Update tests

Author: xEnVrE

test/test_Gaussian_Density_UVR/main.cpp

@@ -69,9 +69,15 @@ class SimulatedMeasurement : public AdditiveMeasurementModel
     }
 
 
-    std::pair<std::size_t, std::size_t> getOutputSize() const override
+    VectorDescription getInputDescription() const override
     {
-        return std::pair<std::size_t, std::size_t>(measurement_.size(), 0);
+        return VectorDescription(3, 0, measurement_.size());
+    }
+
+
+    VectorDescription getMeasurementDescription() const override
+    {
+        return VectorDescription(measurement_.size());
     }
 
 private:
@@ -87,12 +93,6 @@ int main()
     {
         std::cout << "Constructing a simulated scenario..." << std::endl;
 
-        // Unscented Transform parameters
-        double alpha = 1.0;
-        double beta = 2.0;
-        double kappa = 0.0;
-        sigma_point::UTWeight ut_weight(3, alpha, beta, kappa);
-
         // Predicted state
         Gaussian predicted_state(3);
         Matrix3d covariance;
@@ -121,6 +121,12 @@ int main()
 
         SimulatedMeasurement measurement_model(measurement, predicted);
 
+        // Unscented Transform parameters
+        double alpha = 1.0;
+        double beta = 2.0;
+        double kappa = 0.0;
+        sigma_point::UTWeight ut_weight(measurement_model.getInputDescription().noiseless_description(), alpha, beta, kappa);
+
         // Propagate belief
         GaussianMixture predicted_measurement(1, 6);
         MatrixXd cross_covariance;

test/test_UKF/main.cpp

@@ -131,7 +131,7 @@ int main(int argc, char* argv[])
     /* Step 2.2 - Define the prediction step. */
 
     /* Initialize the unscented Kalman filter prediction step and pass the ownership of the state model */
-    std::unique_ptr<UKFPrediction> ukf_prediction = utils::make_unique<UKFPrediction>(std::move(wna), state_size, alpha, beta, kappa);
+    std::unique_ptr<UKFPrediction> ukf_prediction = utils::make_unique<UKFPrediction>(std::move(wna), alpha, beta, kappa);
 
 
     /* Step 3 - Correction */
@@ -158,7 +158,7 @@ int main(int argc, char* argv[])
         simulated_linear_sensor->enable_log("./", "testUKF");
 
     /* Step 3.3 - Initialize the unscented Kalman filter correction step and pass the ownership of the measurement model. */
-    std::unique_ptr<UKFCorrection> ukf_correction = utils::make_unique<UKFCorrection>(std::move(simulated_linear_sensor), state_size, alpha, beta, kappa);
+    std::unique_ptr<UKFCorrection> ukf_correction = utils::make_unique<UKFCorrection>(std::move(simulated_linear_sensor), alpha, beta, kappa);
 
 
     /* Step 4 - Assemble the unscented Kalman filter. */

test/test_UPF/main.cpp

@@ -159,7 +159,7 @@ int main(int argc, char* argv[])
 
     /* Initialize the kalman particle filter prediction step that wraps a Gaussian prediction step,
        in this case an unscented kalman filter prediction step. */
-    std::unique_ptr<GaussianPrediction> upf_prediction = utils::make_unique<UKFPrediction>(std::move(wna), state_size, alpha, beta, kappa);
+    std::unique_ptr<GaussianPrediction> upf_prediction = utils::make_unique<UKFPrediction>(std::move(wna), alpha, beta, kappa);
     std::unique_ptr<PFPrediction> gpf_prediction = utils::make_unique<GPFPrediction>(std::move(upf_prediction));
 
 
@@ -200,7 +200,7 @@ int main(int argc, char* argv[])
 
     /* Initialize the particle filter correction step that wraps a Guassian correction step,
        in this case an unscented kalman filter correction step. */
-    std::unique_ptr<GaussianCorrection> upf_correction = utils::make_unique<UKFCorrection>(std::move(simulated_linear_sensor), state_size, alpha, beta, kappa);
+    std::unique_ptr<GaussianCorrection> upf_correction = utils::make_unique<UKFCorrection>(std::move(simulated_linear_sensor), alpha, beta, kappa);
     std::unique_ptr<PFCorrection> gpf_correction = utils::make_unique<GPFCorrection>(std::move(exp_likelihood), std::move(upf_correction), std::move(transition_probability_model));
 
 

test/test_UPF_MAP/main.cpp

@@ -191,7 +191,7 @@ int main(int argc, char* argv[])
 
     /* Initialize the kalman particle filter prediction step that wraps a Gaussian prediction step,
        in this case an unscented kalman filter prediction step. */
-    std::unique_ptr<GaussianPrediction> upf_prediction = utils::make_unique<UKFPrediction>(std::move(wna), state_size, alpha, beta, kappa);
+    std::unique_ptr<GaussianPrediction> upf_prediction = utils::make_unique<UKFPrediction>(std::move(wna), alpha, beta, kappa);
     std::unique_ptr<PFPrediction> gpf_prediction = utils::make_unique<GPFPrediction>(std::move(upf_prediction));
 
 
@@ -232,7 +232,7 @@ int main(int argc, char* argv[])
 
     /* Initialize the particle filter correction step that wraps a Guassian correction step,
        in this case an unscented kalman filter correction step. */
-    std::unique_ptr<GaussianCorrection> upf_correction = utils::make_unique<UKFCorrection>(std::move(simulated_linear_sensor), state_size, alpha, beta, kappa);
+    std::unique_ptr<GaussianCorrection> upf_correction = utils::make_unique<UKFCorrection>(std::move(simulated_linear_sensor), alpha, beta, kappa);
     std::unique_ptr<PFCorrection> gpf_correction = utils::make_unique<GPFCorrection>(std::move(exp_likelihood), std::move(upf_correction), std::move(transition_probability_model));
 
 

test/test_mixed_KF_SUKF/main.cpp

@@ -156,7 +156,7 @@ int main(int argc, char* argv[])
         simulated_linear_sensor->enable_log("./", "testKF_SUKF");
 
     /* Step 3.3 - Initialize the serial unscented Kalman filter correction step and pass the ownership of the measurement model. */
-    std::unique_ptr<GaussianCorrection> sukf_correction = utils::make_unique<SUKFCorrection>(std::move(simulated_linear_sensor), state_size, alpha, beta, kappa, 2, true);
+    std::unique_ptr<GaussianCorrection> sukf_correction = utils::make_unique<SUKFCorrection>(std::move(simulated_linear_sensor), alpha, beta, kappa, 2, true);
 
 
     /* Step 4 - Assemble the serial unscented Kalman filter */

test/test_mixed_KF_UKF/main.cpp

@@ -157,7 +157,7 @@ int main(int argc, char* argv[])
         simulated_linear_sensor->enable_log("./", "testKF_UKF");
 
     /* Step 3.3 - Initialize the unscented Kalman filter correction step and pass the ownership of the measurement model. */
-    std::unique_ptr<UKFCorrection> ukf_correction = utils::make_unique<UKFCorrection>(std::move(simulated_linear_sensor), state_size, alpha, beta, kappa);
+    std::unique_ptr<UKFCorrection> ukf_correction = utils::make_unique<UKFCorrection>(std::move(simulated_linear_sensor), alpha, beta, kappa);
 
 
     /* Step 4 - Assemble the mixed Kalman filter. */

test/test_mixed_UKF_KF/main.cpp

@@ -132,7 +132,7 @@ int main(int argc, char* argv[])
     /* Step 2.2 - Define the prediction step. */
 
     /* Initialize the unscented Kalman filter prediction step and pass the ownership of the state model. */
-    std::unique_ptr<UKFPrediction> ukf_prediction = utils::make_unique<UKFPrediction>(std::move(wna), state_size, alpha, beta, kappa);
+    std::unique_ptr<UKFPrediction> ukf_prediction = utils::make_unique<UKFPrediction>(std::move(wna), alpha, beta, kappa);
 
 
     /* Step 3 - Correction */