1. Home
  2. Nonlinear Quantile Regression Standard Curves

Nonlinear Quantile Regression Standard Curves

This dialog lets you specify a standard curve for the Nonlinear Quantile Regression menu. Once you have chosen the particular model and completed the parameter initial values and bounds, this fills in the terms and parameters using the parameterization as specified in the FITCURVE directive. For robustness of fitting some models, transforms of the parameters are fitted, but the initial values and bounds are given as reported by FITCURVE. The initial values for the parameters could be obtained by fitting the standard regression model using the Standard Curves menu.

Model

The models to be fitted belong to five general classes of curve; each has several variations. Typical graphs of these models are given here. The help pages below give the formulae and descriptions of these models.

Explanatory variate

The explanatory of x variate for the model must be specified here. The name of the variate can either be typed in, or selected from the dropdown list of variates which match in length to the response variate. This variate will be placed in the model formulae and in the variate to plot the fitted values against.

Nonlinear parameters

These fields give the following information on each nonlinear parameter in the model. Linear parameters do not require initial values or bounds. Initial values are not required if the simplex is being used, but otherwise must be specified. Bounds must be given when the simplex algorithm is being used, but otherwise are optional.

Parameter Name of the nonlinear parameter
Initial value Initial value for the parameter in the optimization algorithm
Lower limit Lower limit of parameter
Upper limit Upper limit of parameter

Action Icons

Clear Clear all fields and list boxes.
Help Open the Help topic for this dialog.

See also

Updated on March 28, 2019

Was this article helpful?