tomz-liblinear-ruby-swig 0.1.0 → 0.1.1

data/History.txt

@@ -1,3 +1,5 @@
 2009-05-10  Tom Zeng (tom.z.zeng@gmail.com)
   * initial check-in into github
 
+2009-07-10  Tom Zeng (tom.z.zeng@gmail.com)
+  * incorporated the _convert_to_feature_node_array fix from the LIBSVM development group

data/README.rdoc

@@ -9,7 +9,7 @@ This is the Ruby LIBLINEAR SWIG (Simplified Wrapper and Interface Generator)
 interface. LIBLINEAR is a high performance machine learning library for large 
 scale text mining(http://www.csie.ntu.edu.tw/~cjlin/liblinear).
 
-A slightly modified version of LIBLINEAR 1.33 is included which allow turnning
+A slightly modified version of LIBLINEAR 1.33 is included which allows turning
 on/off the default debuging/logging messages. You don't need your own copy of 
 SWIG to use this library - all needed files are generated using SWIG already.
 
@@ -25,8 +25,8 @@ Currently the gem is available on linux only(tested on Ubuntu 8 and Fedora 9/10,
 and should work on OS X), and you will need g++ installed to compile the native 
 code. 
 
-  sudo gem sources -a http://gems.github.com   (you only have to do this once)
-  sudo gem install tomz-liblinear-ruby-swig
+    sudo gem sources -a http://gems.github.com   (you only have to do this once)
+    sudo gem install tomz-liblinear-ruby-swig
 
 == SYNOPSIS:
 
@@ -60,6 +60,7 @@ Try the following multiclass problem in irb:
     irb(main):021:0>  pred = m.predict({1=>-0.1,2=>-0.2,3=>0.1,4=>1.1,5=>0.1})
     => 3
     irb(main):022:0> m.save("test.model")
+    irb(main):023:0> "your imagination " * 10
 
 
 For more examples see test*.rb in the liblinear-ruby-swig/liblinear-1.33/ruby
@@ -68,7 +69,7 @@ directory
 == AUTHOR:
 
 Tom Zeng
-http://www.tomzconsulting.com
-http://www.linkedin.com/in/tomzeng
-tom.z.zeng _at_ gmail _dot_ com
+- http://www.tomzconsulting.com
+- http://www.linkedin.com/in/tomzeng
+- tom.z.zeng _at_ gmail _dot_ com
 

data/Rakefile

@@ -1,11 +1,13 @@
 require 'rubygems'
+gem 'hoe', '>=1.8.3','<= 1.12.2'
 require 'hoe'
 
+
 task :default => ["sync_files","make_gem"] 
 
 EXT = "ext/liblinear?.#{Hoe::DLEXT}"
 
-Hoe.new('liblinear-ruby-swig', '0.1.0') do |p|
+Hoe.new('liblinear-ruby-swig', '0.1.1') do |p|
   p.author = 'Tom Zeng'
   p.email = 'tom.z.zeng@gmail.com'
   p.url = 'http://www.tomzconsulting.com'

data/ext/liblinear_wrap.cxx

@@ -2120,6 +2120,13 @@ void feature_node_matrix_set(struct feature_node **matrix, int i, struct feature
 	matrix[i] = array;
 }
 
+void set_bias_index(struct feature_node **matrix, int size, int maxlen, int *len_array)
+{
+        int i;
+        for(i=0; i<size; i++)
+                (matrix[i]+len_array[i])->index = maxlen+1;        
+}
+
 void feature_node_matrix_destroy(struct feature_node **matrix)
 {
 	free(matrix);
@@ -4095,6 +4102,51 @@ fail:
 }
 
 
+SWIGINTERN VALUE
+_wrap_set_bias_index(int argc, VALUE *argv, VALUE self) {
+  feature_node **arg1 = (feature_node **) 0 ;
+  int arg2 ;
+  int arg3 ;
+  int *arg4 = (int *) 0 ;
+  void *argp1 = 0 ;
+  int res1 = 0 ;
+  int val2 ;
+  int ecode2 = 0 ;
+  int val3 ;
+  int ecode3 = 0 ;
+  void *argp4 = 0 ;
+  int res4 = 0 ;
+  
+  if ((argc < 4) || (argc > 4)) {
+    rb_raise(rb_eArgError, "wrong # of arguments(%d for 4)",argc); SWIG_fail;
+  }
+  res1 = SWIG_ConvertPtr(argv[0], &argp1,SWIGTYPE_p_p_feature_node, 0 |  0 );
+  if (!SWIG_IsOK(res1)) {
+    SWIG_exception_fail(SWIG_ArgError(res1), Ruby_Format_TypeError( "", "feature_node **","set_bias_index", 1, argv[0] )); 
+  }
+  arg1 = reinterpret_cast< feature_node ** >(argp1);
+  ecode2 = SWIG_AsVal_int(argv[1], &val2);
+  if (!SWIG_IsOK(ecode2)) {
+    SWIG_exception_fail(SWIG_ArgError(ecode2), Ruby_Format_TypeError( "", "int","set_bias_index", 2, argv[1] ));
+  } 
+  arg2 = static_cast< int >(val2);
+  ecode3 = SWIG_AsVal_int(argv[2], &val3);
+  if (!SWIG_IsOK(ecode3)) {
+    SWIG_exception_fail(SWIG_ArgError(ecode3), Ruby_Format_TypeError( "", "int","set_bias_index", 3, argv[2] ));
+  } 
+  arg3 = static_cast< int >(val3);
+  res4 = SWIG_ConvertPtr(argv[3], &argp4,SWIGTYPE_p_int, 0 |  0 );
+  if (!SWIG_IsOK(res4)) {
+    SWIG_exception_fail(SWIG_ArgError(res4), Ruby_Format_TypeError( "", "int *","set_bias_index", 4, argv[3] )); 
+  }
+  arg4 = reinterpret_cast< int * >(argp4);
+  set_bias_index(arg1,arg2,arg3,arg4);
+  return Qnil;
+fail:
+  return Qnil;
+}
+
+
 SWIGINTERN VALUE
 _wrap_feature_node_matrix_destroy(int argc, VALUE *argv, VALUE self) {
   feature_node **arg1 = (feature_node **) 0 ;
@@ -4520,6 +4572,7 @@ SWIGEXPORT void Init_liblinear(void) {
   rb_define_module_function(mLiblinear, "feature_node_array_destroy", VALUEFUNC(_wrap_feature_node_array_destroy), -1);
   rb_define_module_function(mLiblinear, "feature_node_matrix", VALUEFUNC(_wrap_feature_node_matrix), -1);
   rb_define_module_function(mLiblinear, "feature_node_matrix_set", VALUEFUNC(_wrap_feature_node_matrix_set), -1);
+  rb_define_module_function(mLiblinear, "set_bias_index", VALUEFUNC(_wrap_set_bias_index), -1);
   rb_define_module_function(mLiblinear, "feature_node_matrix_destroy", VALUEFUNC(_wrap_feature_node_matrix_destroy), -1);
 }
 

data/lib/linear.rb

@@ -99,7 +99,7 @@ class LParameter
   end
 end
 
-def _convert_to_feature_node_array(x, bias=-1)
+def _convert_to_feature_node_array(x, maxlen, bias=-1)
   # convert a sequence or mapping to an feature_node array
   
   # Find non zero elements
@@ -121,7 +121,7 @@ def _convert_to_feature_node_array(x, bias=-1)
   if bias >=0
     data = feature_node_array(iter_range.size+2)
     #puts "bias element (#{iter_range.size},#{bias})"
-    feature_node_array_set(data,iter_range.size,iter_range.size,bias)
+    feature_node_array_set(data,iter_range.size,maxlen+1,bias)
     feature_node_array_set(data,iter_range.size+1,-1,0)
   else
     data = feature_node_array(iter_range.size+1)
@@ -154,17 +154,25 @@ class LProblem
     @x_matrix = x_matrix = feature_node_matrix(size)
     @data = []
     @maxlen = 0  #max number of features
+    len_array=[]
+
     for i in (0..size-1)
-      data = _convert_to_feature_node_array(x[i], bias)
+      data = _convert_to_feature_node_array(x[i], @maxlen, bias)
       @data << data
       feature_node_matrix_set(x_matrix,i,data)
+
       if x[i].class == Hash
         if x[i].size > 0
-          @maxlen = [@maxlen,x[i].keys.max].max
+          @maxlen = [@maxlen,x[i].keys.max].max          
         end
       else
         @maxlen = [@maxlen,x[i].size].max
       end
+      len_array << x[i].size      
+    end
+
+    if bias >= 0
+      set_bias_index(x_matrix, size, @maxlen, _int_array(len_array))
     end
     
     prob.y = y_array
@@ -217,7 +225,7 @@ class LModel
   end
   
   def predict(x)
-    data = _convert_to_feature_node_array(x, @model.bias)
+    data = _convert_to_feature_node_array(x, @model.nr_feature, @model.bias)
     ret = Liblinear::predict(@model,data)
     feature_node_array_destroy(data)
     return ret
@@ -235,7 +243,7 @@ class LModel
   def predict_values_raw(x)
     #convert x into feature_node, allocate a double array for return
     n = (@nr_class*(@nr_class-1)/2).floor
-    data = _convert_to_feature_node_array(x, @model.bias)
+    data = _convert_to_feature_node_array(x, @model.nr_feature, @model.bias)
     dblarr = new_double(n)
     Liblinear::predict_values(@model, data, dblarr)
     ret = _double_array_to_list(dblarr, n)
@@ -276,7 +284,7 @@ class LModel
 #    end
     
     #convert x into feature_node, alloc a double array to receive probabilities
-    data = _convert_to_feature_node_array(x,@model.bias)
+    data = _convert_to_feature_node_array(x, @model.nr_feature, @model.bias)
     dblarr = new_double(@nr_class)
     pred = Liblinear::predict_probability(@model, data, dblarr)
     pv = _double_array_to_list(dblarr, @nr_class)

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: tomz-liblinear-ruby-swig
 version: !ruby/object:Gem::Version 
-  version: 0.1.0
+  version: 0.1.1
 platform: ruby
 authors: 
 - Tom Zeng
@@ -9,7 +9,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2009-05-09 00:00:00 -07:00
+date: 2009-07-10 00:00:00 -07:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency