enet
Estimation of an elastic net model, including options for Lasso and ridge regression.
Syntax
equation_name.enet(options) y x1 [x2 x3 ...] [@vw(...)]
List the dependent variable first, followed by a list of the independent variables. Use a “C” if you wish to include a constant or intercept term. Variable weights should be incorporated with the “@vw” tag, e.g., “@vw(x1, 0.5)”.
Options
Specification Options

 penalty=arg (default=“el”) Type of threshold estimation: “el” (elastic net), “ridge” (ridge), “lasso” (Lasso). alpha=arg (default=“.5”) Value of the mixing parameter. Must be a value between zero and one. lambda=arg Value of the penalty parameter. Can be a single number, list of space-delimited numbers, a workfile series object, or left blank for an EViews-supplied list (default). Values must be zero or greater.
General Options

 xtrans=arg (default=“none”) Transformation of the regressor variables: “none” (none), “L1” (L1), “L2” (L2), “stdsmpl” (sample standard deviation), “stdpop” (population standard deviation), “minmax” (min-max). lambdaratio=arg (default=0.0001) Ratio of minimum to maximum lambda for EViews-supplied list. nlambdas=arg (default=100) Number of lambas for EViews-supplied list. maxit=integer Maximum number of iterations. conv=scalar Set convergence criterion. The criterion is based upon the maximum of the percentage changes in the scaled estimates. The criterion will be set to the nearest value between 1e-24 and 0.2. w=arg Weight series or expression. wtype=arg (default=“istdev”) Weight specification type: inverse standard deviation (“istdev”), inverse variance (“ivar”), standard deviation (“stdev”), variance (“var”). wscale=arg Weight scaling: EViews default (“eviews”), average (“avg”), none (“none”).The default setting depends upon the weight type: “eviews” if “wtype=istdev”, “avg” for all others. showopts / ‑showopts [Do / do not] display the starting coefficient values and estimation options in the rotation output. prompt Force the dialog to appear from within a program. p Print basic estimation results.
Cross Validation Options

 cvmethod=arg (default=“kfold_cv”) Cross-validation method: “kfold” (k-fold), “shuffle” (shuffle), “leavepout” (leave p out), “leave1out” (leave one out), “rolling” (rolling window), “expanding” (expanding window). cvmeasure=arg (default=“mse”) Error measurement from cross-validation: “mse” (mean-squared error), “mae” (mean absolute error), “r2” (r-squared). training=arg (default=0.8) Proportion of data or number of data points in training set for shuffle method. test=arg (default=“mse”) Proportion of data or number of data points in test set for shuffle method. nreps=arg (default=1) Number of shuffle method repetitions. nfolds=arg (default=5) Number of folds for k-fold method. leaveout=arg (default=2) Number of data points left out for leave p out method. horizon=arg (default=1) Gap between the training set and future test set for rolling and expanding window methods. initial=arg (default=0) Number of initial data points held out of rolling and expanding window methods. window=arg (default=12) Window size for rolling window method.
Random Number Options

 seed=positive_integer from 0 to 2,147,483,647 Seed the random number generator.If not specified, EViews will seed random number generator with a single integer draw from the default global random number generator. rnd=arg (default=“kn” or method previously set using rndseed). Type of random number generator: improved Knuth generator (“kn”), improved Mersenne Twister (“mt”), Knuth’s (1997) lagged Fibonacci generator used in EViews 4 (“kn4”) L’Ecuyer’s (1999) combined multiple recursive generator (“le”), Matsumoto and Nishimura’s (1998) Mersenne Twister used in EViews 4 (“mt4”).
Examples
test.enet(penalty=lasso, alpha=1, lambda=.04065, conv=1e-8, maxit=5000) ystdz v1 v2 v3 c
estimates a lasso model with a single lambda value, no regressor transformation, and an optimization convergence limit of 1e-8 with a maximum of 5000 iterations.
test.enet(alpha=0, xtrans=stdpop, cvmeasure=mae, nfolds=10) ystdz v1 v2 v3 c
estimates an elastic net model with alpha = 0 (a ridge regression model solved numerically), sample standardization of the regressors, mean absolute error for the cross-validation error selection, and k-fold cross validation with ten folds.
Cross-references
See “Elastic Net and Lasso” for a discussion of elastic net, ridge regression, and Lasso models.