Fit #

Fits the model to training data using least squares.

method : public : Fit(X:Float[,], y:Float[,]) ~ Bool

Parameters

Return

Example

X := [[1.0, 1.0], [1.0, 2.0], [1.0, 3.0]];
y := [[2.0], [4.0], [6.0]];
model := LinearRegression->New();
if(model->Fit(X, y)) {
  "Fitted. R²={$model->GetRSquared()}"->PrintLine();
};

Fit #

Fits the model to training data, optionally estimating an intercept term. When fit_intercept is true a leading column of 1.0s is prepended to the design matrix and coefficient 0 holds the intercept. Computes R-squared, adjusted R-squared, MSE and RMSE; returns false on a singular / ill- conditioned system (the underlying solver yields no solution).

method : public : Fit(X:Float[,], y:Float[,], fit_intercept:Bool) ~ Bool

Parameters

Return

Example

X := [[1.0], [2.0], [3.0]];
y := [[3.0], [5.0], [7.0]];   # y = 2x + 1
model := LinearRegression->New();
model->Fit(X, y, true);       # intercept ~ 1, slope ~ 2
"RMSE={$model->GetRMSE()}"->PrintLine();

GetAdjustedRSquared #

Gets the adjusted R-squared, which penalizes additional predictors.

method : public : GetAdjustedRSquared() ~ Float

Return

GetCoefficients #

Gets the computed coefficients.

method : public : GetCoefficients() ~ Float[]

Return

Example

coeffs := model->GetCoefficients();
each(c in coeffs) {
  c->PrintLine();
};

GetMSE #

Gets the mean squared error over the training data.

method : public : GetMSE() ~ Float

Return

GetRMSE #

Gets the root mean squared error over the training data.

method : public : GetRMSE() ~ Float

Return

GetRSquared #

Gets the R-squared goodness of fit.

method : public : GetRSquared() ~ Float

Return

Example

"Model R²={$model->GetRSquared()}"->PrintLine();

IsFitted #

Whether the model has been fitted.

method : public : IsFitted() ~ Bool

Return

Load # function

Loads a fitted model from a file.

function : Load(filename:String) ~ LinearRegression

Parameters

Return

New # constructor

Constructor

New()

Example

model := LinearRegression->New();

Predict #

Predicts target values for new input data.

method : public : Predict(X:Float[,]) ~ Float[,]

Parameters

Return

Example

preds := model->Predict([[1.0, 5.0], [1.0, 6.0]]);
each(i : preds->Size()[0]) {
  preds[i,0]->PrintLine();
};

Score #

Scores the model on the given data, returning the R-squared (coefficient of determination) of its predictions against the true targets.

method : public : Score(X:Float[,], y:Float[,]) ~ Float

Parameters

Return

Store #

Saves the fitted model (coefficients, intercept convention and training metrics) to a file.

method : public : Store(filename:String) ~ Bool

Parameters

Return