Repost: inverting the failure function

[burkom@comcast.net]

No one responded to the problem stated below a couple of weeks ago, so I'm trying one more time.  I see it as an interpolation of a vector into each of a large set of survival functions, but maybe there's another approach.  Or would another language or environment be more suitable for this problem?  Thanks in advance for any suggestions.

Howard Burkom

I am a STATA 7.0 user with a question on survival data.  I�ve been making failure function tables with commands like:

   

    sts list, by(covariate) failure at(1 2 to 10) compare

 

I would like to invert these failure functions using STATA.  In particular, I�d like to compute a row of values of the interpolated (or at least the nearest) times at which the failure function crosses each of a set of levels, say 0.5, 0.6,�,1.0, with a row like this for each of the many values of covariate.