0
$\begingroup$

I have a zero-inflated model with censored dataenter image description here

Is there a package in R that can combine zero- inflated model with right-censored data?

Alternatively, is there a general way to model right-censored data using zero-inflated/hurdle models?

$\endgroup$
  • 1
    $\begingroup$ Are these really zeros (=events at time 0)?! If the events occur after at least a tiny amount of time, but this somehow gets rounded to zero, then you could treat this as a case of the data being interval censored to lie within the time interval (0, lowest time that would not be zero). You may then simply need some kind of model that allows for an initially very high hazard rate. If we instead talk about people never getting the event, then that would be a "cure rate model". $\endgroup$ – Björn Jan 6 '17 at 9:32
  • 1
    $\begingroup$ Asking for R packages is off topic here. Regarding the unstated statistical issue, the Cox proportional hazards model handles censored data & makes no distributional assumptions, so it doesn't care if you have a lot of 0s. $\endgroup$ – gung Jan 6 '17 at 13:20
  • $\begingroup$ @Björn - yes, those are the events that occur in tiny time. $\endgroup$ – Michael Jan 7 '17 at 10:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.