opener-opinion-detector-basic 1.0.0

data/core/vendor/src/svm_light/svm_loqo.c

@@ -0,0 +1,211 @@
+/***********************************************************************/
+/*                                                                     */
+/*   svm_loqo.c                                                        */
+/*                                                                     */
+/*   Interface to the PR_LOQO optimization package for SVM.            */
+/*                                                                     */
+/*   Author: Thorsten Joachims                                         */
+/*   Date: 19.07.99                                                    */
+/*                                                                     */
+/*   Copyright (c) 1999  Universitaet Dortmund - All rights reserved   */
+/*                                                                     */
+/*   This software is available for non-commercial use only. It must   */
+/*   not be modified and distributed without prior permission of the   */
+/*   author. The author is not responsible for implications from the   */
+/*   use of this software.                                             */
+/*                                                                     */
+/***********************************************************************/
+
+# include <math.h>
+# include "pr_loqo/pr_loqo.h"
+# include "svm_common.h"
+
+/* Common Block Declarations */
+
+long verbosity;
+
+/* /////////////////////////////////////////////////////////////// */
+
+# define DEF_PRECISION_LINEAR    1E-8
+# define DEF_PRECISION_NONLINEAR 1E-14
+
+double *optimize_qp();
+double *primal=0,*dual=0;
+double init_margin=0.15;
+long   init_iter=500,precision_violations=0;
+double model_b;
+double opt_precision=DEF_PRECISION_LINEAR;
+
+/* /////////////////////////////////////////////////////////////// */
+
+void *my_malloc();
+
+double *optimize_qp(qp,epsilon_crit,nx,threshold,learn_parm)
+QP *qp;
+double *epsilon_crit;
+long nx; /* Maximum number of variables in QP */
+double *threshold;
+LEARN_PARM *learn_parm;
+/* start the optimizer and return the optimal values */
+{
+  register long i,j,result;
+  double margin,obj_before,obj_after;
+  double sigdig,dist,epsilon_loqo;
+  int iter;
+ 
+  if(!primal) { /* allocate memory at first call */
+    primal=(double *)my_malloc(sizeof(double)*nx*3);
+    dual=(double *)my_malloc(sizeof(double)*(nx*2+1));
+  }
+  
+  if(verbosity>=4) { /* really verbose */
+    printf("\n\n");
+    for(i=0;i<qp->opt_n;i++) {
+      printf("%f: ",qp->opt_g0[i]);
+      for(j=0;j<qp->opt_n;j++) {
+	printf("%f ",qp->opt_g[i*qp->opt_n+j]);
+      }
+      printf(": a%ld=%.10f < %f",i,qp->opt_xinit[i],qp->opt_up[i]);
+      printf(": y=%f\n",qp->opt_ce[i]);
+    }
+    for(j=0;j<qp->opt_m;j++) {
+      printf("EQ-%ld: %f*a0",j,qp->opt_ce[j]);
+      for(i=1;i<qp->opt_n;i++) {
+	printf(" + %f*a%ld",qp->opt_ce[i],i);
+      }
+      printf(" = %f\n\n",-qp->opt_ce0[0]);
+    }
+}
+
+  obj_before=0; /* calculate objective before optimization */
+  for(i=0;i<qp->opt_n;i++) {
+    obj_before+=(qp->opt_g0[i]*qp->opt_xinit[i]);
+    obj_before+=(0.5*qp->opt_xinit[i]*qp->opt_xinit[i]*qp->opt_g[i*qp->opt_n+i]);
+    for(j=0;j<i;j++) {
+      obj_before+=(qp->opt_xinit[j]*qp->opt_xinit[i]*qp->opt_g[j*qp->opt_n+i]);
+    }
+  }
+
+  result=STILL_RUNNING;
+  qp->opt_ce0[0]*=(-1.0);
+  /* Run pr_loqo. If a run fails, try again with parameters which lead */
+  /* to a slower, but more robust setting. */
+  for(margin=init_margin,iter=init_iter;
+      (margin<=0.9999999) && (result!=OPTIMAL_SOLUTION);) {
+    sigdig=-log10(opt_precision);
+
+    result=pr_loqo((int)qp->opt_n,(int)qp->opt_m,
+		   (double *)qp->opt_g0,(double *)qp->opt_g,
+		   (double *)qp->opt_ce,(double *)qp->opt_ce0,
+		   (double *)qp->opt_low,(double *)qp->opt_up,
+		   (double *)primal,(double *)dual, 
+		   (int)(verbosity-2),
+		   (double)sigdig,(int)iter, 
+		   (double)margin,(double)(qp->opt_up[0])/4.0,(int)0);
+
+    if(isnan(dual[0])) {     /* check for choldc problem */
+      if(verbosity>=2) {
+	printf("NOTICE: Restarting PR_LOQO with more conservative parameters.\n");
+      }
+      if(init_margin<0.80) { /* become more conservative in general */
+	init_margin=(4.0*margin+1.0)/5.0;
+      }
+      margin=(margin+1.0)/2.0;
+      (opt_precision)*=10.0;   /* reduce precision */
+      if(verbosity>=2) {
+	printf("NOTICE: Reducing precision of PR_LOQO.\n");
+      }
+    }
+    else if(result!=OPTIMAL_SOLUTION) {
+      iter+=2000; 
+      init_iter+=10;
+      (opt_precision)*=10.0;   /* reduce precision */
+      if(verbosity>=2) {
+	printf("NOTICE: Reducing precision of PR_LOQO due to (%ld).\n",result);
+      }      
+    }
+  }
+
+  if(qp->opt_m)         /* Thanks to Alex Smola for this hint */
+    model_b=dual[0];
+  else
+    model_b=0;
+
+  /* Check the precision of the alphas. If results of current optimization */
+  /* violate KT-Conditions, relax the epsilon on the bounds on alphas. */
+  epsilon_loqo=1E-10;
+  for(i=0;i<qp->opt_n;i++) {
+    dist=-model_b*qp->opt_ce[i]; 
+    dist+=(qp->opt_g0[i]+1.0);
+    for(j=0;j<i;j++) {
+      dist+=(primal[j]*qp->opt_g[j*qp->opt_n+i]);
+    }
+    for(j=i;j<qp->opt_n;j++) {
+      dist+=(primal[j]*qp->opt_g[i*qp->opt_n+j]);
+    }
+    /*  printf("LOQO: a[%d]=%f, dist=%f, b=%f\n",i,primal[i],dist,dual[0]); */
+    if((primal[i]<(qp->opt_up[i]-epsilon_loqo)) && (dist < (1.0-(*epsilon_crit)))) {
+      epsilon_loqo=(qp->opt_up[i]-primal[i])*2.0;
+    }
+    else if((primal[i]>(0+epsilon_loqo)) && (dist > (1.0+(*epsilon_crit)))) {
+      epsilon_loqo=primal[i]*2.0;
+    }
+  }
+
+  for(i=0;i<qp->opt_n;i++) {  /* clip alphas to bounds */
+    if(primal[i]<=(0+epsilon_loqo)) {
+      primal[i]=0;
+    }
+    else if(primal[i]>=(qp->opt_up[i]-epsilon_loqo)) {
+      primal[i]=qp->opt_up[i];
+    }
+  }
+
+  obj_after=0;  /* calculate objective after optimization */
+  for(i=0;i<qp->opt_n;i++) {
+    obj_after+=(qp->opt_g0[i]*primal[i]);
+    obj_after+=(0.5*primal[i]*primal[i]*qp->opt_g[i*qp->opt_n+i]);
+    for(j=0;j<i;j++) {
+      obj_after+=(primal[j]*primal[i]*qp->opt_g[j*qp->opt_n+i]);
+    }
+  }
+
+  /* if optimizer returned NAN values, reset and retry with smaller */
+  /* working set. */
+  if(isnan(obj_after) || isnan(model_b)) {
+    for(i=0;i<qp->opt_n;i++) {
+      primal[i]=qp->opt_xinit[i];
+    }     
+    model_b=0;
+    if(learn_parm->svm_maxqpsize>2) {
+      learn_parm->svm_maxqpsize--;  /* decrease size of qp-subproblems */
+    }
+  }
+
+  if(obj_after >= obj_before) { /* check whether there was progress */
+    (opt_precision)/=100.0;
+    precision_violations++;
+    if(verbosity>=2) {
+      printf("NOTICE: Increasing Precision of PR_LOQO.\n");
+    }
+  }
+
+  if(precision_violations > 500) { 
+    (*epsilon_crit)*=10.0;
+    precision_violations=0;
+    if(verbosity>=1) {
+      printf("\nWARNING: Relaxing epsilon on KT-Conditions.\n");
+    }
+  }	  
+
+  (*threshold)=model_b;
+
+  if(result!=OPTIMAL_SOLUTION) {
+    printf("\nERROR: PR_LOQO did not converge. \n");
+    return(qp->opt_xinit);
+  }
+  else {
+    return(primal);
+  }
+}
+

data/ext/hack/Rakefile

@@ -0,0 +1,17 @@
+require 'rake'
+require_relative 'support'
+
+desc 'Verifies the requirements'
+task :requirements do
+  verify_requirements
+  require_executable("make")
+end
+
+task :compile => :requirements do
+  compile_vendored_code
+end
+
+task :default => [:compile] do
+  # path = File.join(PYTHON_SITE_PACKAGES, 'pre_install')
+  # pip_install(PRE_INSTALL_REQUIREMENTS, path)
+end

data/ext/hack/support.rb

@@ -0,0 +1,88 @@
+require 'opener/build-tools'
+
+include Opener::BuildTools::Requirements
+include Opener::BuildTools::Python
+include Opener::BuildTools::Files
+
+# Directory of the core
+CORE_DIR = File.expand_path('../../../core', __FILE__)
+
+# Directory where packages will be installed to.
+PYTHON_SITE_PACKAGES = File.expand_path(
+  '../../../core/site-packages',
+  __FILE__
+)
+
+# Directory containing the temporary files.
+TMP_DIRECTORY = File.expand_path('../../../tmp', __FILE__)
+
+# Path to the pip requirements file used to install requirements before
+# packaging the Gem.
+PRE_BUILD_REQUIREMENTS = File.expand_path(
+  '../../../pre_build_requirements.txt',
+  __FILE__
+)
+
+# Path to the pip requirements file used to install requirements upon Gem
+# installation.
+PRE_INSTALL_REQUIREMENTS = File.expand_path(
+  '../../../pre_install_requirements.txt',
+  __FILE__
+)
+
+# Path to the vendor directory for C code.
+VENDOR_DIRECTORY = File.expand_path('../../../core/vendor', __FILE__)
+
+# Path to the directory to install vendored C code into.
+VENDOR_BUILD_DIRECTORY = File.expand_path(
+  '../../../core/vendor/build',
+  __FILE__
+)
+
+# Path to the directory that contains the source C code to compile.
+VENDOR_SRC_DIRECTORY = File.expand_path('../../../core/vendor/src', __FILE__)
+
+
+##
+# Verifies the requirements to install thi Gem.
+#
+def verify_requirements
+  require_executable('python')
+  require_version('python', python_version, '2.6.0')
+  require_executable('pip')
+  require_version('pip', pip_version, '1.3.1')
+end
+
+##
+# Compiles C code in the current directory using `make`.
+#
+# @param [Array] args The arguments to pass to ./configure
+#
+def compile(*args)
+  sh "./configure #{args.join(' ')}"
+  sh 'make'
+  sh 'make install'
+  sh 'make distclean'
+end
+
+##
+# Compiles the C code found in src/vendor.
+#
+def compile_vendored_code
+  src   = VENDOR_SRC_DIRECTORY
+  build = VENDOR_BUILD_DIRECTORY
+
+  Dir.chdir(File.join(src, "liblbfgs")) do
+    compile("--prefix=#{build}")
+  end
+
+  Dir.chdir(File.join(src, "crfsuite")) do
+    compile("--prefix=#{build}", "--with-liblbfgs=#{build}")
+  end
+
+  Dir.chdir(File.join(src, "svm_light")) do
+    sh 'make'
+    sh 'mv svm_classify svm_learn ../../build/bin'
+    sh 'make clean'
+  end
+end

data/lib/opener/opinion_detector_basic.rb

@@ -0,0 +1,91 @@
+require 'open3'
+
+require_relative 'opinion_detector_basic/version'
+
+module Opener
+
+
+  ##
+  # The basic Opinion detector.
+  #
+  # @!attribute [r] args
+  #  @return [Array]
+  # @!attribute [r] options
+  #  @return [Hash]
+  #
+  class OpinionDetectorBasic
+    attr_reader :args, :options
+
+    def initialize(options = {})
+      @args          = options.delete(:args) || []
+      @options       = options
+    end
+
+    ##
+    # Builds the command used to execute the kernel.
+    #
+    # @param [Array] args Commandline arguments passed to the command.
+    #
+    def command
+      return "#{adjust_python_path} python -E -OO #{kernel} #{args.join(' ')}"
+    end
+
+    ##
+    # Runs the command and returns the output of STDOUT, STDERR and the
+    # process information.
+    #
+    # @param [String] input The input to tag.
+    # @return [Array]
+    #
+    def run(input)
+      capture(input)
+    end
+
+    protected
+    ##
+    # @return [String]
+    #
+    def adjust_python_path
+      site_packages =  File.join(core_dir, 'site-packages')
+      "env PYTHONPATH=#{site_packages}:$PYTHONPATH"
+    end
+    ##
+    # capture3 method doesn't work properly with Jruby, so 
+    # this is a workaround
+    #
+    
+    def capture(input)
+      Open3.popen3(*command.split(" ")) {|i, o, e, t|
+        out_reader = Thread.new { o.read }
+        err_reader = Thread.new { e.read }
+        i.write input
+        i.close
+        [out_reader.value, err_reader.value, t.value]
+      }
+    end
+    
+    ##
+    # @return [String]
+    #
+    def core_dir
+      return File.expand_path('../../../core', __FILE__)
+    end
+
+    ##
+    # @return [String]
+    #
+    def kernel
+      return File.join(core_dir, 'opinion_detector_basic_multi.py')
+    end
+
+    ##
+    # @return the language from the KAF
+    #
+    def language(input)
+      document = Nokogiri::XML(input)
+      language = document.at('KAF').attr('xml:lang')
+      return language
+    end
+
+  end # OpinionDetectorBasic
+end # Opener

data/lib/opener/opinion_detector_basic/public/markdown.css

@@ -0,0 +1,284 @@
+
+input[type="text"], textarea
+{
+  width: 500px;
+}
+
+body {
+  font-family: Helvetica, arial, sans-serif;
+  font-size: 14px;
+  line-height: 1.6;
+  padding-top: 10px;
+  padding-bottom: 10px;
+  background-color: white;
+  padding: 30px; }
+
+body > *:first-child {
+  margin-top: 0 !important; }
+body > *:last-child {
+  margin-bottom: 0 !important; }
+
+a {
+  color: #4183C4; }
+a.absent {
+  color: #cc0000; }
+a.anchor {
+  display: block;
+  padding-left: 30px;
+  margin-left: -30px;
+  cursor: pointer;
+  position: absolute;
+  top: 0;
+  left: 0;
+  bottom: 0; }
+
+h1, h2, h3, h4, h5, h6 {
+  margin: 20px 0 10px;
+  padding: 0;
+  font-weight: bold;
+  -webkit-font-smoothing: antialiased;
+  cursor: text;
+  position: relative; }
+
+h1:hover a.anchor, h2:hover a.anchor, h3:hover a.anchor, h4:hover a.anchor, h5:hover a.anchor, h6:hover a.anchor {
+  background: url("../../images/modules/styleguide/para.png") no-repeat 10px center;
+  text-decoration: none; }
+
+h1 tt, h1 code {
+  font-size: inherit; }
+
+h2 tt, h2 code {
+  font-size: inherit; }
+
+h3 tt, h3 code {
+  font-size: inherit; }
+
+h4 tt, h4 code {
+  font-size: inherit; }
+
+h5 tt, h5 code {
+  font-size: inherit; }
+
+h6 tt, h6 code {
+  font-size: inherit; }
+
+h1 {
+  font-size: 28px;
+  color: black; }
+
+h2 {
+  font-size: 24px;
+  border-bottom: 1px solid #cccccc;
+  color: black; }
+
+h3 {
+  font-size: 18px; }
+
+h4 {
+  font-size: 16px; }
+
+h5 {
+  font-size: 14px; }
+
+h6 {
+  color: #777777;
+  font-size: 14px; }
+
+p, blockquote, ul, ol, dl, li, table, pre {
+  margin: 15px 0; }
+
+hr {
+  background: transparent url("../../images/modules/pulls/dirty-shade.png") repeat-x 0 0;
+  border: 0 none;
+  color: #cccccc;
+  height: 4px;
+  padding: 0; }
+
+body > h2:first-child {
+  margin-top: 0;
+  padding-top: 0; }
+body > h1:first-child {
+  margin-top: 0;
+  padding-top: 0; }
+body > h1:first-child + h2 {
+  margin-top: 0;
+  padding-top: 0; }
+body > h3:first-child, body > h4:first-child, body > h5:first-child, body > h6:first-child {
+  margin-top: 0;
+  padding-top: 0; }
+
+a:first-child h1, a:first-child h2, a:first-child h3, a:first-child h4, a:first-child h5, a:first-child h6 {
+  margin-top: 0;
+  padding-top: 0; }
+
+h1 p, h2 p, h3 p, h4 p, h5 p, h6 p {
+  margin-top: 0; }
+
+li p.first {
+  display: inline-block; }
+
+ul, ol {
+  padding-left: 30px; }
+
+ul :first-child, ol :first-child {
+  margin-top: 0; }
+
+ul :last-child, ol :last-child {
+  margin-bottom: 0; }
+
+dl {
+  padding: 0; }
+dl dt {
+  font-size: 14px;
+  font-weight: bold;
+  font-style: italic;
+  padding: 0;
+  margin: 15px 0 5px; }
+dl dt:first-child {
+  padding: 0; }
+dl dt > :first-child {
+  margin-top: 0; }
+dl dt > :last-child {
+  margin-bottom: 0; }
+dl dd {
+  margin: 0 0 15px;
+  padding: 0 15px; }
+dl dd > :first-child {
+  margin-top: 0; }
+dl dd > :last-child {
+  margin-bottom: 0; }
+
+blockquote {
+  border-left: 4px solid #dddddd;
+  padding: 0 15px;
+  color: #777777; }
+blockquote > :first-child {
+  margin-top: 0; }
+blockquote > :last-child {
+  margin-bottom: 0; }
+
+table {
+  padding: 0; }
+table tr {
+  border-top: 1px solid #cccccc;
+  background-color: white;
+  margin: 0;
+  padding: 0; }
+table tr:nth-child(2n) {
+  background-color: #f8f8f8; }
+table tr th {
+  font-weight: bold;
+  border: 1px solid #cccccc;
+  text-align: left;
+  margin: 0;
+  padding: 6px 13px; }
+table tr td {
+  border: 1px solid #cccccc;
+  text-align: left;
+  margin: 0;
+  padding: 6px 13px; }
+table tr th :first-child, table tr td :first-child {
+  margin-top: 0; }
+table tr th :last-child, table tr td :last-child {
+  margin-bottom: 0; }
+
+img {
+  max-width: 100%; }
+
+span.frame {
+  display: block;
+  overflow: hidden; }
+span.frame > span {
+  border: 1px solid #dddddd;
+  display: block;
+  float: left;
+  overflow: hidden;
+  margin: 13px 0 0;
+  padding: 7px;
+  width: auto; }
+span.frame span img {
+  display: block;
+  float: left; }
+span.frame span span {
+  clear: both;
+  color: #333333;
+  display: block;
+  padding: 5px 0 0; }
+span.align-center {
+  display: block;
+  overflow: hidden;
+  clear: both; }
+span.align-center > span {
+  display: block;
+  overflow: hidden;
+  margin: 13px auto 0;
+  text-align: center; }
+span.align-center span img {
+  margin: 0 auto;
+  text-align: center; }
+span.align-right {
+  display: block;
+  overflow: hidden;
+  clear: both; }
+span.align-right > span {
+  display: block;
+  overflow: hidden;
+  margin: 13px 0 0;
+  text-align: right; }
+span.align-right span img {
+  margin: 0;
+  text-align: right; }
+span.float-left {
+  display: block;
+  margin-right: 13px;
+  overflow: hidden;
+  float: left; }
+span.float-left span {
+  margin: 13px 0 0; }
+span.float-right {
+  display: block;
+  margin-left: 13px;
+  overflow: hidden;
+  float: right; }
+span.float-right > span {
+  display: block;
+  overflow: hidden;
+  margin: 13px auto 0;
+  text-align: right; }
+
+code, tt {
+  margin: 0 2px;
+  padding: 0 5px;
+  white-space: nowrap;
+  border: 1px solid #eaeaea;
+  background-color: #f8f8f8;
+  border-radius: 3px; }
+
+pre code {
+  margin: 0;
+  padding: 0;
+  white-space: pre;
+  border: none;
+  background: transparent; }
+
+.highlight pre {
+  background-color: #f8f8f8;
+  border: 1px solid #cccccc;
+  font-size: 13px;
+  line-height: 19px;
+  overflow: auto;
+  padding: 6px 10px;
+  border-radius: 3px; }
+
+pre {
+  background-color: #f8f8f8;
+  border: 1px solid #cccccc;
+  font-size: 13px;
+  line-height: 19px;
+  overflow: auto;
+  padding: 6px 10px;
+  border-radius: 3px; }
+pre code, pre tt {
+  background-color: transparent;
+  border: none; }
+