class Statsample::Regression::Multiple::AlglibEngine

Class for Multiple Regression Analysis Requires Alglib gem and uses a listwise aproach. Faster than GslEngine on massive prediction use, because process is c-based. Prefer GslEngine if you need good memory use. If you need pairwise, use RubyEngine Example:

@a=[1,3,2,4,3,5,4,6,5,7].to_vector(:scale)
@b=[3,3,4,4,5,5,6,6,4,4].to_vector(:scale)
@c=[11,22,30,40,50,65,78,79,99,100].to_vector(:scale)
@y=[3,4,5,6,7,8,9,10,20,30].to_vector(:scale)
ds={'a'=>@a,'b'=>@b,'c'=>@c,'y'=>@y}.to_dataset
lr=Statsample::Regression::Multiple::AlglibEngine.new(ds,'y')

@a=[1,3,2,4,3,5,4,6,5,7].to_vector(:scale)
@b=[3,3,4,4,5,5,6,6,4,4].to_vector(:scale)
@c=[11,22,30,40,50,65,78,79,99,100].to_vector(:scale)
@y=[3,4,5,6,7,8,9,10,20,30].to_vector(:scale)
ds={'a'=>@a,'b'=>@b,'c'=>@c,'y'=>@y}.to_dataset
lr=Statsample::Regression::Multiple::AlglibEngine.new(ds,'y')

Public Class Methods

_load(data) click to toggle source

# File lib/statsample/regression/multiple/alglibengine.rb, line 47
def self._load(data)
    h=Marshal.load(data)
    self.new(h['ds'], h['y_var'])
end

new(ds,y_var, opts=Hash.new) click to toggle source

Calls superclass method

# File lib/statsample/regression/multiple/alglibengine.rb, line 20
def initialize(ds,y_var, opts=Hash.new)
  super    
  @ds=ds.dup_only_valid
  @ds_valid=@ds
  @dy=@ds[@y_var]
  @ds_indep=ds.dup(ds.fields-[y_var])
  # Create a custom matrix
  columns=[]
  @fields=[]
  @ds.fields.each{|f|
      if f!=@y_var
          columns.push(@ds[f].to_a)
          @fields.push(f)
      end
  }
  @dep_columns=columns.dup
  columns.push(@ds[@y_var])
  matrix=Matrix.columns(columns)
  @lr_s=nil
  @lr=::Alglib::LinearRegression.build_from_matrix(matrix)
  @coeffs=assign_names(@lr.coeffs)
  
end

Public Instance Methods

_dump(i) click to toggle source

# File lib/statsample/regression/multiple/alglibengine.rb, line 44
def _dump(i)
    Marshal.dump({'ds'=>@ds,'y_var'=>@y_var})
end

build_standarized() click to toggle source

# File lib/statsample/regression/multiple/alglibengine.rb, line 89
def build_standarized
    @ds_s=@ds.standarize
    columns=[]
    @ds_s.fields.each{|f|
        columns.push(@ds_s[f].to_a) unless f==@y_var
    }
    @dep_columns_s=columns.dup
    columns.push(@ds_s[@y_var])
    matrix=Matrix.columns(columns)
    @lr_s=Alglib::LinearRegression.build_from_matrix(matrix)
end

coeffs() click to toggle source

# File lib/statsample/regression/multiple/alglibengine.rb, line 52
def coeffs
    @coeffs
end

constant() click to toggle source

# File lib/statsample/regression/multiple/alglibengine.rb, line 76
def constant
    @lr.constant
end

lr_s() click to toggle source

# File lib/statsample/regression/multiple/alglibengine.rb, line 83
def lr_s
    if @lr_s.nil?
        build_standarized
    end
    @lr_s
end

matrix_resolution() click to toggle source

Coefficients using a constant Based on www.xycoon.com/ols1.htm

# File lib/statsample/regression/multiple/alglibengine.rb, line 57
def matrix_resolution
    mse_p=mse
    columns=@dep_columns.dup.map {|xi| xi.map{|i| i.to_f}}
    columns.unshift([1.0]*@ds.cases)
    y=Matrix.columns([@dy.data.map  {|i| i.to_f}])
    x=Matrix.columns(columns)
    xt=x.t
    matrix=((xt*x)).inverse*xt
    matrix*y
end

process(v) click to toggle source

# File lib/statsample/regression/multiple/alglibengine.rb, line 100
def process(v)
    @lr.process(v)
end

process_s(v) click to toggle source

# File lib/statsample/regression/multiple/alglibengine.rb, line 103
def process_s(v)
    lr_s.process(v)
end

r() click to toggle source

# File lib/statsample/regression/multiple/alglibengine.rb, line 70
def r
    Bivariate::pearson(@dy,predicted)
end

r2() click to toggle source

# File lib/statsample/regression/multiple/alglibengine.rb, line 67
def r2
    r**2
end

sst() click to toggle source

# File lib/statsample/regression/multiple/alglibengine.rb, line 73
def sst
    @dy.ss
end

standarized_coeffs() click to toggle source

# File lib/statsample/regression/multiple/alglibengine.rb, line 79
def standarized_coeffs
    l=lr_s
    assign_names(l.coeffs)
end

standarized_residuals() click to toggle source

???? Not equal to SPSS output

# File lib/statsample/regression/multiple/alglibengine.rb, line 107
def standarized_residuals
    res=residuals
    red_sd=residuals.sds
    res.collect {|v|
        v.quo(red_sd)
    }.to_vector(:scale)
end