Dear R users, I have seven regression lines I´d like to compare, in order to find out if these are significatively different. The main problem is that these are curves, non normal, non homogeneous data, I´ve tried to linearize them but it has not worked. So I´d like to know if you know any command or source in R which explains how to perform this kind of comparison. Thanks in advance for your help! [[alternative HTML version deleted]]
I'm very much a novice with R, and not a statistician. But as my group has been working with generalized additive models (semi-parametric regression), we have followed Wood's advice about using the R anova function to do model comparison for different regressions. I would imagine at least some of yours might be nested, e.g., adding polynomial functions to some? If so, model comparisons like this might be justified. Will Shadish On 7/29/2012 10:38 PM, Luis Fernando García Hernández wrote:> Dear R users, > > I have seven regression lines I´d like to compare, in order to find out if > these are significatively different. The main problem is that these are > curves, non normal, non homogeneous data, I´ve tried to linearize them but > it has not worked. So I´d like to know if you know any command or source in > R which explains how to perform this kind of comparison. > > Thanks in advance for your help! > > [[alternative HTML version deleted]] > > > > ______________________________________________ > R-help@r-project.org mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code.-- William R. Shadish Distinguished Professor Founding Faculty Chair, Psychological Sciences Mailing Address: William R. Shadish University of California School of Social Sciences, Humanities and Arts 5200 North Lake Rd Merced CA 95343 Physical/Delivery Address: University of California Merced ATTN: William Shadish School of Social Sciences, Humanities and Arts Facilities Services Building A 5200 North Lake Rd. Merced, CA 95343 209-228-4372 voice 209-228-4390 fax wshadish@ucmerced.edu http://faculty.ucmerced.edu/wshadish/index.htm http://psychology.ucmerced.edu [[alternative HTML version deleted]]
Hello, You could try some non-parametric tests for comparison such as Kolmogorov-Smirnov test (ks.test() package: truncgof)or Cramer-von Mises test (cvmts.pval() package: CvM2SL2Test). A.K. ----- Original Message ----- From: Luis Fernando Garc?a Hern?ndez <luysgarcia at gmail.com> To: r-help at r-project.org Cc: Sent: Monday, July 30, 2012 1:38 AM Subject: [R] curve comparison Dear R users, I have seven regression lines I?d like to compare, in order to find out if these are significatively different. The main problem is that these are curves,? non normal, non homogeneous data, I?ve tried to linearize them but it has not worked. So I?d like to know if you know any command or source in R which explains how to perform this kind of comparison. Thanks in advance for your help! ??? [[alternative HTML version deleted]] ______________________________________________ R-help at r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
Hi, Please check this article which details the method for comparison: Statistica Sinica 17(2007), 1115-1137 TESTING FOR THE EQUALITY OF k REGRESSION CURVES Juan Carlos Pardo-Fern ?ndez, Ingrid Van Keilegom and Wenceslao Gonz ?lez-Manteiga Also, check some examples (http://www.stat.umn.edu/geyer/old/5601/examp/kolmogorov.html). I am not very familiar with non-parametric tests. A.K. ________________________________ From: Luis Fernando Garc?a Hern?ndez <luysgarcia at gmail.com> To: arun <smartpink111 at yahoo.com> Sent: Monday, July 30, 2012 9:00 PM Subject: Re: [R] curve comparison Thanks both for your answer,? Will, do you have any reference which shows me how ?how to perform this analysis? I?d like to know ?to find out how to do it. In my case data is similar to compare several diversity curves, I do not know if you are familiar with them. Related to the Kolmogrov Smirnov test, I?ll try, I did not know it allowed curve comparison. This is a great space to share questions, so thanks both for your answer! 2012/7/30 arun <smartpink111 at yahoo.com> Hello,> >You could try some non-parametric tests for comparison such as Kolmogorov-Smirnov test (ks.test() package: truncgof)or Cramer-von Mises test (cvmts.pval() package: CvM2SL2Test). > > >A.K. > > >----- Original Message ----- >From: Luis Fernando Garc?a Hern?ndez <luysgarcia at gmail.com> >To: r-help at r-project.org >Cc: >Sent: Monday, July 30, 2012 1:38 AM >Subject: [R] curve comparison > >Dear R users, > >I have seven regression lines I?d like to compare, in order to find out if >these are significatively different. The main problem is that these are >curves,? non normal, non homogeneous data, I?ve tried to linearize them but >it has not worked. So I?d like to know if you know any command or source in >R which explains how to perform this kind of comparison. > >Thanks in advance for your help! > > >??? [[alternative HTML version deleted]] > > >______________________________________________ >R-help at r-project.org mailing list >https://stat.ethz.ch/mailman/listinfo/r-help >PLEASE do read the posting guide http://www.R-project.org/posting-guide.html >and provide commented, minimal, self-contained, reproducible code. > >