2

I'm looking for clarification on the definition of truncated distributions and on terminology for censored distributions and truncated distributions. I recently had a [dialog on SO][1] regarding [a relatively ancient posting by Prof. Brian Ripley on Rhelp][2] where Ripley had suggested dividing a Weibull density by the Weibull cumulative mass function (CMF) ( in R: pweibull(trunc_point, shape , scale ) ) evaluated at the truncation point. After looking at it a bit longer I decided he probably meant instead to use the expression 1-pweibull(trunc_point, , ) as the denominator for a left-truncated distribution.

After looking at the original posting I see that he included the argument lower=FALSE, so the CMF was being evaluated as I thought it should be with the function value being integral of the density to the left of the first argument. My efforts at levity were not appreciated and I've answered my question, but I see no means of deleting a question which I have now answered for myself.

DWin
  • 7,005
  • 17
  • 32
  • 5
    Please replace your title with one that relates to your question -- a descriptive one that will be more help for people looking answers to similar questions. Please also expand abbreviations on first use where they might not be widely known (MPH, CMF) ... I assume the first is a doctorate in medicine, but it might be a masters in philosophy or something else altogether, I can't be sure. – Glen_b Jan 23 '15 at 05:01
  • Also clarify your question. It sounds like a question to prof. Ripley "what did he meant by this?". If you have a general question about truncated Weilbull please state it and provide a background so we do not have read the discussions you link so to understand what is your question. – Tim Jan 23 '15 at 10:35
  • Also closely related: http://stats.stackexchange.com/questions/113273 and http://stats.stackexchange.com/questions/82495 – whuber Jan 23 '15 at 16:44

0 Answers0