Class: Statsample::MLE::BaseMLE

Inherits:
Object
  • Object
show all
Defined in:
lib/statsample/mle.rb

Direct Known Subclasses

Logit, Normal, Probit

Constant Summary collapse

ITERATIONS = 
100
MIN_DIFF = 
1e-5
MIN_DIFF_PARAMETERS = 
1e-2

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize ⇒ BaseMLE

Model should be a MLE subclass



30
31
32
33
34
35
36
37
 
# File 'lib/statsample/mle.rb', line 30

def initialize()
  @verbose        = false
  @output         = STDOUT
  @stop_criteria  = :parameters
  @var_cov_matrix = nil
  @iterations     = nil
  @parameters     = nil
end

Instance Attribute Details

#iterations ⇒ Object (readonly)

Iterations



23
24
25
 
# File 'lib/statsample/mle.rb', line 23

def iterations
  @iterations
end

#output ⇒ Object

Returns the value of attribute output.



17
18
19
 
# File 'lib/statsample/mle.rb', line 17

def output
  @output
end

#parameters ⇒ Object (readonly)

Parameters (beta coefficients)



25
26
27
 
# File 'lib/statsample/mle.rb', line 25

def parameters
  @parameters
end

#stop_criteria ⇒ Object

Could be :parameters or :mle



19
20
21
 
# File 'lib/statsample/mle.rb', line 19

def stop_criteria
  @stop_criteria
end

#var_cov_matrix ⇒ Object (readonly)

Variance - Covariance matrix



21
22
23
 
# File 'lib/statsample/mle.rb', line 21

def var_cov_matrix
  @var_cov_matrix
end

#verbose ⇒ Object

Returns the value of attribute verbose.



16
17
18
 
# File 'lib/statsample/mle.rb', line 16

def verbose
  @verbose
end

Instance Method Details

#likehood(x, y, b) ⇒ Object

Calculate likehood for matrices x and y, given b parameters



39
40
41
42
43
44
45
46
47
48
 
# File 'lib/statsample/mle.rb', line 39

def likehood(x,y,b)
  prod=1
  x.row_size.times{|i|
    xi=Matrix.rows([x.row(i).to_a.collect{|v| v.to_f}])
    y_val=y[i,0].to_f
    fbx=f(b,x)
    prod=prod*likehood_i(xi, y_val ,b)
  }
  prod
end

#log_likehood(x, y, b) ⇒ Object

Calculate log likehood for matrices x and y, given b parameters



50
51
52
53
54
55
56
57
58
 
# File 'lib/statsample/mle.rb', line 50

def log_likehood(x,y,b)
  sum=0
  x.row_size.times{|i|
    xi=Matrix.rows([x.row(i).to_a.collect{|v| v.to_f}])
    y_val=y[i,0].to_f
    sum+=log_likehood_i(xi,y_val,b)
  }
  sum
end

#newton_raphson(x, y, start_values = nil) ⇒ Object

Newton Raphson with automatic stopping criteria. Based on: Von Tessin, P. (2005). Maximum Likelihood Estimation With Java and Ruby

x

matrix of dependent variables. Should have nxk dimensions

y

matrix of independent values. Should have nx1 dimensions

@m

class for @ming. Could be Normal or Logit

start_values

matrix of coefficients. Should have 1xk dimensions



75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
 
# File 'lib/statsample/mle.rb', line 75

def newton_raphson(x,y, start_values=nil)
  # deep copy?
  if start_values.nil?
      parameters=set_default_parameters(x)
  else
      parameters = start_values.dup
  end
  k=parameters.row_size
  cv=Matrix.rows([([1.0]*k)])
  last_diff=nil
  raise "n on y != n on x" if x.row_size!=y.row_size
  h=nil
  fd=nil
  if @stop_criteria==:mle
    old_likehood=log_likehood(x, y, parameters)
  else
    old_parameters=parameters
  end
  ITERATIONS.times do |i|
    @iterations=i+1
    puts "Set #{i}" if @verbose
    h = second_derivative(x,y,parameters)
    if h.singular?
      raise "Hessian is singular!"
    end
    fd = first_derivative(x,y,parameters)
    parameters = parameters-(h.inverse*(fd))
    
    if @stop_criteria==:parameters
    flag=true
    k.times do |j|
      diff= ( parameters[j,0] - old_parameters[j,0] ) / parameters[j,0]
      flag=false if diff.abs >= MIN_DIFF_PARAMETERS
      @output.puts "Parameters #{j}: #{diff}" if @verbose
    end
    if flag
      @var_cov_matrix = h.inverse*-1.0
      return parameters
    end
    old_parameters=parameters
    else
      begin
        new_likehood = log_likehood(x,y,parameters)
        @output.puts "[#{i}]Log-MLE:#{new_likehood} (Diff:#{(new_likehood-old_likehood) / new_likehood})" if @verbose
        if(new_likehood < old_likehood) or ((new_likehood - old_likehood) / new_likehood).abs < MIN_DIFF
            @var_cov_matrix = h.inverse*-1.0
        #@output.puts "Ok"
            break;
        end
        old_likehood=new_likehood
      rescue =>e
        puts "#{e}"
        #puts "dup"
      end
    end
  end
  @parameters=parameters
  parameters
end

#set_default_parameters(x) ⇒ Object

Creates a zero matrix Mx1, with M=x.M



62
63
64
65
66
 
# File 'lib/statsample/mle.rb', line 62

def set_default_parameters(x)
  fd=[0.0]*x.column_size
  fd.push(0.1)    if self.is_a? Statsample::MLE::Normal
  parameters = Matrix.columns([fd])
end