statsample 0.3.0

data/lib/statsample/graph/svggraph.rb

@@ -0,0 +1,181 @@
+require 'SVG/Graph/Bar'
+require 'SVG/Graph/BarHorizontal'
+require 'SVG/Graph/Pie'
+require 'SVG/Graph/Line'
+require 'SVG/Graph/Plot'
+require 'statsample/graph/svghistogram'
+
+module Statsample
+	class Nominal
+		# Creates a barchart using ruby-gdchart
+		def svggraph_frequencies(file, width=600, height=300, chart_type=SVG::Graph::BarNoOp, options={})
+			labels,data=[],[]
+			self.frequencies.sort.each{|k,v|
+				labels.push(k.to_s)
+				data.push(v) 
+			}
+            options[:height]=height
+            options[:width]=width
+            options[:fields]=labels  
+			graph = chart_type.new(options)
+			graph.add_data(
+            :data => data,
+            :title => "Frequencies"
+			)
+			File.open(file,"w") {|f|
+              f.puts(graph.burn)
+			}
+		end
+	end
+	class Scale < Ordinal
+		def svggraph_histogram(bins, options={})
+            options={:graph_title=>"Histogram", :show_graph_title=>true,:show_normal=>true, :mean=>self.mean, :sigma=>sdp }.merge! options
+            graph = Statsample::Graph::SvgHistogram.new(options)
+            graph.histogram=histogram(bins)
+            graph
+		end
+        # Returns a Run-Sequence Plot
+        # Reference: http://www.itl.nist.gov/div898/handbook/eda/section3/runseqpl.htm
+        def svggraph_runsequence_plot(options={})
+            options={:graph_title=>"Run-Sequence Plot", :show_graph_title=>true, :scale_x_integers => true, :add_popups=>true }.merge! options
+            vx=(1..@data.size).to_a.to_vector(:scale)
+            vy=@data.to_vector(:scale)
+            ds={'index'=>vx,'value'=>vy}.to_dataset            
+            graph = Statsample::Graph::SvgScatterplot.new(ds,options)
+            graph.set_x('index')
+            graph.parse
+            graph
+        end
+        def svggraph_boxplot(options={})
+            options={:graph_title=>"Boxplot", :fields=>['vector'], :show_graph_title=>true}.merge! options
+            vx=@data.to_a.to_vector(:scale)
+            graph = Statsample::Graph::SvgBoxplot.new(options)
+            graph.add_data(:title=>"vector", :data=>@data.to_a)
+            graph
+        end
+        
+        def svggraph_lag_plot(options={})
+            options={:graph_title=>"Lag Plot", :show_graph_title=>true}.merge! options
+            vx=@data[0...(@data.size-1)].to_vector(:scale)
+            vy=@data[1...@data.size].to_vector(:scale)
+            ds={'x_minus_1'=>vx,'x'=>vy}.to_dataset            
+            graph = Statsample::Graph::SvgScatterplot.new(ds,options)
+            graph.set_x('x_minus_1')
+            graph.parse
+            graph
+        end
+        
+        # Returns a Normal Probability Plot
+        # Reference: http://www.itl.nist.gov/div898/handbook/eda/section3/normprpl.htm
+        def svggraph_normalprobability_plot(options={})
+                    extend Statsample::Util
+
+                    options={:graph_title=>"Normal Probability Plot", :show_graph_title=>true}.merge! options
+            n=@data.size
+            vx=(1..@data.size).to_a.collect{|i|
+                GSL::Cdf.gaussian_Pinv(normal_order_statistic_medians(i,n))
+            }.to_vector(:scale)
+            vy=@data.sort.to_vector(:scale)
+            ds={'normal_order_statistics_medians'=>vx, 'ordered_response'=>vy}.to_dataset            
+            graph = Statsample::Graph::SvgScatterplot.new(ds,options)
+            graph.set_x('normal_order_statistics_medians')
+            graph.parse
+            graph
+        end
+	end
+end
+
+# replaces all key and fill classes with similar ones, without opacity
+# this allows rendering of svg and png on rox and gqview without problems
+module SVG
+	module Graph
+		class BarNoOp < Bar
+			def get_css; SVG::Graph.get_css_standard; end
+		end
+		class BarHorizontalNoOp < BarHorizontal
+			def get_css; SVG::Graph.get_css_standard; end
+		end
+		
+		class LineNoOp < Line
+			def get_css; SVG::Graph.get_css_standard; end
+			
+		end
+        class PlotNoOp < Plot
+			def get_css; SVG::Graph.get_css_standard; end            
+        end
+		class PieNoOp < Pie
+			def get_css; SVG::Graph.get_css_standard; end
+			
+		end
+		class << self
+		def get_css_standard
+        return <<EOL
+/* default fill styles for multiple datasets (probably only use a single dataset on this graph though) */
+.key1,.fill1{
+	fill: #ff0000;
+	stroke: none;
+	stroke-width: 0.5px;	
+}
+.key2,.fill2{
+	fill: #0000ff;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key3,.fill3{
+	fill: #00ff00;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key4,.fill4{
+	fill: #ffcc00;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key5,.fill5{
+	fill: #00ccff;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key6,.fill6{
+	fill: #ff00ff;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key7,.fill7{
+	fill: #00ffff;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key8,.fill8{
+	fill: #ffff00;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key9,.fill9{
+	fill: #cc6666;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key10,.fill10{
+	fill: #663399;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key11,.fill11{
+	fill: #339900;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key12,.fill12{
+	fill: #9966FF;
+	stroke: none;
+	stroke-width: 1px;	
+}
+EOL
+      	end
+	end
+	end
+end
+
+require 'statsample/graph/svgscatterplot'
+require 'statsample/graph/svgboxplot'

data/lib/statsample/graph/svghistogram.rb

@@ -0,0 +1,208 @@
+module Statsample
+    module Graph
+class SvgHistogram < SVG::Graph::BarBase
+    attr_accessor :inner_margin, :mean, :sigma, :show_normal
+        def initialize(config)
+            config[:fields]=[:dummy]
+            super(config)
+            
+            @histogram=nil
+        end
+		include REXML
+		
+			# In addition to the defaults set in Graph::initialize, sets
+			# [inner_margin] 14
+			# [key] false
+		def set_defaults 
+		 super
+		 self.top_align = self.top_font = 0
+         init_with({
+            :inner_margin=>16,
+            :key=>false,
+            :show_normal=>false
+            })
+         
+		end
+		
+        def histogram=(h)
+            @histogram=h
+            @data=[{:data=>(0...@histogram.bins).to_a.collect {|i|
+                @histogram[i]
+            }}]
+        end
+		def get_x_labels
+            [""]
+		end
+        
+		def get_y_labels
+            maxvalue = max_value
+            minvalue = min_value
+            range = maxvalue - minvalue
+            
+            top_pad = range == 0 ? 10 : range / 20.0
+            scale_range = (maxvalue + top_pad) - minvalue
+            
+            scale_division = scale_divisions || (scale_range / 10.0)
+            
+            if scale_integers
+            scale_division = scale_division < 1 ? 1 : scale_division.round
+            end
+            
+            rv = []
+            maxvalue = maxvalue%scale_division == 0 ? 
+            maxvalue : maxvalue + scale_division
+            minvalue.step( maxvalue, scale_division ) {|v| rv << v}
+            return rv
+		end
+		
+        def unit_width
+            (@graph_width-(@inner_margin*2)) / (@histogram.max-@histogram.min).to_f
+        end
+        def draw_x_label(v)
+            left = (v - @histogram.min)*unit_width                    
+            x=@inner_margin+left
+            text = @graph.add_element( "text" )
+            text.attributes["class"] = "xAxisLabels"
+            text.text = sprintf("%0.2f",v)
+            y = @graph_height + x_label_font_size + 3
+            text.attributes["x"] = x.to_s
+            text.attributes["y"] = y.to_s
+            
+        end
+	    def draw_x_labels
+            if show_x_labels
+                (0...@histogram.bins).each { |i|
+                    value = @histogram[i]
+                    range = @histogram.get_range(i)
+                    draw_x_label(range[0])
+                    if(i==(@histogram.bins)-1)
+                        draw_x_label(range[1])
+                    end
+                }
+            end
+        end
+		def draw_data
+            minvalue = min_value
+            fieldwidth = field_width
+    
+            unit_size =  (@graph_height.to_f - font_size*2*top_font) / 
+                          (get_y_labels.max - get_y_labels.min)		
+            bottom = @graph_height
+            field_count = 0
+            hist_min=@histogram.min
+            hist_max=@histogram.max
+            range_hist=hist_max-hist_min
+            total=0
+            (0...@histogram.bins).each { |i|
+                dataset_count = 0
+                value = @histogram[i]
+                total=total+value
+                range = @histogram.get_range(i)
+                left = (range[0] - hist_min)*unit_width
+                bar_width = (range[1] - hist_min)*unit_width - left
+                length = (value.abs - (minvalue > 0 ? minvalue : 0)) * unit_size
+                # top is 0 if value is negative
+                top = bottom - (((value < 0 ? 0 : value) - minvalue) * unit_size)
+                @graph.add_element( "rect", {
+                    "x" => (@inner_margin+left).to_s,
+                    "y" => top.to_s,
+                    "width" => bar_width.to_s,
+                    "height" => length.to_s,
+                    "class" => "fill#{dataset_count+1}"
+                })
+                make_datapoint_text(left + @inner_margin+ (bar_width/2), top - 6, value.to_s)
+                field_count += 1
+            }
+            if(show_normal)
+                divs=30
+                path=""
+                0.upto(divs) {|i|
+                    x_abs=hist_min+(range_hist/divs)*i
+                    y=GSL::Ran::gaussian_pdf((x_abs-mean) / sigma)*total
+                    xg=@inner_margin+((x_abs-hist_min)*unit_width)
+                    yg=bottom-(y-minvalue)*unit_size
+                    if i==0
+                        path="M#{xg} #{yg} "
+                    else
+                        path+="L#{xg} #{yg} "
+                    end
+                }
+                @graph.add_element("path", {
+                        "d"=>path,
+                        "style"=>"stroke:black;fill:none"
+                }
+                )
+            end
+	end
+    
+    
+     def get_css
+        return <<EOL
+/* default fill styles for multiple datasets (probably only use a single dataset on this graph though) */
+
+.key1,.fill1{
+	fill: #ff0000;
+	stroke: black;
+	stroke-width: 1px;	
+}
+.key2,.fill2{
+	fill: #0000ff;
+	stroke: black;
+	stroke-width: 1px;	
+}
+.key3,.fill3{
+	fill: #00ff00;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key4,.fill4{
+	fill: #ffcc00;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key5,.fill5{
+	fill: #00ccff;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key6,.fill6{
+	fill: #ff00ff;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key7,.fill7{
+	fill: #00ffff;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key8,.fill8{
+	fill: #ffff00;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key9,.fill9{
+	fill: #cc6666;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key10,.fill10{
+	fill: #663399;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key11,.fill11{
+	fill: #339900;
+	stroke: none;
+	stroke-width: 1px;	
+}
+.key12,.fill12{
+	fill: #9966FF;
+	stroke: none;
+	stroke-width: 1px;	
+}
+EOL
+      	end
+
+end
+end
+end

data/lib/statsample/graph/svgscatterplot.rb

@@ -0,0 +1,111 @@
+module Statsample
+    module Graph
+        class SvgScatterplot < SVG::Graph::Plot
+            attr_accessor :draw_path
+            def initialize(ds,config={})
+                super(config)
+                @ds=ds
+                set_x(@ds.fields[0])
+            end
+            def set_defaults
+                super
+                init_with(
+                    :show_data_values => false,
+                    :draw_path => false
+                    )
+            end
+            def set_x(x)
+                @x=x
+                @y=@ds.fields - [x]
+            end
+            def parse
+                data=@y.inject({}){|a,v| a[v]=[];a}
+                @ds.each{|row|
+                    @y.each{|y|
+                        data[y]+=[row[@x],row[y]] unless row[@x].nil? or row[y].nil?
+                    }
+                }
+                data.each{|y,d|
+                    add_data({
+                            :data=>d, :title=>@ds.vector_label(y)
+                    })
+                }
+            end
+            def get_x_labels
+                values=super
+                values.collect{|x|
+                    if x.is_a? Integer
+                        x 
+                    else
+                        sprintf("%0.2f",x).to_f
+                    end
+                }
+            end
+            def get_y_labels
+                values=super
+                values.collect{|x|
+                    if x.is_a? Integer
+                        x 
+                    else
+                        sprintf("%0.2f",x).to_f
+                    end
+                }
+            end
+            def draw_data
+        line = 1
+
+        x_min, x_max, x_div = x_range
+        y_min, y_max, y_div = y_range
+        x_step = (@graph_width.to_f - font_size*2) / (x_max-x_min)
+        y_step = (@graph_height.to_f -  font_size*2) / (y_max-y_min)
+
+        for data in @data
+          x_points = data[:data][X]
+          y_points = data[:data][Y]
+
+          lpath = "L"
+          x_start = 0
+          y_start = 0
+          x_points.each_index { |idx|
+            x = (x_points[idx] -  x_min) * x_step
+            y = @graph_height - (y_points[idx] -  y_min) * y_step
+            x_start, y_start = x,y if idx == 0
+            lpath << "#{x} #{y} "
+          }
+
+          if area_fill
+            @graph.add_element( "path", {
+              "d" => "M#{x_start} #@graph_height #{lpath} V#@graph_height Z",
+              "class" => "fill#{line}"
+            })
+          end
+          if draw_path
+          @graph.add_element( "path", {
+            "d" => "M#{x_start} #{y_start} #{lpath}",
+            "class" => "line#{line}"
+          })
+          end
+          if show_data_points || show_data_values
+            x_points.each_index { |idx|
+              x = (x_points[idx] -  x_min) * x_step
+              y = @graph_height - (y_points[idx] -  y_min) * y_step
+              if show_data_points
+                @graph.add_element( "circle", {
+                  "cx" => x.to_s,
+                  "cy" => y.to_s,
+                  "r" => "2.5",
+                  "class" => "dataPoint#{line}"
+                })
+                add_popup(x, y, format( x_points[idx], y_points[idx] )) if add_popups
+              end
+              make_datapoint_text( x, y-6, y_points[idx] ) if show_data_values
+            }
+          end
+          line += 1
+        end
+      end
+            
+            
+        end
+    end
+end