feat: add seeded model initialization

README.md

@@ -123,6 +123,20 @@ IrtRuby::TwoParameterModel.new(
   decay_factor: 0.5
 )
 ```
+
+### Reproducible Initialization
+Each model initializes parameters randomly. By default, constructors use Ruby's global random number generator, preserving the historical behavior and honoring any external `srand` calls. For reproducible model initialization without resetting or consuming global RNG state, pass `seed:`:
+
+```ruby
+model_a = IrtRuby::ThreeParameterModel.new(data, seed: 1234)
+model_b = IrtRuby::ThreeParameterModel.new(data, seed: 1234)
+
+# Same data, options, and seed produce identical fitted results.
+model_a.fit == model_b.fit #=> true
+```
+
+The `seed:` keyword is available for `RaschModel`, `TwoParameterModel`, and `ThreeParameterModel`.
+
 ### Parameter Clamping
 For 2PL and 3PL:
 

lib/irt_ruby/rasch_model.rb

@@ -18,7 +18,8 @@ def initialize(data,
                    param_tolerance: 1e-6,
                    learning_rate: 0.01,
                    decay_factor: 0.5,
-                   missing_strategy: :ignore)
+                   missing_strategy: :ignore,
+                   seed: nil)
       # data: A Matrix or array-of-arrays of responses (0/1 or nil for missing).
       # missing_strategy: :ignore (skip), :treat_as_incorrect, :treat_as_correct
 
@@ -36,10 +37,11 @@ def initialize(data,
       raise ArgumentError, "missing_strategy must be one of #{MISSING_STRATEGIES}" unless MISSING_STRATEGIES.include?(missing_strategy)
 
       @missing_strategy = missing_strategy
+      @random = seed.nil? ? nil : Random.new(seed)
 
       # Initialize parameters near zero
-      @abilities    = Array.new(num_rows)  { rand(-0.25..0.25) }
-      @difficulties = Array.new(num_cols)  { rand(-0.25..0.25) }
+      @abilities    = Array.new(num_rows)  { random_between(-0.25..0.25) }
+      @difficulties = Array.new(num_cols)  { random_between(-0.25..0.25) }
 
       @max_iter        = max_iter
       @tolerance       = tolerance
@@ -52,6 +54,11 @@ def sigmoid(x)
       1.0 / (1.0 + Math.exp(-x))
     end
 
+    def random_between(range)
+      @random ? @random.rand(range) : rand(range)
+    end
+    private :random_between
+
     def resolve_missing(resp)
       return [resp, false] unless resp.nil?
 

lib/irt_ruby/three_parameter_model.rb

@@ -20,7 +20,8 @@ def initialize(data,
                    param_tolerance: 1e-6,
                    learning_rate: 0.01,
                    decay_factor: 0.5,
-                   missing_strategy: :ignore)
+                   missing_strategy: :ignore,
+                   seed: nil)
       ModelOptionsValidator.validate!(max_iter: max_iter,
                                       tolerance: tolerance,
                                       param_tolerance: param_tolerance,
@@ -35,12 +36,13 @@ def initialize(data,
       raise ArgumentError, "missing_strategy must be one of #{MISSING_STRATEGIES}" unless MISSING_STRATEGIES.include?(missing_strategy)
 
       @missing_strategy = missing_strategy
+      @random = seed.nil? ? nil : Random.new(seed)
 
       # Initialize parameters
-      @abilities       = Array.new(num_rows)  { rand(-0.25..0.25) }
-      @difficulties    = Array.new(num_cols)  { rand(-0.25..0.25) }
-      @discriminations = Array.new(num_cols)  { rand(0.5..1.5) }
-      @guessings       = Array.new(num_cols)  { rand(0.0..0.3) }
+      @abilities       = Array.new(num_rows)  { random_between(-0.25..0.25) }
+      @difficulties    = Array.new(num_cols)  { random_between(-0.25..0.25) }
+      @discriminations = Array.new(num_cols)  { random_between(0.5..1.5) }
+      @guessings       = Array.new(num_cols)  { random_between(0.0..0.3) }
 
       @max_iter        = max_iter
       @tolerance       = tolerance
@@ -53,6 +55,11 @@ def sigmoid(x)
       1.0 / (1.0 + Math.exp(-x))
     end
 
+    def random_between(range)
+      @random ? @random.rand(range) : rand(range)
+    end
+    private :random_between
+
     # Probability for the 3PL model: c + (1-c)*sigmoid(a*(θ - b))
     def probability(theta, a, b, c)
       c + ((1.0 - c) * sigmoid(a * (theta - b)))

lib/irt_ruby/two_parameter_model.rb

@@ -16,7 +16,7 @@ class TwoParameterModel
 
     def initialize(data, max_iter: 1000, tolerance: 1e-6, param_tolerance: 1e-6,
                    learning_rate: 0.01, decay_factor: 0.5,
-                   missing_strategy: :ignore)
+                   missing_strategy: :ignore, seed: nil)
       ModelOptionsValidator.validate!(max_iter: max_iter,
                                       tolerance: tolerance,
                                       param_tolerance: param_tolerance,
@@ -31,12 +31,13 @@ def initialize(data, max_iter: 1000, tolerance: 1e-6, param_tolerance: 1e-6,
       raise ArgumentError, "missing_strategy must be one of #{MISSING_STRATEGIES}" unless MISSING_STRATEGIES.include?(missing_strategy)
 
       @missing_strategy = missing_strategy
+      @random = seed.nil? ? nil : Random.new(seed)
 
       # Initialize parameters
       # Typically: ability ~ 0, difficulty ~ 0, discrimination ~ 1
-      @abilities       = Array.new(num_rows)  { rand(-0.25..0.25) }
-      @difficulties    = Array.new(num_cols)  { rand(-0.25..0.25) }
-      @discriminations = Array.new(num_cols)  { rand(0.5..1.5) }
+      @abilities       = Array.new(num_rows)  { random_between(-0.25..0.25) }
+      @difficulties    = Array.new(num_cols)  { random_between(-0.25..0.25) }
+      @discriminations = Array.new(num_cols)  { random_between(0.5..1.5) }
 
       @max_iter         = max_iter
       @tolerance        = tolerance
@@ -49,6 +50,11 @@ def sigmoid(x)
       1.0 / (1.0 + Math.exp(-x))
     end
 
+    def random_between(range)
+      @random ? @random.rand(range) : rand(range)
+    end
+    private :random_between
+
     def resolve_missing(resp)
       return [resp, false] unless resp.nil?
 

spec/irt_ruby/rasch_model_spec.rb

@@ -5,6 +5,7 @@
 RSpec.describe IrtRuby::RaschModel do
   it_behaves_like "response data validation"
   it_behaves_like "model optimization option validation"
+  it_behaves_like "seeded model initialization", %i[abilities difficulties]
 
   let(:data_array) do
     [

spec/irt_ruby/three_parameter_model_spec.rb

@@ -5,6 +5,7 @@
 RSpec.describe IrtRuby::ThreeParameterModel do
   it_behaves_like "response data validation"
   it_behaves_like "model optimization option validation"
+  it_behaves_like "seeded model initialization", %i[abilities difficulties discriminations guessings]
 
   let(:data_array) do
     [

spec/irt_ruby/two_parameter_model_spec.rb

@@ -5,6 +5,7 @@
 RSpec.describe IrtRuby::TwoParameterModel do
   it_behaves_like "response data validation"
   it_behaves_like "model optimization option validation"
+  it_behaves_like "seeded model initialization", %i[abilities difficulties discriminations]
 
   let(:data_array) do
     [

spec/spec_helper.rb

@@ -59,6 +59,46 @@
   end
 end
 
+RSpec.shared_examples "seeded model initialization" do |parameter_names|
+  let(:seeded_fit_options) { { max_iter: 50, learning_rate: 0.05 } }
+
+  def seeded_parameter_snapshot(model, parameter_names)
+    parameter_names.to_h do |parameter_name|
+      [parameter_name, model.instance_variable_get("@#{parameter_name}").dup]
+    end
+  end
+
+  it "produces identical initial and fitted parameters with the same seed" do
+    model1 = described_class.new(data_array, **seeded_fit_options, seed: 12_345)
+    model2 = described_class.new(data_array, **seeded_fit_options, seed: 12_345)
+
+    expect(seeded_parameter_snapshot(model1, parameter_names)).to eq(
+      seeded_parameter_snapshot(model2, parameter_names)
+    )
+    expect(model1.fit).to eq(model2.fit)
+  end
+
+  it "produces different initial parameters with different seeds" do
+    model1 = described_class.new(data_array, **seeded_fit_options, seed: 12_345)
+    model2 = described_class.new(data_array, **seeded_fit_options, seed: 54_321)
+
+    expect(seeded_parameter_snapshot(model1, parameter_names)).not_to eq(
+      seeded_parameter_snapshot(model2, parameter_names)
+    )
+  end
+
+  it "does not reset or consume Ruby's global random number generator" do
+    srand(98_765)
+    expected_values = Array.new(5) { rand }
+
+    srand(98_765)
+    described_class.new(data_array, **seeded_fit_options, seed: 12_345)
+    actual_values = Array.new(5) { rand }
+
+    expect(actual_values).to eq(expected_values)
+  end
+end
+
 RSpec.configure do |config|
   # Enable flags like --only-failures and --next-failure
   config.example_status_persistence_file_path = ".rspec_status"