Packages

c

scalation.analytics

NonLinRegression

class NonLinRegression extends Predictor with Error

The NonLinRegression class supports non-linear regression. In this case, 'x' can be multi-dimensional '[1, x1, ... xk]' and the function 'f' is non-linear in the parameters 'b'. Fit the parameter vector 'b' in the regression equation

y = f(x, b) + e

where 'e' represents the residuals (the part not explained by the model). Use Least-Squares (minimizing the residuals) to fit the parameter vector 'b' by using Non-linear Programming to minimize Sum of Squares Error 'SSE'.

See also

www.bsos.umd.edu/socy/alan/stats/socy602_handouts/kut86916_ch13.pdf

Linear Supertypes
Error, Predictor, AnyRef, Any
Ordering
  1. Alphabetic
  2. By Inheritance
Inherited
  1. NonLinRegression
  2. Error
  3. Predictor
  4. AnyRef
  5. Any
  1. Hide All
  2. Show All
Visibility
  1. Public
  2. All

Instance Constructors

  1. new NonLinRegression(x: MatrixD, y: VectorD, f: (VectoD, VectoD) ⇒ Double, b_init: VectorD)

    x

    the input/design matrix augmented with a first column of ones

    y

    the response vector

    f

    the non-linear function f(x, b) to fit

    b_init

    the initial guess for the parameter vector b

Value Members

  1. def coefficient: VectoD

    Return the vector of coefficient/parameter values.

    Return the vector of coefficient/parameter values.

    Definition Classes
    Predictor
  2. def diagnose(yy: VectoD): Unit

    Compute diagostics for the regression model.

    Compute diagostics for the regression model.

    yy

    the response vector

    Definition Classes
    NonLinRegression → Predictor
  3. def fit: VectorD

    Return the quality of fit.

    Return the quality of fit.

    Definition Classes
    NonLinRegression → Predictor
  4. def fitLabels: Seq[String]

    Return the labels for the fit.

    Return the labels for the fit.

    Definition Classes
    NonLinRegression → Predictor
  5. final def flaw(method: String, message: String): Unit
    Definition Classes
    Error
  6. def predict(z: VectoD): Double

    Predict the value of y = f(z) by evaluating the formula y = f(z, b), i.e.0, (b0, b1) dot (1.0, z1).

    Predict the value of y = f(z) by evaluating the formula y = f(z, b), i.e.0, (b0, b1) dot (1.0, z1).

    z

    the new vector to predict

    Definition Classes
    NonLinRegression → Predictor
  7. def predict(z: VectoI): Double

    Given a new discrete data vector z, predict the y-value of f(z).

    Given a new discrete data vector z, predict the y-value of f(z).

    z

    the vector to use for prediction

    Definition Classes
    Predictor
  8. def residual: VectoD

    Return the vector of residuals/errors.

    Return the vector of residuals/errors.

    Definition Classes
    Predictor
  9. def sseF(b: VectoD): Double

    Function to compute the Sum of Squares Error 'SSE' for given values for the parameter vector 'b'.

    Function to compute the Sum of Squares Error 'SSE' for given values for the parameter vector 'b'.

    b

    the parameter vector

  10. def train(yy: VectoD): Unit

    Train the predictor by fitting the parameter vector (b-vector) in the non-linear regression equation for the response passed into the class 'y'.

    Train the predictor by fitting the parameter vector (b-vector) in the non-linear regression equation for the response passed into the class 'y'.

    yy

    the response vector

    Definition Classes
    NonLinRegression → Predictor
  11. def train(): Unit

    Train the predictor by fitting the parameter vector (b-vector) in the non-linear regression equation

    Train the predictor by fitting the parameter vector (b-vector) in the non-linear regression equation

    y = f(x, b)

    using the least squares method. Caveat: Optimizer may converge to an unsatisfactory local optima. If the regression can be linearized, use linear regression for starting solution.

    Definition Classes
    NonLinRegression → Predictor