[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

From |
"Nick Cox" <n.j.cox@durham.ac.uk> |

To |
<statalist@hsphsun2.harvard.edu> |

Subject |
RE: st: Calculating area under a curve |

Date |
Fri, 4 Dec 2009 08:55:00 -0000 |

Thanks for the clarification, which is very helpful. On a small point of terminology, -cumul- and -integ- are to Stata commands, not functions. Nick n.j.cox@durham.ac.uk Padmakumar Sivadasan I realized after reading your initial post that what I was not on a producutive path and so I shelved it. I did find the -cumul- and -integ- functions ,that I was not aware of, useful for other projects of mine. On Thu, Dec 3, 2009 at 12:49 PM, Nick Cox <n.j.cox@durham.ac.uk> wrote: > It's kind of you to thank us all but you do not distinguish between quite different comments. > > Others can speak for themselves for Steve (Samuels) and I couldn't see why this would be interesting or useful. (Steve's at liberty to dissent if that's not his view.) In effect, you are integrating twice, as -cumul- is clearly numerical integration of the density function, even though not named as such, and you then integrate once more. > > Could you please explain why this is a good idea? > > Nick > n.j.cox@durham.ac.uk > > Padmakumar Sivadasan > > Thank you all (Martin, Philippe, Nick, Steve, Maarten, Bob) for your > valuable suggestions! > > As suggested, I used -cumul- to calculate the cumulative distributions > and -integ- to calculate the area under the curve.. > > On Fri, Nov 27, 2009 at 8:22 AM, Padmakumar Sivadasan > <pmathru2@illinois.edu> wrote: >> Dear All, >> >> I am analyzing the performance of companies indicated by a variable >> v1. Variable v1 has a range 0-10 where higher values indicate poorer >> performance. I am attempting to compare the performance of companies >> for the country as a whole and to that at the local level >> (Metropolitan Statistical Area). I am interested not only in the mean >> value of v1 but also the variability of v1. One suggestion I got was >> to compute the cumulative probabilities at the national and local >> levels and then compare the area under the cumulative probability >> distributions at the local level to that at the national level. >> >> I understand that I can use the -cumul- function in Stata to calculate >> the cumulative probabilities but I couldn't find a method to calculate >> the area under the cumulative probability curve. I have two questions >> in this regard >> (1) Is there a way in Stata to calculate the area under cumulative >> probability curve? >> (2) Could someone point me to reference that I can use to read up on >> this method? * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

**References**:**Re: st: Calculating area under a curve***From:*Padmakumar Sivadasan <pmathru2@illinois.edu>

**RE: st: Calculating area under a curve***From:*"Nick Cox" <n.j.cox@durham.ac.uk>

**Re: st: Calculating area under a curve***From:*Padmakumar Sivadasan <pmathru2@illinois.edu>

- Prev by Date:
**st: calculating p-values from pcorr's returned results** - Next by Date:
**st: Fwd: Calculating weighted kappa for multiple raters** - Previous by thread:
**Re: st: Calculating area under a curve** - Next by thread:
**st: "no; data in memory would be lost" warning after preserve?** - Index(es):

© Copyright 1996–2015 StataCorp LP | Terms of use | Privacy | Contact us | What's new | Site index |