statsample 0.9.0 → 0.10.0

data/lib/statsample/regression/multiple/baseengine.rb

@@ -3,17 +3,22 @@ module Statsample
     module Multiple
       # Base class for Multiple Regression Engines
       class BaseEngine
-        
-        include GetText
-        bindtextdomain("statsample")
+        include Statsample::Summarizable
         # Name of analysis
         attr_accessor :name
-        
+        # Minimum number of  valid case for pairs of correlation
+        attr_reader :cases
+        # Number of valid cases (listwise)
+        attr_reader :valid_cases
+        # Number of total cases (dataset.cases)
+        attr_reader :total_cases
         def self.univariate?
           true
         end
         def initialize(ds, y_var, opts = Hash.new)
           @ds=ds
+          @predictors_n=@ds.fields.size-1
+          @total_cases=@ds.cases
           @cases=@ds.cases
           @y_var=y_var
           @r2=nil
@@ -23,12 +28,16 @@ module Statsample
           }
         end
         # Calculate F Test
-        def f_test
-          @f_test||=Statsample::Test::F.new(ssr, sse, df_r, df_e, :name_numerator=>_("Regression"), :name_denominator=>_("Error"), :name=>"ANOVA")
+        def anova
+          @anova||=Statsample::Anova::OneWay.new(:ss_num=>ssr, :ss_den=>sse, :df_num=>df_r, :df_den=>df_e, :name_numerator=>_("Regression"), :name_denominator=>_("Error"), :name=>"ANOVA")
+        end
+        # Standard error of estimate
+        def se_estimate
+          Math::sqrt(sse.quo(df_e))
         end
         # Retrieves a vector with predicted values for y
         def predicted
-          (0...@ds.cases).collect { |i|
+          @total_cases.times.collect { |i|
             invalid=false
             vect=@dep_columns.collect {|v| invalid=true if v[i].nil?; v[i]}
             if invalid
@@ -44,7 +53,7 @@ module Statsample
         end
         # Retrieves a vector with residuals values for y
         def residuals
-          (0...@ds.cases).collect{|i|
+          (0...@total_cases).collect{|i|
             invalid=false
             vect=@dep_columns.collect{|v| invalid=true if v[i].nil?; v[i]}
             if invalid or @ds[@y_var][i].nil?
@@ -62,6 +71,9 @@ module Statsample
         def sst
           raise "You should implement this"
         end
+        def r2_adjusted
+          r2-((1-r2)*@predictors_n).quo(df_e)
+        end
         # Sum of squares (regression)
         def ssr
           r2*sst
@@ -89,19 +101,19 @@ module Statsample
         end
         # Degrees of freedom for regression
         def df_r
-          @dep_columns.size
+          @predictors_n
         end
         # Degrees of freedom for error
         def df_e
-          @ds_valid.cases-@dep_columns.size-1
+          @valid_cases-@predictors_n-1
         end
         # Fisher for Anova
         def f
-          f_test.f
+          anova.f
         end
         # p-value of Fisher
         def probability
-          f_test.probability
+          anova.probability
         end
         # Tolerance for a given variable
         # http://talkstats.com/showthread.php?t=5056
@@ -130,6 +142,7 @@ module Statsample
           out
         end
         # Estandar error of R^2
+        # ????
         def se_r2
           Math::sqrt((4*r2*(1-r2)**2*(df_e)**2).quo((@cases**2-1)*(@cases+3)))
         end
@@ -139,10 +152,11 @@ module Statsample
         def estimated_variance_covariance_matrix
           mse_p=mse
           columns=[]
-          @ds_valid.each_vector{|k,v|
+          @ds_valid.fields.each{|k|
+            v=@ds_valid[k]
             columns.push(v.data) unless k==@y_var
           }
-          columns.unshift([1.0]*@ds_valid.cases)
+          columns.unshift([1.0]*@valid_cases)
           x=Matrix.columns(columns)
           matrix=((x.t*x)).inverse * mse
           matrix.collect {|i| Math::sqrt(i) if i>0 }
@@ -155,25 +169,22 @@ module Statsample
         def constant_se
           estimated_variance_covariance_matrix[0,0]
         end
-        def summary
-          rp=ReportBuilder.new()
-          rp.add(self)
-          rp.to_text
-        end
         def report_building(b)
           b.section(:name=>@name) do |g|
             c=coeffs
-            g.text(_("Engine: %s") % self.class)
-            g.text(_("Cases(listwise)=%d(%d)") % [@ds.cases, @ds_valid.cases])
-            g.text("R=#{sprintf('%0.3f',r)}")
-            g.text("R^2=#{sprintf('%0.3f',r2)}")
+            g.text _("Engine: %s") % self.class
+            g.text(_("Cases(listwise)=%d(%d)") % [@total_cases, @valid_cases])
+            g.text _("R=%0.3f") % r
+            g.text _("R^2=%0.3f") % r2
+            g.text _"R^2 Adj=%0.3f" % r2_adjusted
+            g.text _("Std.Error R=%0.3f") % se_estimate
             
             g.text(_("Equation")+"="+ sprintf('%0.3f',constant) +" + "+ @fields.collect {|k| sprintf('%0.3f%s',c[k],k)}.join(' + ') )
             
-            g.parse_element(f_test)
+            g.parse_element(anova)
             sc=standarized_coeffs
             cse=coeffs_se
-            g.table(:name=>"Beta coefficients", :header=>%w{coeff b beta se t}.collect{|field| _(field)} ) do |t|
+            g.table(:name=>_("Beta coefficients"), :header=>%w{coeff b beta se t}.collect{|field| _(field)} ) do |t|
               t.row([_("Constant"), sprintf("%0.3f", constant), "-", sprintf("%0.3f", constant_se), sprintf("%0.3f", constant_t)])
               @fields.each do |f|
                 t.row([f, sprintf("%0.3f", c[f]), sprintf("%0.3f", sc[f]), sprintf("%0.3f", cse[f]), sprintf("%0.3f", c[f].quo(cse[f]))])

data/lib/statsample/regression/multiple/gslengine.rb

@@ -21,6 +21,7 @@ class GslEngine < BaseEngine
     super
     @ds=ds.dup_only_valid
     @ds_valid=@ds
+    @valid_cases=@ds_valid.cases
     @dy=@ds[@y_var]
     @ds_indep=ds.dup(ds.fields-[y_var])
     # Create a custom matrix
@@ -111,6 +112,19 @@ class GslEngine < BaseEngine
         v.quo(red_sd)
     }.to_vector(:scale)
   end
+  
+    # Standard error for coeffs
+  def coeffs_se
+    out={}
+    evcm=estimated_variance_covariance_matrix
+    @ds_valid.fields.each_with_index do |f,i|
+      mi=i+1
+      next if f==@y_var
+      out[f]=evcm[mi,mi]
+    end
+    out
+  end
+  
 end
 end
 end

data/lib/statsample/regression/multiple/matrixengine.rb

@@ -50,7 +50,7 @@ class MatrixEngine < BaseEngine
     @fields=matrix.fields-[y_var]
     
     @n_predictors=@fields.size
-    
+    @predictors_n=@n_predictors
     @matrix_x= @matrix_cor.submatrix(@fields)
     @matrix_x_cov= @matrix_cov.submatrix(@fields)
     
@@ -88,11 +88,11 @@ class MatrixEngine < BaseEngine
       }
     else
       @coeffs_stan=result_matrix.column(0).to_a
-      
       @coeffs=standarized_coeffs.collect {|k,v|
         standarized_coeffs[k]*@y_sd.quo(@x_sd[k])
       } 
     end
+    @total_cases=@valid_cases=@cases
   end
   def cases
     raise "You should define the number of valid cases first" if @cases.nil?
@@ -144,6 +144,7 @@ class MatrixEngine < BaseEngine
   # Reference:
   # * http://talkstats.com/showthread.php?t=5056
   def tolerance(var)
+    return 1 if @matrix_x.column_size==1
     lr=Statsample::Regression::Multiple::MatrixEngine.new(@matrix_x, var)
     1-lr.r2
   end
@@ -165,7 +166,7 @@ class MatrixEngine < BaseEngine
   # t value for constant
   def constant_t
     return nil if constant_se.nil?
-    constant.to_f/constant_se
+    constant.to_f / constant_se
   end
   # Standard error for constant.
   # This method recreates the estimaded variance-covariance matrix
@@ -197,35 +198,6 @@ class MatrixEngine < BaseEngine
     matrix.collect {|i| Math::sqrt(i) if i>0 }[0,0]
   end
   
-  def report_building(builder) # :nodoc:
-    builder.section(:name=>_("Multiple Regression: ")+@name) do |g|
-      c=coeffs
-      g.text(_("Engine: %s") % self.class)
-      g.text(_("Cases=%d") % [@cases])
-      g.text("R=#{sprintf('%0.3f',r)}")
-      g.text("R^2=#{sprintf('%0.3f',r2)}")
-      
-      g.text(_("Equation")+"="+ sprintf('%0.3f',constant) +" + "+ @fields.collect {|k| sprintf('%0.3f%s',c[k],k)}.join(' + ') )
-
-      g.parse_element(f_test)
-
-      
-      
-      sc=standarized_coeffs
-      cse=coeffs_se
-      g.table(:name=>"Beta coefficients", :header=>%w{coeff b beta se t}.collect{|field| _(field)} ) do |t|
-        if (constant_se.nil?)
-          t.row([_("Constant"), sprintf("%0.3f", constant),"--","?","?"])
-        else
-          t.row([_("Constant"), sprintf("%0.3f", constant), "-", sprintf("%0.3f", constant_se), sprintf("%0.3f", constant_t)])
-        end
-        
-        @fields.each do |f|
-          t.row([f, sprintf("%0.3f", c[f]), sprintf("%0.3f", sc[f]), sprintf("%0.3f", cse[f]), sprintf("%0.3f", c[f].quo(cse[f]))])
-        end  
-      end
-    end    
-  end
 end
 end
 end

data/lib/statsample/regression/multiple/rubyengine.rb

@@ -31,11 +31,9 @@ class RubyEngine < MatrixEngine
     @ds=ds
     @dy=ds[@y_var]
     @ds_valid=ds.dup_only_valid
+    @total_cases=@ds.cases
+    @valid_cases=@ds_valid.cases
     @ds_indep = ds.dup(ds.fields-[y_var])
-    
-#    p obtain_predictor_matrix
-#    p @matrix_x.correlation
-    
     set_dep_columns
   end
   
@@ -82,8 +80,6 @@ class RubyEngine < MatrixEngine
     @ds_indep.update_valid_data
     set_dep_columns
   end
-  
-
   # Standard error for constant
   def constant_se
     estimated_variance_covariance_matrix[0,0]

data/lib/statsample/regression/simple.rb

@@ -70,7 +70,7 @@ module Statsample
         end
       end
       def init_vectors(vx,vy)
-        @vx,@vy=Statsample.only_valid(vx,vy)
+        @vx,@vy=Statsample.only_valid_clone(vx,vy)
         x_m=@vx.mean
         y_m=@vy.mean
         num=den=0

data/lib/statsample/reliability.rb

@@ -7,9 +7,9 @@ module Statsample
         ds=ods.dup_only_valid
         n_items=ds.fields.size
         sum_var_items=ds.vectors.inject(0) {|ac,v|
-        ac+v[1].variance_sample }
+        ac+v[1].variance }
         total=ds.vector_sum
-        (n_items / (n_items-1).to_f) * (1-(sum_var_items/ total.variance_sample))
+        (n_items.quo(n_items-1)) * (1-(sum_var_items.quo(total.variance)))
       end
       # Calculate Chonbach's alpha for a given dataset
       # using standarized values for every vector.
@@ -17,16 +17,16 @@ module Statsample
 
       def cronbach_alpha_standarized(ods)
         ds=ods.dup_only_valid.fields.inject({}){|a,f|
-          a[f]=ods[f].vector_standarized; a
+          a[f]=ods[f].standarized; a
         }.to_dataset
         cronbach_alpha(ds)
       end
     end
     class ItemCharacteristicCurve
-      attr_reader :totals, :counts,:vector_total
+      attr_reader :totals, :counts, :vector_total
       def initialize (ds, vector_total=nil)
         vector_total||=ds.vector_sum
-        raise "Total size != Dataset size" if vector_total.size!=ds.cases
+        raise ArgumentError, "Total size != Dataset size" if vector_total.size!=ds.cases
         @vector_total=vector_total
         @ds=ds
         @totals={}
@@ -48,19 +48,21 @@ module Statsample
           i+=1
         end
       end
+      # Return a hash with p for each different value on a vector
       def curve_field(field, item)
         out={}
         item=item.to_s
-        @totals.each{|value,n|
+        @totals.each do |value,n|
           count_value= @counts[field][value][item].nil? ? 0 : @counts[field][value][item]
-          out[value]=count_value.to_f/n.to_f
-        }
+          out[value]=count_value.quo(n)
+        end
         out
       end
     end
     class ItemAnalysis
       attr_reader :mean, :sd,:valid_n, :alpha , :alpha_standarized
-      def initialize(ds)
+      attr_accessor :name
+      def initialize(ds,opts=Hash.new)
         @ds=ds.dup_only_valid
         @total=@ds.vector_sum
         @item_mean=@ds.vector_mean.mean
@@ -70,11 +72,15 @@ module Statsample
         @kurtosis=@total.kurtosis
         @sd = @total.sd
         @valid_n = @total.size
+        opts_default={:name=>"Reliability Analisis"}
+        @opts=opts_default.merge(opts)
+        @name=@opts[:name]
+        
         begin
           @alpha = Statsample::Reliability.cronbach_alpha(ds)
           @alpha_standarized = Statsample::Reliability.cronbach_alpha_standarized(ds)
         rescue => e
-          raise DatasetException.new(@ds,e), "Problem on calculate alpha"
+          raise DatasetException.new(@ds,e), "Error calculating alpha"
         end
       end
       # Returns a hash with structure
@@ -201,51 +207,33 @@ module Statsample
           a
         end
       end
-      def html_summary
-				html = <<EOF
-<p><strong>Summary for scale:</strong></p>
-<ul>
-<li>Items=#{@ds.fields.size}</li>
-<li>Total Mean=#{@mean}</li>
-<li>Item Mean=#{@item_mean}</li>
-<li>Std.Dv.=#{@sd}</li>
-<li>Median=#{@median}</li>
-<li>Skewness=#{sprintf("%0.3f",@skew)}</li>
-<li>Kurtosis=#{sprintf("%0.3f",@kurtosis)}</li>
-
-<li>Valid n:#{@valid_n}</li>
-<li>Cronbach alpha: #{@alpha}</li>
-</ul>
-<table><thead><th>Variable</th>
-
-<th>Mean</th>
-<th>StDv.</th>
-<th>Mean if deleted</th><th>Var. if
-deleted</th><th>	StDv. if
-deleted</th><th>	Itm-Totl
-Correl.</th><th>Alpha if
-deleted</th></thead>
-EOF
+      def summary
+        ReportBuilder.new(:no_title=>true).add(self).to_text
+      end
+      def report_building(builder)
+        builder.section(:name=>@name) do |s|
+          s.table(:name=>"Summary") do |t|
+            t.row ["Items", @ds.fields.size]
+            t.row ["Total Mean", @mean]
+            t.row ["Item Mean", @item_mean]
+            t.row ["S.D.", @sd]
+            t.row ["Median", @median]
+            t.row ["Skewness", "%0.4f" % @skew]
+            t.row ["Kurtosis", "%0.4f" % @kurtosis]
+            t.row ["Valid n", @valid_n]
+            t.row ["Cronbach's alpha", "%0.4f" % @alpha]
+            t.row ["Standarized Cronbach's alpha", "%0.4f" % @alpha_standarized]
+          end
+          itc=item_total_correlation
+          sid=stats_if_deleted
+          is=item_statistics
 
-        itc=item_total_correlation
-        sid=stats_if_deleted
-        is=item_statistics
-        @ds.fields.each {|f|
-          html << <<EOF
-          <tr>
-          <td>#{f}</td>
-          <td>#{sprintf("%0.5f",is[f][:mean])}</td>
-          <td>#{sprintf("%0.5f",is[f][:sds])}</td>
-          <td>#{sprintf("%0.5f",sid[f][:mean])}</td>
-          <td>#{sprintf("%0.5f",sid[f][:variance_sample])}</td>
-          <td>#{sprintf("%0.5f",sid[f][:sds])}</td>
-          <td>#{sprintf("%0.5f",itc[f])}</td>
-          <td>#{sprintf("%0.5f",sid[f][:alpha])}</td>
-          </tr>
-EOF
-        }
-        html << "</table><hr />"
-        html
+          s.table(:name=>"Items report", :header=>["item","mean","sd", "mean if deleted", "var if deleted", "sd if deleted"," item-total correl.", "alpha if deleted"]) do |t|
+            @ds.fields.each do |f|
+              t.row(["#{@ds[f].name}(#{f})", sprintf("%0.5f",is[f][:mean]), sprintf("%0.5f",is[f][:sds]), sprintf("%0.5f",sid[f][:mean]), sprintf("%0.5f",sid[f][:variance_sample]), sprintf("%0.5f",sid[f][:sds]),  sprintf("%0.5f",itc[f]), sprintf("%0.5f",sid[f][:alpha])])
+            end
+          end
+          end
       end
     end
   end

data/lib/statsample/test.rb

@@ -8,11 +8,12 @@ module Statsample
     autoload(:F, 'statsample/test/f')
     # Returns probability of getting a value lower or higher
     # than sample, using cdf and number of tails.
-    # * For one tail left, return the cdf
-    # * For one tail right, return 1-cdf
-    # * For two tails, returns 2*right_tail(cdf.abs)
+    # 
+    # * <tt>:left</tt> : For one tail left, return the cdf
+    # * <tt>:right</tt> : For one tail right, return 1-cdf
+    # * <tt>:both</tt> : For both tails, returns 2*right_tail(cdf.abs)
     def p_using_cdf(cdf, tails=:both)
-      tails=:both if tails==2
+      tails=:both if tails==2 or tails==:two
       tails=:right if tails==1 or tails==:positive
       tails=:left if tails==:negative
       case tails
@@ -40,8 +41,11 @@ module Statsample
         }
         sum
       end
-      def u_mannwhitney(v1p,v2p)
-        Statsample::Test::UMannWhitney.new(v1p,v2p)
+      # Shorthand for Statsample::Test::UMannWhitney.new
+      # 
+      # * <tt>v1</tt> and <tt>v2</tt> should be Statsample::Vector.
+      def u_mannwhitney(v1, v2)
+        Statsample::Test::UMannWhitney.new(v1,v2)
       end
       # Shorthand for Statsample::Test::T::OneSample.new
       def t_one_sample(vector, opts=Hash.new)