CLsqFit Class Description

The CLsqFit class performs least squares analysis of sample data. This class uses the linear least squares method to fit up to 100 coefficients for 1 to 10 independent variables. Options include fitting single channel or RGB data, built-in or user-defined basis functions, iterative data rejection, point editing, and forcing coefficients to specified values.

Despite the power and versatility of this class, its user interface is remarkably simple. For example, the AddPt method adds data points to the sample; using various argument lists, the same method is used to specify 1 to 10 values of the independent variables as well as the observed value for either numeric or RGB data.

CLsqFit Class Topics

Using the CLsqFit Class

Basis Functions

Using Multiple Independent Variables

Working with RGB data

Forcing Coefficient Values

Rejecting Outliers from the Fit

Using CLsqFit with Image Pixels

Plotting a Least Squares Fit

Usage

Construction

Object = new_lsqfit()

Object = CLsqFit:new()

Destruction

Object:delete()

Creation - Non-class Functions

This function is normally used to create a new CLsqFit object as an alternative to the class-based new method.

new_lsqfit

Creates a new instance of the CLsqFit class.

Construction and Destruction - Class Methods

new	Creates a new instance of the CLsqFit class.
delete	Deletes the instance of the CLsqFit object.

Operations

Configuring the Fit

SetNumCoefs	Sets the number of coefficients and, optionally, the number of independent variables, or "basis dimensions".
GetNumCoefs	Returns the number of coefficients in total or for the specified basis dimension.
SetBasisFunc	Specifies the basis function. The options are an n-dimensional polynomial (the default), an n-dimensional hyperplane or a user provided function.
GetBasisFunc	Returns the basis function that is selected.
GetBasisDim	Returns the number of independent variables, or "basis dimensions".
SetNumChannels	Sets the number of data channels to fit. The options are 1 (the default), 3, or 4.
GetNumChannels	Returns the number of channels being fit.

Computing the Fit

Fit	Computes the fit.
ResetPoints	Initializes the fit data by deleting all points that have been added.
DoRejection	Enables (the default) or disables iterative data rejection.
ForceCoef	Forces a coefficient to a specified value and does not estimate it in the fit.
UnforceCoef	Restores a forced coefficient to the pool of coefficients being fit.
IsCoefForced	Returns the state of a coefficient (estimated or forced).
ResetForcedCoefs	Resets all forced coefficients so they all will be estimated by the fit.
GetErrMsg	Returns an error message if the fit fails.
GetErrChan	Returns the channel number where the fit failed.

Working with Fit Results

GetCoef	Returns a coefficient value from the fit.
GetCoefErr	Returns the error in the coefficient value from the fit.
GetCovar	Returns a covariance from the fit covariance matrix.
GetSigmaFit	Returns the standard deviation of the fit.
Eval	Evaluates the predicted value of a specified point using the fit results.
GetResid	Returns the residual for a point used in the fit.
GetSampleMean	Returns the mean value for the specified basis dimension.
GetRangeX	Returns the min and max values of the independent variable(s).
GetRangeObs	Returns the min and max values of the observations.
GetNumPts	Returns the number of points added to the sample.
GetNumPtsUsed	Returns the number of points used in the fit.
GetNumPtsRejected	Returns the number of points rejected from the fit when data rejection is enabled.
GetNumPtsDeleted	Returns the number of points that were manually deleted.
GetPolyCoefIndex	Returns the index of a coefficient based on its dimensional index (used only for the default n-dimensional polynomial basis function).

Working with Sample Points

AddPt	Adds a point to the sample data. The point is added at the next point index.
AddPtWt	Adds a point to the sample data with a point weight. The point is added at the next point index.
GetPtX	Returns the values of the independent variable for a point by the point's index.
SetPtX	Sets the values of the independent variables for a point by the point's index.
GetPtObs	Returns the observed value of a sample point by the point's index.
SetPtObs	Sets the observed value at an existing point by the point's index.
GetPtWt	Returns the weight of a point.
SetPtWt	Sets the weight of a point by the point's index.
GetPtStatus	Returns the status code of a sample point. This is a number 0, 1, or 2 for used, deleted, or rejected. It is useful when data rejection is enabled.
GetPtStatusStr	Returns the status string of a sample point as "use", "rej", or "del" (used, deleted, or rejected). It is useful when data rejection is enabled.
GetX	Returns an table containing all the point X (basis) values for one of the basis dimensions.
GetObs	Returns a table containing all the point Y (observed) values.
GetWt	Rreturns a table containing all the point weights.
SetPtStatus	Sets the status of an existing data point to used, rejected, or deleted.
FindPt	Returns the index of a point based on a match to its values.
DeletePt	Manually deletes a point from the fit.
UndeletePt	Undeletes a point that was deleted from the fit.

Configuring Data Rejection

DoRejection	Enables (the default) or disables iterative data rejection.
SetRejDefault	Sets default values for data rejection parameters.
SetRejSigmaHigh	Sets the high sigma value used for Rejecting Outliers having residuals above the fit.
GetRejSigmaHigh	Returns the high sigma value used for Rejecting Outliers having residuals above the fit.
SetRejSigmaLow	Sets the low sigma value used for Rejecting Outliers having residuals below the fit.
GetRejSigmaLow	Returns the low sigma value used for Rejecting Outliers having residuals below the fit.
SetRejCycles	Sets the number of cycles used for Rejecting Outliers from the fit.
GetRejCycles	Returns the number of cycles used for Rejecting Outliers from the fit.