version 4.1-2

Author: trevorhastie

DESCRIPTION

@@ -1,8 +1,8 @@
 Package: glmnet
 Type: Package
 Title: Lasso and Elastic-Net Regularized Generalized Linear Models
-Version: 4.1-1
-Date: 2021-02-17
+Version: 4.1-2
+Date: 2021-06-21
 Authors@R: c(person("Jerome", "Friedman", role=c("aut")),
 	     person("Trevor", "Hastie", role=c("aut", "cre"), email = "[email protected]"),
 	     person("Rob", "Tibshirani", role=c("aut")),
@@ -22,7 +22,7 @@ URL: https://glmnet.stanford.edu,
         https://dx.doi.org/10.18637/jss.v039.i05
 RoxygenNote: 7.1.1
 NeedsCompilation: yes
-Packaged: 2021-02-18 01:22:34 UTC; hastie
+Packaged: 2021-06-22 19:59:09 UTC; hastie
 Author: Jerome Friedman [aut],
   Trevor Hastie [aut, cre],
   Rob Tibshirani [aut],
@@ -32,4 +32,4 @@ Author: Jerome Friedman [aut],
   Junyang Qian [ctb]
 Maintainer: Trevor Hastie <[email protected]>
 Repository: CRAN
-Date/Publication: 2021-02-21 17:40:03 UTC
+Date/Publication: 2021-06-24 06:30:11 UTC

MD5

@@ -1,9 +1,9 @@
-1c79ad913d0f27ee763728e18f7884bc *DESCRIPTION
+405719048147a83a40afd4658a2e83c6 *DESCRIPTION
 a4e9841f3d7c3df90e1e5e013d53e5ec *NAMESPACE
-a1debac8a87a4bf3cb4e29c5c169341e *NEWS.md
+8173da17d52644d17b983af8304e9936 *NEWS.md
 b7e403cbac1975dfabbd5b1e6ec68cfc *R/Cindex.R
 4bd3eb09b5b5e4d5dd4a6d885c734b0d *R/assess.coxnet.R
-2f177df723d0565cbae683046d5ad550 *R/assess.glmnet.R
+7172891b9097b44ef3d8951153eee111 *R/assess.glmnet.R
 6d3a93c6972ccbfee9aa0c264cb08ad3 *R/auc.R
 8cd96999dce298c1d81ce51ea2f6d2a3 *R/auc.mat.R
 8b2f65e1c21c123294a132887081c627 *R/bigGlm.R
@@ -12,6 +12,7 @@ d8808544bfbafb423ce60e359ade4b64 *R/blend.relaxed.R
 70162e7910431ac9e89fcb5526469b84 *R/buildPredmat.coxnetlist.R
 019453d01483ecda68866eaa73e1df3c *R/buildPredmat.default.R
 885a478709d197d973cdc621cd3e5225 *R/check.dots.R
+612fef55db8dfaf923f5b986d8dad11e *R/check.exclude.R
 05d1834f8f3dc30e0428eb6d1dcb1ba1 *R/coef.cv.glmnet.R
 da1e68b0c4ae43a6ffcaa2ea3d7c287a *R/coef.cv.relaxed.R
 f288cf974491d93b56e5c4d6d204d658 *R/coef.glmnet.R
@@ -35,6 +36,7 @@ ece778c00ca6b58e088f6a700d23e15b *R/cv.multnet.R
 c739878e7fc617dab75e8cd65cf9b356 *R/cvcompute.R
 2b98d1871792cda71bd9323e9ee2a0b0 *R/cvstats.R
 90e2fb9130f89ddaef65e43202dc4261 *R/cvtype.R
+9a25662a16138900203c201bb4bd057b *R/data.R
 959eaac0e190795f3ee1ced251d60a90 *R/deviance.glmnet.R
 fb1ae53f9285f39d93592b6a0de61f6f *R/elnet.R
 0c6c0b1dfa5c98b55f2e9e097df3747d *R/error.bars.R
@@ -45,8 +47,8 @@ c65547cba8e3d5ad16b28a84881e7fca *R/getOptcv.glmnet.R
 b0325f5e584fbe646aedb73d02f6b86d *R/getOptcv.relaxed.R
 dd14d7acb21ae8f1bb87a54ea763950b *R/getcoef.R
 c1a8bceef71c915555f8bffc7298137e *R/getcoef.multinomial.R
-cc5c36c94de80e5399ef2db9355d4b09 *R/glmnet-package.R
-86176ceaf91d77a9f843143c05d38bc0 *R/glmnet.R
+495b7a692179c35799664c5510e04f3f *R/glmnet-package.R
+0b3c808ebf16acd3be2f7d5a83f2a69c *R/glmnet.R
 e0da0e93d1a87cc5d5342050a15e6354 *R/glmnet.control.R
 741a587449cd2f3613c2e52f25bc0280 *R/glmnet.measures.R
 d3b6ab46fd14433716fdc40cfd72807c *R/glmnetFlex.R
@@ -73,11 +75,11 @@ ea50b48783cf8f7f19578990a2fcd656 *R/plot.mrelnet.R
 18015a23b5efc79de433ac2b1fa5746a *R/plot.relaxed.R
 636feb1ac600bb09dccdfabfecc52c1d *R/plotCoef.R
 fa0986bb58c46a8481b78b877a6e7636 *R/predict.coxnet.R
-c56faa301b0f15275d45b356b364e2c6 *R/predict.cv.glmnet.R
-e4f6bd75906fdb9c58057143edbacbe5 *R/predict.cv.relaxed.R
+56a20beb9e09cba4456a611d6e385ca4 *R/predict.cv.glmnet.R
+46ef747bea2ff68cef5ef2ffbb5bb466 *R/predict.cv.relaxed.R
 459ca7b37d845f842e8b5681a2e3a5aa *R/predict.elnet.R
 0f470237229a0f3b973d29ef3683d73b *R/predict.fishnet.R
-fd9ff0509dd08f46b0d98d933bb0f02e *R/predict.glmnet.R
+1702dafa0f89817b8c77d231ad9224c3 *R/predict.glmnet.R
 3272244ba2c54519d973d5a4cbc2df8a *R/predict.lognet.R
 1bc0c68ce624df57b07bdabfd1e84597 *R/predict.mrelnet.R
 02fc6050ced4badab91cc882e06f3bb5 *R/predict.multnet.R
@@ -96,35 +98,42 @@ e44a2d10d5135f6931a13542d8996d56 *R/survfit.coxnet.R
 e2f22c34e46e1ec2964d1968e7116342 *R/survfit.cv.glmnet.R
 0d040824da1ffcd1edb57b252f6b1f6e *R/zeromat.R
 8920f300fb3d337d7048ef8af005d41e *README.md
-9305e0262835e4bf9ec618fb06fe020b *build/vignette.rds
-058ca4ea844ab733e99395fa1488546b *data/BinomialExample.RData
+70bc1a33f67d10989622ea3f4e680ebb *build/vignette.rds
+0814c5a080c9ba7874362051b6cbd89f *data/BinomialExample.rda
 7a404bc6a255ac95a6cc06cf4fe6ddbf *data/CVXResults.RData
-64d84fd86c91dfde6cbda5f13da82327 *data/CoxExample.RData
-2740a6e77cb5875cca0de3f5a6b2c2ed *data/MultiGaussianExample.RData
-f00302d6b4d512cd1f24c30c21335211 *data/MultinomialExample.RData
-c4f36770e60c6421603e9933930fec06 *data/PoissonExample.RData
-beaffe6c0ae597f8803216e3bbd4a2d1 *data/QuickStartExample.RData
-7056d5f98a624c060d0522045f1b68f8 *data/SparseExample.RData
+9497636bb0edb7f76c792f33c47ba681 *data/CoxExample.rda
+a72ec47b364e0eb3a85fc2d402c7ac16 *data/MultiGaussianExample.rda
+892fa8ad5b2ff24f9c989ebcec41bfb8 *data/MultinomialExample.rda
+cac6b8e217715098a5ee94e4066c1a4b *data/PoissonExample.rda
+35754bd4878bc234696cc2d9b880d90e *data/QuickStartExample.rda
+54067bd791ff172cff97ff9d26b6de12 *data/SparseExample.rda
 edcd84fb25d51203916ae3a39b1fc655 *inst/CITATION
-c2a99f2de8eaafe08dcdfdb754e1772e *inst/doc/Coxnet.R
-fae7a9d02341e6576b9560e26701ad9b *inst/doc/Coxnet.Rmd
-0190e2afe30e6b71eb740dd027b0918f *inst/doc/Coxnet.pdf
-2d93b78f53f037a964659ffaabe3aaac *inst/doc/glmnet.R
-bdb8b818d178d74a2af22608038932a1 *inst/doc/glmnet.Rmd
-731c80df0edc379d7a22fd2c85b002f7 *inst/doc/glmnet.pdf
+a93a7d8728d7660982e08bba82f720dc *inst/doc/Coxnet.R
+0bb3a929eff4b5695d75d5c08ae8360c *inst/doc/Coxnet.Rmd
+d8a790a361d27e7edd1988cf9728630a *inst/doc/Coxnet.pdf
+9ae5eaa92f330dbf58b156e1b5871dda *inst/doc/glmnet.R
+79ccc7b5e10f841e74ed9612b0e10cfd *inst/doc/glmnet.Rmd
+d00ef4196b199e0b00d9627b81487273 *inst/doc/glmnet.pdf
 85579e41e7c36b006e98a47bb0ca7f6f *inst/doc/glmnetFamily.R
 cab72afa072399ca5ae1906987ef89f4 *inst/doc/glmnetFamily.Rmd
-3f91549589947d5815b735619ea0cba7 *inst/doc/glmnetFamily.pdf
-4a3dfcb5e091bb6803bf35f9c689c16f *inst/doc/relax.R
-690ad421e350cde18389898140f41325 *inst/doc/relax.Rmd
-105554c99d10a7e387cc9305d54512ac *inst/doc/relax.pdf
-6b106667c04d867e9c1e279eedeff266 *inst/mortran/README.Rmd
-eca43bff4d1443c399d8305003777a9b *inst/mortran/glmnet5dpclean.m
+a79621fc63d35dbb310cf2c88dd6f2ef *inst/doc/glmnetFamily.pdf
+f3b0466d566df6079e63fe164a809f82 *inst/doc/relax.R
+e1dc532562ac69525a0b9c3c095be562 *inst/doc/relax.Rmd
+dc7816a0686ee4acc325cbbc3e10c2d6 *inst/doc/relax.pdf
+64c471277309951b4697ffdd1ee15fb8 *inst/mortran/README.Rmd
+a475108243c0db9c33089a0275c7eaba *inst/mortran/glmnet5dpclean.m
 d8516f118db49ea4fbde5d5b2f6e67a6 *inst/mortran/wls.m
 52e95162e7dc0e62ec329602b05cd001 *inst/testscripts/save_results.R
+6e03129049a39674c7ea15de8339511d *man/BinomialExample.Rd
 5a4389a6279bf8f6db41b6c7c5862bf9 *man/Cindex.Rd
-dbc0b807ae35fcb67ab8f8f849daa1d2 *man/assess.glmnet.Rd
-d0a7f0da9827e41030fd7644b421bb58 *man/beta_CVX.Rd
+9f14f46f2d349831ddf8fd8958bf69ef *man/CoxExample.Rd
+e472e3fbbaa3c1286891a59a561123ad *man/MultiGaussianExample.Rd
+2023e16a435a534ce184ddbb01a04bb7 *man/MultinomialExample.Rd
+287fe75c09580f4475f97e6a4653a2ca *man/PoissonExample.Rd
+81a89d362e387093991343cbbe045fce *man/QuickStartExample.Rd
+d9cb26ef685c853112acfa17afee001f *man/SparseExample.Rd
+db803cacf2e2783af427a53ae29000de *man/assess.glmnet.Rd
+404ddf86d813ed00f2ddfaf0cc88bf5f *man/beta_CVX.Rd
 413f29637810f0c84718ca2f5bb00ab1 *man/bigGlm.Rd
 9b08d40ce7e666b059b2679f2c878d9f *man/cox.fit.Rd
 e6388e65131a54875a8f44c9e536e17e *man/cox.path.Rd
@@ -142,7 +151,7 @@ d465886739abcf9ba23dd6928504e073 *man/elnet.fit.Rd
 1e9b7d5537b29ef0a4c097e26e0fb2a7 *man/get_start.Rd
 ec561c83a75b43232a6a10335e911ac3 *man/glmnet-internal.Rd
 d6e356e0500e6f9f2678cb03f2ec828d *man/glmnet-package.Rd
-322ad6277d1da769781a9d18404ddf25 *man/glmnet.Rd
+3011fcd9b2eed1ab588f6f17c8b01125 *man/glmnet.Rd
 5cc46f5cf2cc5389a89709d6ae129d0c *man/glmnet.control.Rd
 6d468edc0d0317c2098d27c64e17955e *man/glmnet.fit.Rd
 fed187cbcdf90f65defb1e5de08b3539 *man/glmnet.measures.Rd
@@ -167,15 +176,15 @@ a62678b3f19900ec74d299d7eebafbd7 *man/stratifySurv.Rd
 f2447246ed1bc0682526b694b2ff1daa *man/survfit.cv.glmnet.Rd
 24b1289dbc6e8d8bcc0a90fe57286599 *man/use.cox.path.Rd
 ef3f52f020e0348152d26890bb7e717e *man/weighted_mean_sd.Rd
-496f9de703e1b3d9ce71d4039b0ecaf8 *src/glmnet5dpclean.f
+df1527bce45036b9b3b532e9c376f645 *src/glmnet5dpclean.f
 32b239d462d1f062fe7389da6d2b2a27 *src/glmnet_init.c
 3965fc0eb8205870ea724035d96d8945 *src/pb.c
 f2fefcc9ad61b762fa701181df5ab618 *src/wls.f
-fae7a9d02341e6576b9560e26701ad9b *vignettes/Coxnet.Rmd
+0bb3a929eff4b5695d75d5c08ae8360c *vignettes/Coxnet.Rmd
 789121c8d4e2c1681d05f19d7a69a054 *vignettes/assets/coxnet.RDS
 3969a7d30dda77ddac9a8dcea40f8f58 *vignettes/assets/glmnet_refs.bib
 6b485f932628ec1755201811c3e05f49 *vignettes/assets/vignette_binomial.png
 3f2e5f9cf200b832dd68492f66f63f9e *vignettes/assets/vignette_gaussian.png
-bdb8b818d178d74a2af22608038932a1 *vignettes/glmnet.Rmd
+79ccc7b5e10f841e74ed9612b0e10cfd *vignettes/glmnet.Rmd
 cab72afa072399ca5ae1906987ef89f4 *vignettes/glmnetFamily.Rmd
-690ad421e350cde18389898140f41325 *vignettes/relax.Rmd
+e1dc532562ac69525a0b9c3c095be562 *vignettes/relax.Rmd

NEWS.md

@@ -1,3 +1,14 @@
+# glmnet 4.1-2
+
+A new feature added, as well as some minor fixes to documentation.
+* The exclude argument has come to life. Users can now pass a function
+  that can take arguments x, y and weights, or a subset of these, for
+  filtering variables. Details in documentation and vignette.
+* Prediction with single `newx` observation failed before. This is
+  fixed.
+* Labeling of predictions from `cv.glmnet` improved.
+* Fixed a bug in mortran/fortran that caused program to loop ad infinitum
+
 # glmnet 4.1-1
 
 Fixed some bugs in the coxpath function to do with sparse X.

R/assess.glmnet.R

@@ -18,7 +18,7 @@
 #'
 #' @aliases assess.glmnet confusion.glmnet roc.glmnet
 #' @param object Fitted \code{"glmnet"} or \code{"cv.glmnet"}, \code{"relaxed"}
-#' or \code{"cv.relaxed"} object, or a matrix of predictions (for
+#' or \code{"cv.relaxed"} object, OR a matrix of predictions (for
 #' \code{roc.glmnet} or \code{assess.glmnet}). For \code{roc.glmnet} the model
 #' must be a 'binomial', and for \code{confusion.glmnet} must be either
 #' 'binomial' or 'multinomial'
@@ -42,6 +42,7 @@
 #' @examples
 #'
 #' data(QuickStartExample)
+#' x <- QuickStartExample$x; y <- QuickStartExample$y
 #' set.seed(11)
 #' train = sample(seq(length(y)),70,replace=FALSE)
 #' fit1 = glmnet(x[train,], y[train])
@@ -53,13 +54,15 @@
 #' plot(fit1c$lambda, log="x",fit1a$mae,xlab="Log Lambda",ylab="Mean Absolute Error")
 #' abline(v=fit1c$lambda.min, lty=2, col="red")
 #' data(BinomialExample)
+#' x <- BinomialExample$x; y <- BinomialExample$y
 #' fit2 = glmnet(x[train,], y[train], family = "binomial")
 #' assess.glmnet(fit2,newx = x[-train,], newy=y[-train], s=0.1)
 #' plot(roc.glmnet(fit2, newx = x[-train,], newy=y[-train])[[10]])
 #' fit2c = cv.glmnet(x, y, family = "binomial", keep=TRUE)
 #' idmin = match(fit2c$lambda.min, fit2c$lambda)
 #' plot(roc.glmnet(fit2c$fit.preval, newy = y)[[idmin]])
 #' data(MultinomialExample)
+#' x <- MultinomialExample$x; y <- MultinomialExample$y
 #' set.seed(103)
 #' train = sample(seq(length(y)),100,replace=FALSE)
 #' fit3 = glmnet(x[train,], y[train], family = "multinomial")

R/check.exclude.R

@@ -0,0 +1,6 @@
+check.exclude <- function(exclude,nvars){
+      exclude <- unique(exclude)
+      if(length(exclude)> (nvars-2))stop("cannot retain 1 or less variables")
+      if(length(exclude)==0)exclude <- NULL
+      exclude
+      }

R/data.R

@@ -0,0 +1,112 @@
+#' Synthetic dataset with binary response
+#'
+#' Randomly generated data for binomial regression example.
+#'
+#' @name BinomialExample
+#' @docType data
+#' @usage data(BinomialExample)
+#' @keywords data
+#' @format List containing the following elements:
+#' \describe{
+#'   \item{x}{100 by 30 matrix of numeric values.}
+#'   \item{y}{Numeric vector of length 100 containing 44 zeros and 56 ones.}
+#' }
+"BinomialExample"
+
+#' Synthetic dataset with right-censored survival response
+#'
+#' Randomly generated data for Cox regression example.
+#'
+#' @name CoxExample
+#' @docType data
+#' @usage data(CoxExample)
+#' @keywords data
+#' @format List containing the following elements:
+#' \describe{
+#'   \item{x}{1,000 by 30 matrix of numeric values.}
+#'   \item{y}{1,000 by 2 matrix with column names "time" and "status". The
+#'     first column consists of positive numbers representing time to event,
+#'     while the second column represents the status indicator
+#'     (0=right-censored, 1=observed).}
+#' }
+"CoxExample"
+
+#' Synthetic dataset with multiple Gaussian responses
+#'
+#' Randomly generated data for multi-response Gaussian regression example.
+#'
+#' @name MultiGaussianExample
+#' @docType data
+#' @usage data(MultiGaussianExample)
+#' @keywords data
+#' @format List containing the following elements:
+#' \describe{
+#'   \item{x}{100 by 20 matrix of numeric values.}
+#'   \item{y}{100 by 4 matrix of numeric values, each column representing
+#'     one response vector.}
+#' }
+"MultiGaussianExample"
+
+#' Synthetic dataset with multinomial response
+#'
+#' Randomly generated data for multinomial regression example.
+#'
+#' @name MultinomialExample
+#' @docType data
+#' @usage data(MultinomialExample)
+#' @keywords data
+#' @format List containing the following elements:
+#' \describe{
+#'   \item{x}{500 by 30 matrix of numeric values.}
+#'   \item{y}{Numeric vector of length 500 containing 142 ones, 174 twos
+#'     and 184 threes.}
+#' }
+"MultinomialExample"
+
+#' Synthetic dataset with count response
+#'
+#' Randomly generated data for Poisson regression example.
+#'
+#' @name PoissonExample
+#' @docType data
+#' @usage data(PoissonExample)
+#' @keywords data
+#' @format List containing the following elements:
+#' \describe{
+#'   \item{x}{500 by 20 matrix of numeric values.}
+#'   \item{y}{Numeric vector of length 500 consisting of non-negative
+#'     integers.}
+#' }
+"PoissonExample"
+
+#' Synthetic dataset with Gaussian response
+#'
+#' Randomly generated data for Gaussian regression example.
+#'
+#' @name QuickStartExample
+#' @docType data
+#' @usage data(QuickStartExample)
+#' @keywords data
+#' @format List containing the following elements:
+#' \describe{
+#'   \item{x}{100 by 20 matrix of numeric values.}
+#'   \item{y}{Numeric vector of length 100.}
+#' }
+"QuickStartExample"
+
+#' Synthetic dataset with sparse design matrix
+#'
+#' Randomly generated data for Gaussian regression example with the
+#' design matrix x being in sparse matrix format.
+#'
+#' @name SparseExample
+#' @docType data
+#' @usage data(SparseExample)
+#' @keywords data
+#' @format List containing the following elements:
+#' \describe{
+#'   \item{x}{100 by 20 matrix of numeric values. x is in sparse matrix
+#'     format, having class "dgCMatrix".}
+#'   \item{y}{Numeric vector of length 100.}
+#' }
+"SparseExample"

R/glmnet-package.R

@@ -23,6 +23,7 @@
 #' @examples
 #'
 #' data(QuickStartExample)
+#' x <- QuickStartExample$x; y <- QuickStartExample$y
 #' glmnet(x, y)
 #'
 NULL

R/glmnet.R

@@ -146,7 +146,13 @@
 #' very large \code{nvars}, if a partial path is desired.
 #' @param pmax Limit the maximum number of variables ever to be nonzero
 #' @param exclude Indices of variables to be excluded from the model. Default
-#' is none. Equivalent to an infinite penalty factor (next item).
+#' is none. Equivalent to an infinite penalty factor for the variables excluded (next item).
+#' Users can supply instead an \code{exclude} function that generates the list of indices.
+#' This function is most generally defined as \code{function(x, y, weights, ...)},
+#' and is called inside \code{glmnet} to generate the indices for excluded variables.
+#' The \code{...} argument is required, the others are optional.
+#' This is useful for filtering wide data, and works correctly with \code{cv.glmnet}.
+#' See the vignette 'Introduction' for examples.
 #' @param penalty.factor Separate penalty factors can be applied to each
 #' coefficient. This is a number that multiplies \code{lambda} to allow
 #' differential shrinkage. Can be 0 for some variables, which implies no
@@ -346,6 +352,11 @@ glmnet=function(x,y,family=c("gaussian","binomial","poisson","multinomial","cox"
     if(is.null(np)|(np[2]<=1))stop("x should be a matrix with 2 or more columns")
     nobs=as.integer(np[1])
     nvars=as.integer(np[2])
+    if(is.null(weights))weights=rep(1,nobs)
+    else if(length(weights)!=nobs)stop(paste("number of elements in weights (",length(weights),") not equal to the number of rows of x (",nobs,")",sep=""))
+    if(is.function(exclude))exclude <- check.exclude(exclude(x=x,y=y,weights=weights),nvars)
+    if (length(penalty.factor) != nvars)
+        stop("the length of penalty.factor does not match the number of variables")
 
 ### See whether its a call to glmnet or to glmnet.path, based on family arg
     if(!is.character(family)){
@@ -378,8 +389,6 @@ glmnet=function(x,y,family=c("gaussian","binomial","poisson","multinomial","cox"
       alpha=as.double(alpha)
       nlam=as.integer(nlambda)
       y=drop(y) # we dont like matrix responses unless we need them
-      if(is.null(weights))weights=rep(1,nobs)
-      else if(length(weights)!=nobs)stop(paste("number of elements in weights (",length(weights),") not equal to the number of rows of x (",nobs,")",sep=""))
       dimy=dim(y)
       nrowy=ifelse(is.null(dimy),length(y),dimy[1])
       if(nrowy!=nobs)stop(paste("number of observations in y (",nrowy,") not equal to the number of rows of x (",nobs,")",sep=""))

R/predict.cv.glmnet.R

@@ -59,6 +59,7 @@ predict.cv.glmnet=function(object,newx,s=c("lambda.1se","lambda.min"),...){
     if(is.character(s)){
       s=match.arg(s)
       lambda=object[[s]]
+      names(lambda)=s
     }
     else stop("Invalid form for s")
   predict(object$glmnet.fit,newx,s=lambda,...)