Phoenix contains four Linear models used to perform linear regression. In the following table the y-unit is the preferred unit of the y (concentration) variable and the t-unit is the preferred unit of the t (time) variable. See “Units panel” to set preferred units for the output parameters.
The linear models do not require dosing information.
Note:The default minimization method, Gauss-Newton (Hartley) (located in the Engine Settings tab), and the Parameter Boundaries option Do Not Use Bounds are recommended for all linear models.
Linear models include:
y(t)=CONSTANT
Estimated Parameters: CONSTANT (y-unit)
y(t)=INT+SLOPE*t
Estimated Parameters: INT (y-unit), SLOPE (y-unit/t-unit)
y(t)=A0+A1*t+A2*t2
Estimated Parameters: A0 (y-unit), A1 (y-unit/t-unit), A2 (y-unit/t-unit2)
y(t)=A0+A1*t+A2*t2+A3*t3
Estimated Parameters: A0 (y-unit), A1 (y-unit/t-unit), A2 (y-unit/t-unit2), A3 (y-unit/t-unit3)
