- fix #130

DESCRIPTION

@@ -1,8 +1,8 @@
 Package: FeatureHashing
 Type: Package
 Title: Creates a Model Matrix via Feature Hashing with a Formula Interface
-Version: 0.9.1.3
-Date: 2018-04-10
+Version: 0.9.1.4
+Date: 2019-11-24
 Authors@R:   c(
   person("Wush", "Wu", email = "wush978@gmail.com", role = c("aut", "cre")), 
   person("Michael", "Benesty", email = "michael@benesty.fr", role = c("aut", "ctb")))
@@ -23,8 +23,8 @@ Imports:
   Matrix,
   digest(>= 0.6.8),
   magrittr (>= 1.5)
-LinkingTo: Rcpp, digest(>= 0.6.8), BH
-Suggests: RUnit, glmnet, knitr, xgboost, rmarkdown
+LinkingTo: Rcpp, digest(>= 0.6.8), BH(>= 1.54.0-1)
+Suggests: RUnit, glmnet, knitr, xgboost, rmarkdown, pROC
 SystemRequirements: C++11
 BugReports: https://github.com/wush978/FeatureHashing/issues
 URL: https://github.com/wush978/FeatureHashing

src/hash_function.h

@@ -19,13 +19,15 @@
 #ifndef __HASH_FUNCTION_HPP__
 #define __HASH_FUNCTION_HPP__
 
-#include <boost/detail/endian.hpp>
+#include <boost/predef/other/endian.h>
+#include <boost/endian/conversion.hpp>
 #include "digestlocal.h"
-#include "bswap_32.h"
 
 class HashFunction {
 
 public:
+
+  virtual ~HashFunction() { }
 
   virtual uint32_t operator()(const char* buf, int size, bool is_interaction = false) = 0;
 
@@ -34,6 +36,8 @@ class HashFunction {
 class NullHashFunction : public HashFunction {
 
   public:
+
+  virtual ~NullHashFunction() { }
 
   virtual uint32_t operator()(const char* buf, int size, bool is_interaction = false) {
     return 1;
@@ -48,6 +52,8 @@ class MurmurHash3HashFunction : public HashFunction {
 public :
 
   MurmurHash3HashFunction(uint32_t _seed) : seed(_seed) { }
+
+  virtual ~MurmurHash3HashFunction() { }
 
   virtual uint32_t operator()(const char* buf, int size, bool is_interaction = false) {
     return ::PMurHash32(seed, buf, size);
@@ -66,22 +72,26 @@ class MurmurHash3LogHashFunction : public HashFunction {
   : HashFunction(), seed(_seed), e(_e)
   { }
 
+  virtual ~MurmurHash3LogHashFunction() { }
+
   virtual uint32_t operator()(const char* buf, int size, bool is_interaction = false) {
     uint32_t retval = PMurHash32(seed, buf, size);
     if (is_interaction) {
       const uint32_t* src = reinterpret_cast<const uint32_t*>(buf);
-      #ifdef BOOST_BIG_ENDIAN
-      if (inverse_mapping.find(bswap_32(src[0])) == inverse_mapping.end()) throw std::logic_error("interaction is hashed before main effect!");
-      if (inverse_mapping.find(bswap_32(src[1])) == inverse_mapping.end()) throw std::logic_error("interaction is hashed before main effect!");
-      std::string key(inverse_mapping[bswap_32(src[0])]);
+      #if BOOST_ENDIAN_BIG_BYTE && !BOOST_ENDIAN_LITTLE_BYTE
+      if (inverse_mapping.find(boost::endian::endian_reverse(src[0])) == inverse_mapping.end()) throw std::logic_error("interaction is hashed before main effect!");
+      if (inverse_mapping.find(boost::endian::endian_reverse(src[1])) == inverse_mapping.end()) throw std::logic_error("interaction is hashed before main effect!");
+      std::string key(inverse_mapping[boost::endian::endian_reverse(src[0])]);
       key.append(":");
-      key.append(inverse_mapping[bswap_32(src[1])]);
-      #else
+      key.append(inverse_mapping[boost::endian::endian_reverse(src[1])]);
+      #elif !BOOST_ENDIAN_BIG_BYTE && BOOST_ENDIAN_LITTLE_BYTE
       if (inverse_mapping.find(src[0]) == inverse_mapping.end()) throw std::logic_error("interaction is hashed before main effect!");
       if (inverse_mapping.find(src[1]) == inverse_mapping.end()) throw std::logic_error("interaction is hashed before main effect!");
       std::string key(inverse_mapping[src[0]]);
       key.append(":");
       key.append(inverse_mapping[src[1]]);
+      #else
+      #error Unknown endianness
       #endif
       e[key.c_str()] = Rcpp::wrap((int) retval);
       inverse_mapping[retval] = key;

src/hash_internal.cpp

@@ -18,7 +18,6 @@
 
 #include <cstring>
 #include <deque>
-#include <boost/detail/endian.hpp>
 #include <boost/algorithm/string.hpp>
 #include <Rcpp.h>
 #include "hash_function.h"
@@ -52,12 +51,14 @@ IntegerVector h(CharacterVector src) {
 
 MH_UINT32 interaction(MH_UINT32 a, MH_UINT32 b) {
   MH_UINT32 buf[2];
-#ifdef BOOST_BIG_ENDIAN
-  buf[0] = bswap_32(a);
-  buf[1] = bswap_32(b);
-#else
+#if BOOST_ENDIAN_BIG_BYTE && !BOOST_ENDIAN_LITTLE_BYTE
+  buf[0] = boost::endian::endian_reverse(a);
+  buf[1] = boost::endian::endian_reverse(b);
+#elif !BOOST_ENDIAN_BIG_BYTE && BOOST_ENDIAN_LITTLE_BYTE
   buf[0] = a;
   buf[1] = b;
+#else
+  #error Unknown endianness
 #endif
   return PMurHash32(MURMURHASH3_H_SEED, reinterpret_cast<char*>(buf), sizeof(MH_UINT32) * 2);
 }

src/hashed_model_matrix.cpp

@@ -173,6 +173,9 @@ const ConvertersVec get_converters(
       } else {
         pVectorConverter q(*retval.rbegin());
         Param param("", _h_main, _h_binary, hash_size);
+        #ifdef NOISY_DEBUG
+        Rprintf("Initialize InteractionConverter\n");
+        #endif
         *retval.rbegin() = pInteractionConverter(new InteractionConverter(q, p, param));
       }
     }
@@ -222,7 +225,7 @@ SEXP hashed_model_matrix(RObject tf, DataFrameLike data, unsigned long hash_size
         });
         Rprintf("\n");
         #endif
-        std::for_each(i_origin.begin(), i_origin.end(), [&ivec, &xvec, &hash_size](uint32_t hashed_value) {
+        std::for_each(i_origin.begin(), i_origin.end(), [&ivec](uint32_t hashed_value) {
           ivec.push_back(hashed_value);
         });
         xvec.insert(xvec.end(), x_origin.begin(), x_origin.end());
@@ -248,7 +251,7 @@ SEXP hashed_model_matrix(RObject tf, DataFrameLike data, unsigned long hash_size
         const std::vector<uint32_t>& i_origin(p->get_feature(i));
         const std::vector<double>& x_origin(p->get_value(i));
         auto x_value = x_origin.begin();
-        std::for_each(i_origin.begin(), i_origin.end(), [&cache, &hash_size, &x_value, &i](uint32_t hashed_value) {
+        std::for_each(i_origin.begin(), i_origin.end(), [&cache, &x_value, &i](uint32_t hashed_value) {
           std::pair< std::vector<int>, std::vector<double> >& k(cache[hashed_value]);
           k.first.push_back(i);
           k.second.push_back(*(x_value++));

src/hashed_model_matrix.h

@@ -21,7 +21,6 @@
 
 #include <cstring>
 #include <memory>
-#include <boost/detail/endian.hpp>
 #include <Rcpp.h>
 #include "hash_function.h"
 #include "vector_converter.h"

src/intToRaw.cpp

@@ -1,16 +1,9 @@
 #include <Rcpp.h>
-#include <boost/detail/endian.hpp>
+#include <boost/predef/other/endian.h>
+#include <boost/endian/conversion.hpp>
 
-#ifdef linux
-#include <byteswap.h>
-#endif
-
-#ifndef bswap_32
-#ifdef __APPLE__
-#include <libkern/OSByteOrder.h>
-#else
-uint32_t bswap_32(uint32_t x);
-#endif
+#ifndef BOOST_ENDIAN_BIG_BYTE
+  #error No BOOST_ENDIAN_BIG_BYTE
 #endif
 
 using namespace Rcpp;
@@ -23,10 +16,12 @@ using namespace Rcpp;
 SEXP intToRaw(int src) {
   RawVector retval(4);
   uint32_t *p = (uint32_t*) &retval[0];
-  #ifdef BOOST_BIG_ENDIAN
-  p[0] = bswap_32(src);
-  #else
+  #if BOOST_ENDIAN_BIG_BYTE && !BOOST_ENDIAN_LITTLE_BYTE
+  p[0] = boost::endian::endian_reverse(src);
+  #elif !BOOST_ENDIAN_BIG_BYTE && BOOST_ENDIAN_LITTLE_BYTE
   p[0] = src;
+  #else
+  #error Unknown endianness
   #endif
   return retval;
 }

vignettes/FeatureHashing.Rmd

@@ -92,7 +92,9 @@ library(glmnet)
 cv.g.lr <- cv.glmnet(m.train, ipinyou.train$IsClick,
   family = "binomial")#, type.measure = "auc")
 p.lr <- predict(cv.g.lr, m.test, s="lambda.min")
-auc(ipinyou.test$IsClick, p.lr)
+
+library(pROC)
+auc(ipinyou.test$IsClick, c(p.lr))
 ```
 
 ## Gradient Boosted Decision Tree with [`xgboost`](https://cran.r-project.org/package=xgboost)
@@ -103,9 +105,9 @@ Following the script above,
 # GBDT with xgboost
 if(require("xgboost")){
   cv.g.gdbt <- xgboost(m.train, ipinyou.train$IsClick, max.depth=7, eta=0.1, subsample = 0.7, colsample_bytree = 0.7,
-    nround = 100, objective = "binary:logistic", verbose = ifelse(interactive(), 1, 0))
+    nround = 10, objective = "binary:logistic", verbose = ifelse(interactive(), 1, 0))
   p.lm <- predict(cv.g.gdbt, m.test)
-  glmnet::auc(ipinyou.test$IsClick, p.lm)  
+  auc(ipinyou.test$IsClick, p.lm)  
 }
 
 ```