Rules for initial parameter values

Why rules?

Before it can perform nonlinear regression, Prism must have initial values for each parameter in the equation. You can enter initial values at the time you fit curves, but it is helpful to define rules for generating the initial values at the time you enter a new equation. Then Prism will calculate the initial values automatically. If you don't enter rules for initial values, you will need to enter the initial values for every variable, for every data set, every time you fit data.

Rule syntax

While entering or editing a user-defined equation, click on the tab labeled "Rules for initial values". For each parameter, enter a number in the first column and select a multiplier from the drop-down list in the second column.

The first choice on the drop-down list is "(Initial value, to be fit)". This means that the value you entered will be the initial value for all data sets. The initial value will not depend on the range of the data.

The rest of the choices on the drop-down list are used to calculate the initial value from your data. The abbreviation YMIN is the minimum value of Y; YMAX is the maximum value, and YMID is the average of YMIN and YMAX. For example, if you enter "0.5" in the first column and select "YMAX" in the second column, Prism sets the initial value to half of YMAX (which differs for each data set).

No rules

If you don't enter rules for every parameter, Prism will insist that you enter initial values each time you fit data to the equation.