Skip to content

TESTING.md

Latest commit

 

History

History
195 lines (140 loc) · 4.78 KB

TESTING.md

File metadata and controls

195 lines (140 loc) · 4.78 KB

Testing Guide for randomProcesses

This directory contains comprehensive test coverage for the randomProcesses MATLAB class.

Test Files

  • test_randomProcesses.m - Main test suite using MATLAB's unit testing framework
  • run_all_tests.m - Convenience script to run all tests with formatted output

Running Tests

Option 1: Using the test runner script (Recommended)

>> run_all_tests

Option 2: Using MATLAB's runtests function

>> results = runtests('test_randomProcesses')

Option 3: From command line

matlab -batch "run_all_tests"

Option 4: Run specific test methods

>> results = runtests('test_randomProcesses', 'Name', 'testBrownianPrices')

Test Coverage

The test suite covers:

1. Class Initialization (4 tests)

  • Constructor with default values
  • Constructor with custom values
  • Dependent property r0 (initial rate)
  • Dependent property dt (time delta)

2. Stock Price Methods (5 tests)

  • Brownian Motion (brownian_prices)
  • Geometric Brownian Motion (gbm_prices)
  • Merton Jump-Diffusion Model (merton_prices)
  • Heston Model (heston_prices)
  • Single path generation

3. Interest Rate Methods (2 tests)

  • Vasicek Interest Rate Model (vas_rates)
  • Cox-Ingersoll-Ross Model (cir_rates)

4. Utility Methods (4 tests)

  • Order Flow (order_flow)
  • Tick Imbalance Bars (tib)
  • Volume Imbalance Bars (vib)
  • Dollar Imbalance Bars (dib)

5. Input Validation (8 tests)

  • Invalid tick data format detection
  • Invalid window size detection
  • Parameter bounds checking (eta, M)
  • Vector shape conversion
  • Error message verification

6. Static Methods (1 test)

  • Exponential Weighted Moving Average (ewma)

7. Edge Cases (2 tests)

  • Zero price differences handling
  • Small dataset handling

8. Stochastic Volatility Options (3 tests)

  • Brownian with constant volatility
  • GBM with stochastic volatility
  • Merton with stochastic volatility

Total: 29 test methods

Test Structure

Each test method follows this pattern:

function testMethodName(testCase)
    % Test description
    
    % 1. Setup: Create test data
    % 2. Execute: Call the method being tested
    % 3. Verify: Assert expected results
    
    testCase.verifyEqual(actual, expected);
    testCase.verifySize(result, expectedSize);
    testCase.verifyError(@() method(), 'expectedErrorID');
end

Verification Methods Used

  • verifyEqual - Check exact equality (with tolerance)
  • verifySize - Check array dimensions
  • verifyTrue/verifyFalse - Check boolean conditions
  • verifyError - Check that expected errors are thrown
  • verifyGreaterThanOrEqual/verifyLessThanOrEqual - Check ordering
  • verifyLessThan - Check strict inequality

Adding New Tests

To add a new test:

  1. Open test_randomProcesses.m
  2. Add a new test method in the appropriate section:
function testNewFeature(testCase)
    % Test description
    result = testCase.sim.newMethod();
    testCase.verifyEqual(result, expectedValue);
end
  1. Run the tests to verify the new test works

Understanding Test Results

When you run tests, you'll see output like:

Running testConstructorDefaultValues
.
Running testBrownianPrices
.
...

Total Tests:  29
Passed:       29
Failed:       0
Incomplete:   0
Duration:     5.23 seconds

✓ All tests passed!

If a test fails, you'll see details:

Failed Tests:
  - testBrownianPrices
    Reason: Verification failed...
    Expected: [252, 5]
    Actual: [252, 4]

Continuous Integration

These tests can be integrated into CI/CD pipelines:

# Example GitHub Actions workflow
- name: Run MATLAB tests
  run: matlab -batch "run_all_tests"

Test Data

Tests use randomly generated data with fixed parameters to ensure reproducibility. Each test method runs independently with fresh instances created in the setup phase.

Known Limitations

  • Tests use random number generation, so exact values may vary between runs
  • Some stochastic processes may occasionally produce outliers
  • Tests verify dimensions, ranges, and statistical properties rather than exact values

Troubleshooting

Error: "Undefined function or variable 'randomProcesses'"

  • Solution: Make sure you're running from the repository root directory

Error: "Test framework not found"

  • Solution: Requires MATLAB R2013a or later with the Testing Framework

Tests run but some fail intermittently

  • This is expected with stochastic processes
  • Failing tests should still pass on retry
  • If consistently failing, check the implementation

Additional Resources