CRAN compabilities changes.

AnthonyChristidis · Sep 1, 2019 · efb2ff6 · efb2ff6
1 parent d953e7a
commit efb2ff6
Show file tree

Hide file tree

Showing 20 changed files with 70 additions and 26 deletions.
diff --git a/DESCRIPTION b/DESCRIPTION
@@ -2,19 +2,25 @@ Package: RPEGLMEN
 Type: Package
 Title: Efficient Implementation of glmnet using Rcpp and RcppEigen
 Version: 0.1.0
-Author: Anthony Christidis, Xin Chen, Daniel Hanson
+Date: 2019-08-30
+Author: Anthony Christidis <anthony.christidis@stat.ubc.ca>, 
+  Xin Chen <chenx26@uw.edu>, 
+  Daniel Hanson <hansondj@uw.edu>
 Maintainer: Anthony Christidis <anthony.christidis@stat.ubc.ca>
 Description: Implements the FISTA algorithm to fit a Gamma distribution 
              with an elastic net penalty as described in Chen, Arakvin and
              Martin (2018) <https://arxiv.org/abs/1804.07780>. An implementation 
              for the case of the exponential distribution is also available, with 
              details available in Chen and Martin (2018) 
              <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3085672>.
-License: GPL
+License: GPL (>= 2)
 Encoding: UTF-8
 LazyData: true
-Imports: Rcpp (>= 0.12.11)
+Imports: Rcpp (>= 1.0.2), RcppEigen (>= 0.3.3.5.0)
 LinkingTo: Rcpp, RcppEigen
 RoxygenNote: 6.1.1
 Suggests: 
     testthat
+NeedsCompilation: yes
+Biarch: true
+SystemRequirements: C++11
diff --git a/NAMESPACE b/NAMESPACE
@@ -1,5 +1,8 @@
-exportPattern("^[[:alpha:]]+")
-importFrom(Rcpp, evalCpp)
+# Generated by roxygen2: do not edit by hand
+
+export(fit.glmGammaNet)
+export(glmnet_exp)
 import(stats)
 import(utils)
-useDynLib(RPEGLMEN)
+importFrom(Rcpp,evalCpp)
+useDynLib(RPEGLMEN, .registration = TRUE)
diff --git a/R/FISTA.R b/R/FISTA.R
@@ -1,7 +1,4 @@
-## This is the R implementation of FISTA, based on
-## https://people.rennes.inria.fr/Cedric.Herzet/Cedric.Herzet/Sparse_Seminar/Entrees/2012/11/12_A_Fast_Iterative_Shrinkage-Thresholding_Algorithmfor_Linear_Inverse_Problems_(A._Beck,_M._Teboulle)_files/Breck_2009.pdf
-## Author: Xin Chen
-## 03/21/2018
+## This is the R implementation of FISTA
 FISTA = function(x0, f, g, f_grad, g_prox, ABSTOL = 1e-8, maxiter = 1000){
   lambda = 1
   beta = 0.5

diff --git a/R/RcppExports.R b/R/RcppExports.R
@@ -2,30 +2,30 @@
 # Generator token: 10BE3573-1514-4C36-9D1C-5A225CD40393
 
 scalarMultiplication <- function(c, M) {
-    .Call('_RPEGLMEN_scalarMultiplication', PACKAGE = 'RPEGLMEN', c, M)
+    .Call(`_RPEGLMEN_scalarMultiplication`, c, M)
 }
 
 addReals <- function(x, y) {
-    .Call('_RPEGLMEN_addReals', PACKAGE = 'RPEGLMEN', x, y)
+    .Call(`_RPEGLMEN_addReals`, x, y)
 }
 
 fitGlmFixed <- function(predictor_matrix, response_vector, alpha = 1, num_lambda = 100L, glm_type = 1L, max_iter = 100L, abs_tol = 1.0e-4, rel_tol = 1.0e-2, normalize_grad = FALSE, k_fold = 5L) {
-    .Call('_RPEGLMEN_fitGlmFixed', PACKAGE = 'RPEGLMEN', predictor_matrix, response_vector, alpha, num_lambda, glm_type, max_iter, abs_tol, rel_tol, normalize_grad, k_fold)
+    .Call(`_RPEGLMEN_fitGlmFixed`, predictor_matrix, response_vector, alpha, num_lambda, glm_type, max_iter, abs_tol, rel_tol, normalize_grad, k_fold)
 }
 
 fitGlmCv <- function(predictor_matrix, response_vector, alpha = 1, num_lambda = 100L, glm_type = 1L, max_iter = 100L, abs_tol = 1.0e-4, rel_tol = 1.0e-2, normalize_grad = FALSE, k_fold = 5L, has_intercept = TRUE, k_fold_iter = 5L) {
-    .Call('_RPEGLMEN_fitGlmCv', PACKAGE = 'RPEGLMEN', predictor_matrix, response_vector, alpha, num_lambda, glm_type, max_iter, abs_tol, rel_tol, normalize_grad, k_fold, has_intercept, k_fold_iter)
+    .Call(`_RPEGLMEN_fitGlmCv`, predictor_matrix, response_vector, alpha, num_lambda, glm_type, max_iter, abs_tol, rel_tol, normalize_grad, k_fold, has_intercept, k_fold_iter)
 }
 
 ExpNegativeLogLikelihood_cpp <- function(x, predictor_matrix, response_vector, alpha = 1, num_lambda = 100L, glm_type = 1L, max_iter = 100L, abs_tol = 1.0e-4, rel_tol = 1.0e-2, normalize_grad = FALSE, k_fold = 5L) {
-    .Call('_RPEGLMEN_ExpNegativeLogLikelihood_cpp', PACKAGE = 'RPEGLMEN', x, predictor_matrix, response_vector, alpha, num_lambda, glm_type, max_iter, abs_tol, rel_tol, normalize_grad, k_fold)
+    .Call(`_RPEGLMEN_ExpNegativeLogLikelihood_cpp`, x, predictor_matrix, response_vector, alpha, num_lambda, glm_type, max_iter, abs_tol, rel_tol, normalize_grad, k_fold)
 }
 
 GradExpNegativeLogLikelihood_cpp <- function(x, predictor_matrix, response_vector, alpha = 1, num_lambda = 100L, glm_type = 1L, max_iter = 100L, abs_tol = 1.0e-4, rel_tol = 1.0e-2, normalize_grad = FALSE, k_fold = 5L) {
-    .Call('_RPEGLMEN_GradExpNegativeLogLikelihood_cpp', PACKAGE = 'RPEGLMEN', x, predictor_matrix, response_vector, alpha, num_lambda, glm_type, max_iter, abs_tol, rel_tol, normalize_grad, k_fold)
+    .Call(`_RPEGLMEN_GradExpNegativeLogLikelihood_cpp`, x, predictor_matrix, response_vector, alpha, num_lambda, glm_type, max_iter, abs_tol, rel_tol, normalize_grad, k_fold)
 }
 
 ProxGradDescent_cpp <- function(predictor_matrix, response_vector, lambda = 0, alpha = 1, glm_type = 1L, max_iter = 100L, abs_tol = 1.0e-4, rel_tol = 1.0e-2, normalize_grad = FALSE, k_fold = 5L) {
-    .Call('_RPEGLMEN_ProxGradDescent_cpp', PACKAGE = 'RPEGLMEN', predictor_matrix, response_vector, lambda, alpha, glm_type, max_iter, abs_tol, rel_tol, normalize_grad, k_fold)
+    .Call(`_RPEGLMEN_ProxGradDescent_cpp`, predictor_matrix, response_vector, lambda, alpha, glm_type, max_iter, abs_tol, rel_tol, normalize_grad, k_fold)
 }
 
diff --git a/R/glmGammaNet.R b/R/glmGammaNet.R
@@ -1,4 +1,13 @@
-#' Fit glmnet model for Gamma distributed response data.
+#' @useDynLib RPEGLMEN, .registration = TRUE
+#' @importFrom Rcpp evalCpp
+NULL
+#'
+#' @import stats
+#' @import utils
+
+#' @title Elastic Net Penalized Gamma or Exponentially Distributed Response Variables
+#'
+#' @description \code{git.glmGammaNet} Fit glmnet model for Gamma distributed response data.
 #'
 #' @param A The matrix of independent variables.
 #' @param b The vector of response variables.
@@ -17,7 +26,9 @@
 #' @param ... Additional parameters.
 #'
 #' @return vector of optimal coefficient for the glm model
+#' 
 #' @export
+#' 
 fit.glmGammaNet <- function(A,
                             b,
                             exponential.dist = FALSE,

diff --git a/R/glmnet_exp.R b/R/glmnet_exp.R
@@ -15,10 +15,11 @@
 #' @param ... Additional Parameters.
 #'
 #' @return vector of optimal coefficient for the glm model
+#' 
 #' @export
+#' 
 glmnet_exp = function(A,
                       b,
-                      ...,
                       alpha.EN = 0.5,
                       num_lambda = 100L,
                       glm_type = 1L,
@@ -28,7 +29,8 @@ glmnet_exp = function(A,
                       normalize_grad = FALSE,
                       k_fold = 5L,
                       has_intercept = TRUE,
-                      k_fold_iter = 5L){
+                      k_fold_iter = 5L,
+                      ...){
   return(fitGlmCv(A,
          b,
          alpha = alpha.EN,

diff --git a/cleanup b/cleanup
diff --git a/configure b/configure
diff --git a/man/fit.glmGammaNet.Rd b/man/fit.glmGammaNet.Rd
diff --git a/src/GlmNetCpp.cpp b/src/GlmNetCpp.cpp
@@ -4,7 +4,7 @@
  * and open the template in the editor.
  */
 
-#include "GlmNetCpp.h"
+#include "GlmNetCpp.h"   
 
 GlmNetCpp::GlmNetCpp(const Eigen::MatrixXd& predictor_matrix,
         const Eigen::VectorXd& response_vector, double alpha,

diff --git a/src/GlmNetCpp.h b/src/GlmNetCpp.h
@@ -6,7 +6,7 @@
 
 /*
  * File:   glmnetcpp.h
- * Author: Xin Chen
+ * Author: Xin Chen 
  *
  * Created on June 2, 2017, 6:39 PM
  */

diff --git a/src/GlmNetCpp.o b/src/GlmNetCpp.o
diff --git a/src/GlmNetCvCpp.o b/src/GlmNetCvCpp.o
diff --git a/src/Makevars b/src/Makevars
diff --git a/src/Makevars.in b/src/Makevars.in
@@ -0,0 +1,12 @@
+## With Rcpp 0.11.0 and later, we no longer need to set PKG_LIBS as there is
+## no user-facing library. The include path to headers is already set by R.
+#PKG_LIBS = 
+
+## With R 3.1.0 or later, you can uncomment the following line to tell R to 
+## enable compilation with C++11 (or even C++14) where available
+CXX_STD = CXX11
+
+PKG_LIBS = $(LAPACK_LIBS) $(BLAS_LIBS) $(FLIBS) -pthread
+PKG_CFLAGS= -D__STDC_LIMIT_MACROS
+PKG_CXXFLAGS= -D__STDC_LIMIT_MACROS
+PKG_CPPFLAGS= -pthread
diff --git a/src/Makevars.win b/src/Makevars.win
@@ -0,0 +1,14 @@
+## With Rcpp 0.11.0 and later, we no longer need to set PKG_LIBS as there is
+## no user-facing library. The include path to headers is already set by R.
+#PKG_LIBS = 
+
+## With R 3.1.0 or later, you can uncomment the following line to tell R to 
+## enable compilation with C++11 (or even C++14) where available
+CXX_STD = CXX11
+
+## Use the R_HOME indirection to support installations of multiple R version
+PKG_LIBS = $(shell "${R_HOME}/bin${R_ARCH_BIN}/Rscript.exe" -e "Rcpp:::LdFlags()") $(LAPACK_LIBS) $(BLAS_LIBS) $(FLIBS) $(SHLIB_PTHREAD_FLAGS)
+
+PKG_CFLAGS= -D__STDC_LIMIT_MACROS
+PKG_CXXFLAGS= -D__STDC_LIMIT_MACROS
+PKG_CPPFLAGS= $(SHLIB_PTHREAD_FLAGS)
diff --git a/src/MyClass.o b/src/MyClass.o
diff --git a/src/RcppExports.o b/src/RcppExports.o
diff --git a/src/glmnetRcppInterface.cpp b/src/glmnetRcppInterface.cpp
@@ -65,7 +65,7 @@ Eigen::MatrixXd fitGlmFixed(const Eigen::MatrixXd& predictor_matrix,
   return X;
 }
 
-// [[Rcpp::export]]
+// [[Rcpp::export]]  
 Eigen::MatrixXd fitGlmCv(const Eigen::MatrixXd& predictor_matrix,
                             const Eigen::VectorXd& response_vector,
                             double alpha = 1, int num_lambda = 100, int glm_type = 1,

diff --git a/src/glmnetRcppInterface.o b/src/glmnetRcppInterface.o