add comupte regression mean logic

Author: nawazishkhan1-nk

Libs/Particles/ParticleShapeStatistics.cpp

@@ -492,6 +492,27 @@ ParticleShapeStatistics::ParticleShapeStatistics(std::shared_ptr<Project> projec
     groups.push_back(1);
   }
   import_points(points, groups);
+  // TODO: importing regression params doesn't make sense here. take a look again later.
+}
+
+ Eigen::VectorXd ParticleShapeStatistics::compute_regression_mean(const std::vector<double>& explanatory_variables) const
+{
+      // Map explanatory variables to an Eigen vector
+    Eigen::VectorXd t = Eigen::Map<const Eigen::VectorXd>(explanatory_variables.data(), explanatory_variables.size());
+    
+    // Ensure slope and intercept are initialized
+    if (slope_.size() == 0 || intercept_.size() == 0) {
+        throw std::runtime_error("Slope and Intercept not initialized yet!");
+    }
+
+    // Handle scalar and vector cases for t
+    if (t.size() == 1) {
+        return slope_ + intercept_ * t[0]; // Scalar broadcasting
+    } else if (t.size() == slope_.size()) {
+        return slope_ + intercept_.cwiseProduct(t); // Element-wise multiplication
+    } else {
+        throw std::invalid_argument("Size mismatch: t must be either scalar or match dimensions of slope and intercept.");
+    }
 }
 
 //---------------------------------------------------------------------------

Libs/Particles/ParticleShapeStatistics.h

@@ -87,7 +87,9 @@ class ParticleShapeStatistics {
   const Eigen::VectorXd& get_mean_rel_pos() { return mean_rel_pose_; }
 
   //! Returns the mean shape.
-  const Eigen::VectorXd& get_mean() const { return mean_; }
+  const Eigen::VectorXd get_mean() const { return mean_; }
+
+ Eigen::VectorXd compute_regression_mean(const std::vector<double>& explanatory_variables) const; 
   const Eigen::VectorXd& get_group1_mean() const { return mean1_; }
   const Eigen::VectorXd& get_group2_mean() const { return mean2_; }
 
@@ -136,7 +138,7 @@ class ParticleShapeStatistics {
   void set_meshes(const std::vector<Mesh>& meshes) { meshes_ = meshes; }
 
   // import estimated parameters for regression
-  inline bool import_regression_parameters(Eigen::VectorXd slope, Eigen::VectorXd intercept) { slope_ = slope; intercept_ = intercept; };
+  inline bool import_regression_parameters(Eigen::VectorXd& slope, Eigen::VectorXd& intercept) { slope_ = slope; intercept_ = intercept; return true;};
 
  private:
   unsigned int num_samples_group1_;

Studio/Analysis/AnalysisTool.cpp

@@ -664,10 +664,13 @@ bool AnalysisTool::compute_stats() {
   can_run_regression_ = check_explanatory_variable_limits();
   if (can_run_regression_) {
     auto slope = load_regression_parameters(
-        session_->get_regression_param_file("slope"));
+        session_->get_regression_param_file("slope")); // dM vector
     auto intercept = load_regression_parameters(
-        session_->get_regression_param_file("intercept"));
-    stats_.import_regression_parameters(slope, intercept);
+        session_->get_regression_param_file("intercept")); // dM vector
+    stats_.import_regression_parameters(slope, intercept); // set slope and intercept in stats object
+    ui_->regression_groupbox->setVisible(true);
+    ui_->explanatoryVariableSlider->setVisible(true);
+    ui_->enableRegressionCheckBox->setVisible(true);
   }
   else {
     ui_->regression_groupbox->setVisible(false);
@@ -790,7 +793,7 @@ Particles AnalysisTool::get_shape_points(int mode, double value) {
     ui_->explained_variance->setText("");
     ui_->cumulative_explained_variance->setText("");
   }
-  auto mean = !regression_enabled_ ? stats_.get_mean() : stats_.get_regression_mean(ui_->get_explanatory_variable_value());
+  auto mean = !regression_enabled_ ? stats_.get_mean() : stats_.compute_regression_mean(ui_->get_explanatory_variable_value());
   temp_shape_ = mean + (e * (value * lambda));
 
   auto positions = temp_shape_;
@@ -1157,9 +1160,9 @@ double AnalysisTool::get_pca_value() {
 }
 
 
-double AnalysisTool::get_explanatory_variable_value() {
+std::vector<double> AnalysisTool::get_explanatory_variable_value() {
   int slider_value = ui_->explanatoryVariableSlider->value();
-  return t_min + (static_cast<double>(slider_value) / 100.0) * (t_max - t_min);
+  return {t_min + (static_cast<double>(slider_value) / 100.0) * (t_max - t_min)};
 
 }
 

Studio/Analysis/AnalysisTool.h

@@ -77,6 +77,8 @@ class AnalysisTool : public QWidget {
 
   double get_pca_value();
 
+  std::vector<double> get_explanatory_variable_value();
+
   bool pca_animate();
   McaMode get_mca_level() const;
 
@@ -282,7 +284,7 @@ class AnalysisTool : public QWidget {
   std::string feature_map_;
 
   std::vector<double> explanatory_variable_limits_;
-  bool can_run_regression_;
+  bool can_run_regression_; // decide if necessary variables are present to run regression in analysis
 
   std::vector<std::string> current_group_names_;
   std::vector<std::string> current_group_values_;