looking to create an alternative life distribution model, for example, a defective subpopulation exponential CDF with time offset of the form
F(t)=p(1-exp(-[(t-t0)/alpha])
with parameter p being the defective subpopulation fraction, t0 the time offset and alpha the time constant
where the platform would provide the full complement of profilers, covariance matrices, etc
is there a way to insert a custom life distribution such as this?
thanks
You cannot augment the model library in Life Distribution. You can fit practically any model in Nonlinear, though. Think of it as a general purpose maximum likelihood estimate procedure masquerading as a nonlinear curve fitter. The 'back door' is the definition of a custom model and loss function so that you can define the likelihood under censoring or truncation.
Please see Help > Books > Predictive and Specialized Modeling > Nonlinear chapter.