statsample 0.17.0 → 0.18.0

data/lib/statsample/matrix.rb

@@ -1,7 +1,26 @@
+class ::Vector
+  def to_matrix
+    ::Matrix.columns([self.to_a])
+  end
+end
 class ::Matrix
   def to_matrix
     self
   end
+  def to_dataset
+    f = (self.respond_to? :fields_y) ? fields_y : column_size.times.map {|i| _("VAR_%d") % (i+1) }
+    ds=Statsample::Dataset.new(f)
+    f.each do |ff|
+      ds[ff].type=:scale
+      ds[ff].name=ff
+    end
+    row_size.times {|i|
+      ds.add_case_array(self.row(i).to_a)
+    }
+    ds.update_valid_data
+    ds.name=self.name if self.respond_to? :name
+    ds
+  end
   if defined? :eigenpairs
     alias_method :eigenpairs_ruby, :eigenpairs
   end
@@ -31,6 +50,13 @@ class ::Matrix
 end
 
 module GSL
+  class Vector
+    class Col
+      def to_matrix
+      ::Matrix.columns([self.size.times.map {|i| self[i]}])
+      end
+    end
+  end
   class Matrix
     def to_gsl
       self
@@ -45,13 +71,54 @@ module GSL
 end
 
 module Statsample
+  # Module to add names to X and Y fields
+  module NamedMatrix
+    include Summarizable  
+
+    def fields
+    raise "Should be square" if !square?
+    fields_x
+    end
+    def fields=(v)
+    raise "Matrix should be square" if !square?
+    @fields_x=v
+    @fields_y=v
+    end
+    def fields_x=(v)
+    raise "Size of fields != row_size" if v.size!=row_size
+    @fields_x=v
+    end
+    def fields_y=(v)
+    raise "Size of fields != column_size" if v.size!=column_size
+    @fields_y=v
+    end
+    def fields_x
+    @fields_x||=row_size.times.collect {|i| _("X%d") % i} 
+    end
+    def fields_y
+    @fields_y||=column_size.times.collect {|i| _("Y%d") % i} 
+    end
+
+    def name
+      @name||=get_new_name
+    end
+    def name=(v)
+      @name=v
+    end
+    def get_new_name
+      @@named_matrix||=0
+      @@named_matrix+=1
+      _("Matrix %d") % @@named_matrix
+    end
+    
+  end
   # Module to add method for variance/covariance and correlation matrices
   # == Usage
   #  matrix=Matrix[[1,2],[2,3]]
   #  matrix.extend CovariateMatrix
   # 
   module CovariateMatrix
-    include Summarizable
+    include NamedMatrix
     @@covariatematrix=0
 
     # Get type of covariate matrix. Could be :covariance or :correlation
@@ -90,45 +157,19 @@ module Statsample
         self
       end
     end
-    def fields
-      raise "Should be square" if !square?
-      fields_x
-    end
-    def fields=(v)
-      raise "Matrix should be square" if !square?
-      @fields_x=v
-      @fields_y=v
-    end
-    def fields_x=(v)
-      raise "Size of fields != row_size" if v.size!=row_size
-      @fields_x=v
-    end
-    def fields_y=(v)
-      raise "Size of fields != column_size" if v.size!=column_size
-      @fields_y=v
-    end
-    def fields_x
-      @fields_x||=row_size.times.collect {|i| _("X%d") % i} 
-    end
-    def fields_y
-      @fields_y||=column_size.times.collect {|i| _("Y%d") % i} 
-    end
     
-    def name=(v)
-      @name=v
-    end
-    def name
-      @name||=get_new_name
-    end
+    
     # Get variance for field k
     # 
     def variance(k)
       submatrix([k])[0,0]
     end
+    
     def get_new_name
       @@covariatematrix+=1
       _("Covariate matrix %d") % @@covariatematrix
     end
+    
     # Select a submatrix of factors. If you have a correlation matrix
     # with a, b and c, you could obtain a submatrix of correlations of
     # a and b, b and c or a and b
@@ -172,7 +213,9 @@ module Statsample
       @name||= (type==:correlation ? _("Correlation"):_("Covariance"))+_(" Matrix")
       generator.table(:name=>@name, :header=>[""]+fields_y) do |t|
         row_size.times {|i|
-          t.row([fields_x[i]]+@rows[i].collect {|i1| sprintf("%0.3f",i1).gsub("0.",".")})
+          t.row([fields_x[i]]+@rows[i].collect {|i1|
+              i1.nil? ? "--" : sprintf("%0.3f",i1).gsub("0.",".")
+          })
         }
       end
     end

data/lib/statsample/multiset.rb

@@ -20,6 +20,36 @@ module Statsample
         }
         ms
     end
+    # Generate a new dataset as a union of partial dataset
+    # If block given, this is applied to each dataset before union
+    def union(&block)
+      union_field={}
+      types={}
+      names={}
+      labels={}
+      each do |k,ds|
+        if block
+          ds=ds.dup
+          yield k,ds
+        end
+        @fields.each do |f|
+          union_field[f]||=Array.new
+          union_field[f].concat(ds[f].data)
+          types[f]||=ds[f].type
+          names[f]||=ds[f].name
+          labels[f]||=ds[f].labels
+        end
+      end
+      
+      @fields.each do |f|
+        union_field[f]=union_field[f].to_vector(types[f])
+        union_field[f].name=names[f]
+        union_field[f].labels=labels[f]
+      end
+      ds_union=union_field.to_dataset
+      ds_union.fields=@fields
+      ds_union
+    end
     def datasets_names
         @datasets.keys.sort
     end
@@ -55,6 +85,12 @@ module Statsample
     def[](i)
       @datasets[i]
     end
+    def each(&block)
+      @datasets.each {|k,ds|
+        next if ds.cases==0
+        block.call(k,ds)
+      }
+    end
   end
   class StratifiedSample
     class << self

data/lib/statsample/reliability/multiscaleanalysis.rb

@@ -32,23 +32,32 @@ module Statsample
       attr_accessor :summary_pca
       # Add Principal Axis to summary
       attr_accessor :summary_principal_axis
-      # Add Parallel Analysis to summary
-      attr_accessor :summary_parallel_analysis
       # Options for Factor::PCA object
       attr_accessor :pca_options
       # Options for Factor::PrincipalAxis 
       attr_accessor :principal_axis_options
+      
+      # Add Parallel Analysis to summary
+      attr_accessor :summary_parallel_analysis
       # Options for Parallel Analysis
       attr_accessor :parallel_analysis_options
+      
+      # Add MPA to summary
+      attr_accessor :summary_map
+      # Options for MAP
+      attr_accessor :map_options
+      
+      
       # Generates a new MultiScaleAnalysis
       # Opts could be any accessor of the class 
       # * :name, 
       # * :summary_correlation_matrix
       # * :summary_pca
       # * :summary_principal_axis
+      # * :summary_map
       # * :pca_options
       # * :factor_analysis_options
-      #
+      # * :map_options
       # If block given, all methods should be called
       # inside object environment.
       # 
@@ -60,9 +69,11 @@ module Statsample
                         :summary_pca=>false,
                         :summary_principal_axis=>false,
                         :summary_parallel_analysis=>false,
+                        :summary_map=>false,
                         :pca_options=>Hash.new,
                         :principal_axis_options=>Hash.new,
-                        :parallel_analysis_options=>Hash.new
+                        :parallel_analysis_options=>Hash.new,
+                        :map_options=>Hash.new
         }
         @opts=opts_default.merge(opts)
         @opts.each{|k,v|
@@ -99,6 +110,10 @@ module Statsample
         opts||=pca_options        
         Statsample::Factor::PCA.new(correlation_matrix, opts)
       end
+      def map(opts=nil)
+        opts||=map_options
+        Statsample::Factor::MAP.new(correlation_matrix, opts)
+      end
       # Retrieves a PrincipalAxis Analysis (Factor::PrincipalAxis)
       # using all scales, using <tt>opts</tt> a options.
       def principal_axis_analysis(opts=nil)
@@ -151,7 +166,11 @@ module Statsample
               s2.parse_element(parallel_analysis)
             end
           end 
-          
+          if summary_map
+            s.section(:name=>_("MAP for %s") % name) do |s2|
+              s2.parse_element(map)
+            end
+          end           
         end
       end
     end

data/lib/statsample/reliability/scaleanalysis.rb

@@ -14,6 +14,7 @@ module Statsample
       include Summarizable
       attr_reader :ds,:mean, :sd,:valid_n, :alpha , :alpha_standarized, :variances_mean, :covariances_mean, :cov_m
       attr_accessor :name
+      attr_accessor :summary_histogram
       def initialize(ds, opts=Hash.new)
         @dumped=ds.fields.find_all {|f|
           ds[f].variance==0
@@ -40,9 +41,13 @@ module Statsample
         @sd = @total.sd
         @variance=@total.variance
         @valid_n = @total.size
-        opts_default={:name=>_("Reliability Analisis")}
+        opts_default={
+          :name=>_("Reliability Analisis"),
+          :summary_histogram=>true
+        }
         @opts=opts_default.merge(opts)
-        @name=@opts[:name]
+        @opts.each{|k,v| self.send("#{k}=",v) if self.respond_to? k }
+        
         @cov_m=Statsample::Bivariate.covariance_matrix(@ds)
         # Mean for covariances and variances
         @variances=@k.times.map {|i| @cov_m[i,i]}.to_scale
@@ -154,14 +159,13 @@ module Statsample
                 t.row [_("Items"), @ods.fields.size]
                 t.row [_("Sum mean"),     "%0.4f" % @o_total.mean]
                 t.row [_("S.d. mean"),     "%0.4f" % @o_total.sd]
-
               end
-            
               s.table(:name=>_("Deleted items"), :header=>['item','mean']) do |t|
                 @dumped.each do |f|
                   t.row(["#{@ods[f].name}(#{f})", "%0.5f" % @ods[f].mean])
                 end
               end
+              s.parse_element(Statsample::Graph::Histogram.new(@o_total)) if @summary_histogram
             end
           end
           
@@ -225,7 +229,7 @@ module Statsample
               t.row row
             end # end each
           end # table
-          
+          s.parse_element(Statsample::Graph::Histogram.new(@total)) if @summary_histogram
         end # section
       end # def
     end # class

data/lib/statsample/reliability/skillscaleanalysis.rb

@@ -18,7 +18,7 @@ module Statsample
       attr_accessor :summary_show_problematic_items
       def initialize(ds,key,opts=Hash.new)
         opts_default={
-          :name=>_("Skill Scale Reliability Analysis"),
+          :name=>_("Skill Scale Reliability Analysis (%s)") % ds.name,
           :summary_minimal_item_correlation=>0.10,
           :summary_show_problematic_items=>true
         }
@@ -28,9 +28,17 @@ module Statsample
         @opts.each{|k,v| self.send("#{k}=",v) if self.respond_to? k }
         @cds=nil
       end
+      # Dataset only corrected vectors
       def corrected_dataset_minimal
         cds=corrected_dataset
-        @key.keys.inject({}) {|ac,v| ac[v]=cds[v];ac}.to_dataset
+        dsm=@key.keys.inject({}) {|ac,v| ac[v]=cds[v];ac}.to_dataset
+        @key.keys.each do |k|
+          dsm[k].name=_("%s(corrected)") % @ds[k].name
+          dsm[k].labels=@ds[k].labels
+        end
+        
+        dsm.name=_("Corrected dataset from %s") % @ds.name
+        dsm
       end
       def vector_sum
         corrected_dataset_minimal.vector_sum
@@ -74,11 +82,19 @@ module Statsample
             s.section(:name=>_("Problematic Items")) do |spi|
               count=0
               sa.item_total_correlation.each do |k,v|
-                if v<summary_minimal_item_correlation
+                if v < summary_minimal_item_correlation
                   count+=1
                   spi.section(:name=>_("Item: %s") % @ds[k].name) do |spii|
                     spii.text _("Correct answer: %s") % @key[k]
-                    spii.parse_element(@ds[k])
+                    spii.text _("p: %0.3f") % corrected_dataset[k].mean
+                    props=@ds[k].proportions.inject({}) {|ac,v| ac[v[0]] = v[1].to_f;ac}
+                    
+                    spi.table(:name=>"Proportions",:header=>[_("Value"), _("%")]) do |table|
+                      props.each do |k1,v|
+                        table.row [ @ds[k].labeling(k1), "%0.3f" % v]
+                      end
+                    end
+                    
                   end
                 end
               end

data/lib/statsample/vector.rb

@@ -91,12 +91,12 @@ module Statsample
     # Note: data, missing_values and labels are duplicated, so
     # changes on original vector doesn't propages to copies.
     def dup
-      Vector.new(@data.dup,@type, :missing_values => @missing_values.dup, :labels => @labels.dup, :name=>@name.dup)
+      Vector.new(@data.dup,@type, :missing_values => @missing_values.dup, :labels => @labels.dup, :name=>@name)
     end
     # Returns an empty duplicate of the vector. Maintains the type,
     # missing values and labels.
     def dup_empty
-      Vector.new([],@type, :missing_values => @missing_values.dup, :labels => @labels.dup, :name=>@name.dup)
+      Vector.new([],@type, :missing_values => @missing_values.dup, :labels => @labels.dup, :name=> @name)
     end
     # Raises an exception if type of vector is inferior to t type
     def check_type(t)
@@ -105,31 +105,47 @@ module Statsample
     private :check_type
 
     # Return a vector usign the standarized values for data
-    # with sd with denominator n-1. With variance=0, returns nil
+    # with sd with denominator n-1. With variance=0 or mean nil,
+    # returns a vector of equal size full of nils
     # 
     
     def vector_standarized(use_population=false)
       check_type :scale
+      return ([nil]*size).to_scale if mean.nil?
       m=mean
       sd=use_population ? sdp : sds
-      return nil if sd==0.0
-      @data_with_nils.collect{|x|
+      return ([nil]*size).to_scale if sd==0.0
+      vector=@data_with_nils.collect{|x|
         if !x.nil?
           (x.to_f - m).quo(sd)
         else
           nil
         end
       }.to_vector(:scale)
+      vector.name=_("%s(standarized)")  % @name
+      vector
+    end
+    # Return a centered vector
+    def vector_centered
+      check_type :scale
+      m=mean
+      vector=@data_with_nils.collect {|x|
+        x.nil? ? nil : x.to_f-m
+      }.to_scale
+      vector.name=_("%s(centered)") % @name
+      vector
     end
     
     alias_method :standarized, :vector_standarized
+    alias_method  :centered, :vector_centered
     # Return a vector with values replaced with the percentiles
     # of each values
     def vector_percentil
       check_type :ordinal
-      c=size
-      ranked.map {|i| (i.quo(c)*100).to_f }.to_vector(@type)
-      
+      c=@valid_data.size
+      vector=ranked.map {|i| i.nil? ? nil : (i.quo(c)*100).to_f }.to_vector(@type)
+      vector.name=_("%s(percentil)")  % @name
+      vector
     end
     def box_cox_transformation(lambda) # :nodoc:
     raise "Should be a scale" unless @type==:scale
@@ -181,17 +197,17 @@ module Statsample
     # If parameter if defined, this value and lower
     # will be 0 and higher, 1
     def dichotomize(low=nil)
-    fs=factors
-    low||=factors.min
-    @data_with_nils.collect{|x|
-      if x.nil?
-        nil
-      elsif x>low
-        1
-      else
-        0
-      end
-    }.to_scale
+      fs=factors
+      low||=factors.min
+      @data_with_nils.collect{|x|
+        if x.nil?
+          nil
+        elsif x>low
+          1
+        else
+          0
+        end
+      }.to_scale
     end
     # Iterate on each item.
     # Equivalent to
@@ -268,7 +284,7 @@ module Statsample
     # Retrieves label for value x. Retrieves x if
     # no label defined.
     def labeling(x)
-    @labels.has_key?(x) ? @labels[x].to_s : x.to_s
+      @labels.has_key?(x) ? @labels[x].to_s : x.to_s
     end
     # Returns a Vector with data with labels replaced by the label.
     def vector_labeled
@@ -304,22 +320,22 @@ module Statsample
     # if update_valid = false, you should use
     # set_valid_data after all changes
     def missing_values=(vals)
-    @missing_values = vals
-    set_valid_data
+      @missing_values = vals
+      set_valid_data
     end
     # Set data considered as "today" on data vectors
     def today_values=(vals)
-    @today_values = vals
-    set_valid_data
+      @today_values = vals
+      set_valid_data
     end
     # Set level of measurement. 
     def type=(t)
-    @type=t	
-    set_scale_data if(t==:scale)
-    set_date_data if (t==:date)
+      @type=t	
+      set_scale_data if(t==:scale)
+      set_date_data if (t==:date)
     end
     def to_a
-    @data.dup
+      @data.dup
     end
     alias_method :to_ary, :to_a 
     
@@ -532,12 +548,12 @@ module Statsample
     # Ugly name. Really, create a Vector for standard 'matrix' package.
     # <tt>dir</tt> could be :horizontal or :vertical
     def to_matrix(dir=:horizontal)
-    case dir
-    when :horizontal
-      Matrix[@data]
-    when :vertical
-      Matrix.columns([@data])
-    end
+      case dir
+      when :horizontal
+        Matrix[@data]
+      when :vertical
+        Matrix.columns([@data])
+      end
     end
     def inspect
       self.to_s
@@ -553,22 +569,22 @@ module Statsample
     end
     end
     if Statsample::STATSAMPLE__.respond_to?(:frequencies)
-    # Returns a hash with the distribution of frecuencies for
-    # the sample                
-    def frequencies
-      Statsample::STATSAMPLE__.frequencies(@valid_data)
-    end
+      # Returns a hash with the distribution of frecuencies for
+      # the sample                
+      def frequencies
+        Statsample::STATSAMPLE__.frequencies(@valid_data)
+      end
     else
-    def frequencies #:nodoc:
-      _frequencies
-    end
+      def frequencies #:nodoc:
+        _frequencies
+      end
     end
     def _frequencies #:nodoc:
-    @valid_data.inject(Hash.new) {|a,x|
-      a[x]||=0
-      a[x]=a[x]+1
-      a
-    }
+      @valid_data.inject(Hash.new) {|a,x|
+        a[x]||=0
+        a[x]=a[x]+1
+        a
+      }
     end
     
     # Returns the most frequent item.
@@ -812,8 +828,7 @@ module Statsample
       end
       def mean # :nodoc:
       check_type :scale
-          
-          @gsl.mean
+        @gsl.nil? ? nil : @gsl.mean
       end				
       def variance_sample(m=nil) # :nodoc:
           check_type :scale
@@ -822,6 +837,7 @@ module Statsample
       end
       def standard_deviation_sample(m=nil) # :nodoc:
           check_type :scale
+          return nil if @gsl.nil?
           m||=mean
           @gsl.sd(m)
       end