Main Page | Modules | Class Hierarchy | Alphabetical List | Class List | File List | Class Members | File Members

Huber


Classes

class  HuberData
class  HuberLinsys
class  HuberResiduals
class  HuberVars

Detailed Description

Solve the linear Huber regression problem.

Input consists of dimensions "nobservations" (number of observations) and "npredictors" (number of predictor variables); a matrix of dimensions nobservations x npredictors containing values of the predictors, a right-hand side vector of dimension "nobservations"; and a cutoff value. The cutoff value represents the point at which the loss function turns from being a least-squares function to an absolute-value function.


Generated on Wed Mar 22 13:58:35 2006 for OOQP by doxygen 1.3.5