Fitting Life Distribution mixture of two Weibull with custom anchor point

[#1]

I am trying to fit a mixture of two weibull distributions to get two separate alphas and betas from my data but I want to determine what part of my data should the first weibull should fit to and what part of my data should the second weibull should fit, something like following. What it currently does with the Fit Mixture 2 two weibull ingredients is determines the first and second weibulls on its own

[#2]

Hi @FrequencyBison9 : Can you provide more information re exactly what you want to get out of this? 

If you are fitting a mixture of two distributions, then the pdf(mixture) = p1*pdf1 + p2*pdf2 where p1+ p2 = 1 (p1>=0, p2>=0) . Estimating all the parameters (in your case, 5) is usually done via maximum likelihood. Defining and estimating an "anchor point" is not part of that; a mixture distribution is not estimated via fitting a distribution up to some anchor, then fitting another distribution from that anchor forward (though it's temping to think of it that way!). 

What would you like to do with this pdf(mixture)?  Why do you want to "customize"?

Edit: If you want to do what you are describing, you may be able to do it via non-linear regression. It's a fairly complex topic, but maybe this will get you started.

https://community.jmp.com/t5/JMP-Blog/Fitting-piecewise-functions-with-JMP-s-Nonlinear-platform/ba-p...