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

From |
"Verkuilen, Jay" <JVerkuilen@gc.cuny.edu> |

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

Subject |
RE: st: RE: Brute force |

Date |
Mon, 16 Mar 2009 16:08:15 -0400 |

I'm not sure what you mean by "brute force." This is a nonlinear regression, and you haven't fully specified the model. It's not obviously linearizable so you'll need to use nonlinear least squares, -nl-, assuming that the model is really Ai=B0*Ci*(Di^B1) + ei, where ei is a Gaussian error. If not, you need to figure out what the error term is. Multiplicative error would be Ai=B0*Ci*(Di^B1)*exp(ei). I'd be worried about identification so you should make very sure to give good starting values. Also, this model may not be appropriate for all variables. In particular if Ai, Ci, and Di aren't ratio scales (e.g., time, mass, etc.) I am not sure it makes any sense at all. JV -----Original Message----- From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Shun Chonabayashi Sent: Monday, March 16, 2009 3:51 PM To: statalist@hsphsun2.harvard.edu Subject: RE: st: RE: Brute force Thank you so much for your replies. Then, for example, if I have observations of variables Ai, Ci and Di and want to obtain the best estimates for constants B0 and B1 in the formula below, do you think I can apply "brute force"? If so, could you let me know how to command this in Stata? Thanks for your time. Ai=B0*Ci*(Di^B1) Shun -----Original Message----- From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Nick Cox Sent: Monday, March 16, 2009 2:15 PM To: statalist@hsphsun2.harvard.edu Subject: RE: st: RE: Brute force Jeph is based in the USA. That may not be evident from his email address. I think we agree, w.r.t. Shun's question, is that there isn't a prescription for applying brute force, any more than there is a prescription for elegance or simplicity. Nick n.j.cox@durham.ac.uk Jeph Herrin This may be a transatlantic distinction, but on this side of the pond, "brute force" is often both admirable and desirable in all respects except efficiency. That is, the brute force method may be the best approach but too costly (in terms of time or resources) to implement. In analysis, at least; in areas such as diplomacy it indeed lacks an admirable connotation. Nick Cox wrote: > This question probably arises as a understandable puzzle about someone's > English usage, possibly mine! > > Some approach may be described as "brute force", but that means only > that the technique will work, but is otherwise not admirable or > desirable. > > There isn't a converse recipe for "brute force". > > Some idea of context would be helpful. > > Nick > n.j.cox@durham.ac.uk > > Shun Chonabayashi > > Could anybody explain how to do "brute force" in Stata or refer to any > page which explains about it? Thanks. > * * 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/ * * 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/ * * 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/

**Follow-Ups**:**Re: st: RE: Brute force***From:*David Airey <david.airey@vanderbilt.edu>

**References**:**st: Brute force***From:*"Shun Chonabayashi" <schomeirin@yahoo.co.jp>

**st: RE: Brute force***From:*"Nick Cox" <n.j.cox@durham.ac.uk>

**Re: st: RE: Brute force***From:*Jeph Herrin <junk@spandrel.net>

**RE: st: RE: Brute force***From:*"Nick Cox" <n.j.cox@durham.ac.uk>

**RE: st: RE: Brute force***From:*"Shun Chonabayashi" <schomeirin@yahoo.co.jp>

- Prev by Date:
**Re: st: Is there a similar output command as outreg for Summary?** - Next by Date:
**st: Brute force** - Previous by thread:
**RE: st: RE: Brute force** - Next by thread:
**Re: st: RE: Brute force** - Index(es):

© Copyright 1996–2017 StataCorp LLC | Terms of use | Privacy | Contact us | What's new | Site index |