I have a dataset that describes the estimated number of seeds being released from a tree on a minute by minute basis, and environmental variables such as wind speed, temperature and relative humidity that have been recorded at the same time as the seed release. I would like to build a multivariate regression model such that some combination of the environmental variables predict how many seeds are released in a minute.

My seed data is interval censored - while taking the data I just had a number of categories that described that the number of seeds seen were in a certain interval. My categories were:

0 1 2-9 10-29 30-99 100-299 300-999 1000-2999 3000-10000

So as you see I only know what the number of seeds is between an interval, I do not know the exact number of seeds released.

I basically have no idea how to do this although I have been trying to figure it out. My stats background is very weak. It seems that generally people are dealing with interval censored data are using it in the context of survival time and I have found very little information about using it in a context more like mine and I am not sure if the same techniques can be used. Additionally I could only find information about if the data is univariate - not multivariate.

Any help in any direction would be really appreciated. I am using R so anything R-specific would be good.

  • $\begingroup$ By "multivariate", do you mean you have multiple predictors you would like to include? Or that the response for each individual has several values? $\endgroup$ – Cliff AB Jul 13 '18 at 1:59
  • $\begingroup$ Thanks for your comment. I was referring to multiple predictors, not multiple response. $\endgroup$ – agorapotatoes Jul 27 '18 at 0:05

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.