0.01102907
[8,] 1.9371274 0.38725659 0.00000000 0.000000000 0.000000000 0.08526449
[9,] 1.8858539 0.41333196 0.00000000 0.000000000 0.000000000 0.15949991
[10,] 1.7462431 0.43253601 0.02752108 0.000000000 0.000000000 0.22028323
[11,] 1.5266610 0.44551948 0.07995688 0.000000000 0.000000000 0.26888842
[12,] 1.3070788 0.45850294 0.13239268 0.000000000 0.000000000 0.31749362
[13,] 1.0874967 0.47148641 0.18482849 0.000000000 0.000000000 0.36609881
[14,] 0.8679146 0.48446987 0.23726429 0.000000000 0.000000000 0.41470401
[15,] 0.7284811 0.49365402 0.26818634 0.000000000 0.009282588 0.45505849
[16,] 0.6186210 0.50024970 0.29093423 0.000000000 0.021521822 0.48803364
[17,] 0.5087610 0.50684539 0.31368212 0.000000000 0.033761056 0.52100878
[18,] 0.3989010 0.51344107 0.33643002 0.000000000 0.046000291 0.55398393
[19,] 0.4140072 0.51893849 0.35553434 -0.001664361 0.056178983 0.57442325
[20,] 0.5142130 0.52368800 0.37215748 -0.004462120 0.064954486 0.58632593
[21,] 0.6144188 0.52843752 0.38878061 -0.007259879 0.073729989 0.59822860
[22,] 0.6576089 0.53207720 0.40600224 -0.010015486 0.082086242 0.61110512
[23,] 0.6687680 0.53772360 0.41600759 -0.011756627 0.086998743 0.62829926
[24,] 0.6688581 0.54476635 0.42150088 -0.012882424 0.089863829 0.64799306
[25,] 0.6689483 0.55180911 0.42699418 -0.014008221 0.092728916 0.66768687
[26,] 0.6690384 0.55885186 0.43248747 -0.015134019 0.095594003 0.68738067
[27,] 0.6691286 0.56589462 0.43798076 -0.016259816 0.098459090 0.70707447
[28,] 0.6692187 0.57293737 0.44347406 -0.017385613 0.101324177 0.72676828
[29,] 0.6693089 0.57998013 0.44896735 -0.018511410 0.104189264 0.74646208
[30,] 0.6693990 0.58702288 0.45446064 -0.019637208 0.107054351 0.76615588
lcp gleason pgg45
[1,] 0.000000000 0.000000000 0.0000000000
[2,] 0.000000000 0.000000000 0.0000000000
[3,] 0.000000000 0.000000000 0.0000000000
[4,] 0.000000000 0.000000000 0.0000000000
[5,] 0.000000000 0.000000000 0.0000000000
[6,] 0.000000000 0.000000000 0.0000000000
[7,] 0.000000000 0.000000000 0.0000000000
[8,] 0.000000000 0.000000000 0.0000000000
[9,] 0.000000000 0.000000000 0.0000000000
[10,] 0.000000000 0.000000000 0.0000000000
[11,] 0.000000000 0.000000000 0.0000000000
[12,] 0.000000000 0.000000000 0.0000000000
[13,] 0.000000000 0.000000000 0.0000000000
[14,] 0.000000000 0.000000000 0.0000000000
[15,] 0.000000000 0.000000000 0.0001812107
[16,] 0.000000000 0.000000000 0.0005707550
[17,] 0.000000000 0.000000000 0.0009602994
[18,] 0.000000000 0.000000000 0.0013498437
[19,] 0.000000000 0.000000000 0.0017000959
[20,] 0.000000000 0.000000000 0.0020235909
[21,] 0.000000000 0.000000000 0.0023470858
[22,] 0.000000000 0.008508797 0.0025069611
[23,] -0.008582305 0.015186708 0.0027130057
[24,] -0.022423914 0.019465173 0.0029719083
[25,] -0.036265523 0.023743639 0.0032308108
[26,] -0.050107133 0.028022104 0.0034897134
[27,] -0.063948742 0.032300569 0.0037486160
[28,] -0.077790351 0.036579034 0.0040075185
[29,] -0.091631960 0.040857499 0.0042664211
[30,] -0.105473570 0.045135964 0.0045253236
Relative and absolute L1 bounds and the Lagrangians:
rel.bound abs.bound Lagrangian
[1,] 0.03333333 0.06146616 7.548890e+01
[2,] 0.06666667 0.12293233 6.958815e+01
[3,] 0.10000000 0.18439849 6.368740e+01
[4,] 0.13333333 0.24586466 5.778665e+01
[5,] 0.16666667 0.30733082 5.188590e+01
[6,] 0.20000000 0.36879698 4.598514e+01
[7,] 0.23333333 0.43026315 4.028653e+01
[8,] 0.26666667 0.49172931 3.574636e+01
[9,] 0.30000000 0.55319548 3.120619e+01
[10,] 0.33333333 0.61466164 2.747686e+01
[11,] 0.36666667 0.67612780 2.448159e+01
[12,] 0.40000000 0.73759397 2.148631e+01
[13,] 0.43333333 0.79906013 1.849104e+01
[14,] 0.46666667 0.86052630 1.549576e+01
[15,] 0.50000000 0.92199246 1.305806e+01
[16,] 0.53333333 0.98345862 1.089104e+01
[17,] 0.56666667 1.04492479 8.724026e+00
[18,] 0.60000000 1.10639095 6.557010e+00
[19,] 0.63333333 1.16785712 5.138082e+00
[20,] 0.66666667 1.22932328 4.228589e+00
[21,] 0.70000000 1.29078944 3.319095e+00
[22,] 0.73333333 1.35225561 2.440644e+00
[23,] 0.76666667 1.41372177 1.938086e+00
[24,] 0.80000000 1.47518794 1.661216e+00
[25,] 0.83333333 1.53665410 1.384347e+00
[26,] 0.86666667 1.59812026 1.107478e+00
[27,] 0.90000000 1.65958643 8.306082e-01
[28,] 0.93333333 1.72105259 5.537388e-01
[29,] 0.96666667 1.78251876 2.768694e-01
[30,] 1.00000000 1.84398492 7.549517e-15
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.