Hi All I received a raw data set with one record per tennis player (both male and female) and then i cured it by aggregation i.e by 4 age groups, 2 gender levels and 6 income levels. Gender and Income are categorical variables. Please advise me how to use 'R' to model this data set (Actually, i want to know the right regression technique and steps to do that, including removing outliers at the begining and looking for hidden interactions). Thanks Warm Regards Dev
1. response variable : rankings for a tennis player(not taking account career breaks/injury) 2 rankings Vs player performance & player performance (age, gender, income). Different gender might throw up surprises as in female tennis rankings, we see not much change in one playing season where as in men's tennis it's fluctuating. Also, i would like to which age group reflects the prime of a tennis player and hence i've changed continuous variables to categorical. Please advise Thanks -Dev>From: "Peter Flom" <flom at ndri.org> >To: <devshruti at hotmail.com> >Subject: Re: [R] Regression Modeling query >The first question is what you are trying to find out. What is your >dependent variable? What are your hypotheses? > >The second question is why you categorized continuous varaibles (age >and income)? This is almost always a bad idea > > > >Peter > >Peter L. Flom, PhD >Assistant Director, Statistics and Data Analysis Core >Center for Drug Use and HIV Research >National Development and Research Institutes >71 W. 23rd St >www.peterflom.com >New York, NY 10010 >(212) 845-4485 (voice) >(917) 438-0894 (fax) > > > > >>> "devshruti pahuja" <devshruti at hotmail.com> 6/22/2004 4:25:30 PM > >>> >Hi All > >I received a raw data set with one record per tennis player (both male >and >female) and then i cured it by aggregation i.e by 4 age groups, 2 >gender >levels and 6 income levels. Gender and Income are categorical >variables. >Please advise me how to use 'R' to model this data set (Actually, i >want to >know the right regression technique and steps to do that, including >removing >outliers at the begining and looking for hidden interactions). > >Thanks > >Warm Regards >Dev > >______________________________________________ >R-help at stat.math.ethz.ch mailing list >https://www.stat.math.ethz.ch/mailman/listinfo/r-help >PLEASE do read the posting guide! >http://www.R-project.org/posting-guide.html
Please ..... this is not a homework problem and i don't want spoon feeding but just the right way to approach the problem or the reference from regular texts. I'm doing this for my own learning and understanding of regression modeling. I'm familiar with most of the techniques of regression modeling but had a problem with my last project which had 17 predictors and i was able to unravel very few interactions despite using most of the techniques that were there in some standard textbooks. I want an insight from those who have modelled fuzzy data --> logical tables and then tried to get some behavioral insights. I don't mind coding my own R functions and possess a sound familiarity with Numerical Linear Algebra and Matrix theory. Secondly, i don't want to dig deep into books to find little useful. Any references that you provide will be pretty helpful Please advise -Dev>From: Berton Gunter <gunter.berton at gene.com> >To: devshruti pahuja <devshruti at hotmail.com> >Subject: Re: [R] Regression Modeling query >Date: Tue, 22 Jun 2004 13:44:13 -0700 >MIME-Version: 1.0 >X-Sender: "Berton Gunter" <bgunter at smtp.gene.com> >Received: from compton.gene.com ([192.12.78.250]) by mc8-f14.hotmail.com >with Microsoft SMTPSVC(5.0.2195.6824); Tue, 22 Jun 2004 13:45:15 -0700 >Received: from gene.com (dhcp164-33.gene.com [128.137.164.33])by >compton.gene.com (Switch-3.1.4/Switch-3.1.0) with ESMTP id >i5MKiDr3021487for <devshruti at hotmail.com>; Tue, 22 Jun 2004 13:44:13 -0700 >(PDT) >X-Message-Info: JGTYoYF78jHuGzi9i5wyM1lYLSA9WFC7 >Message-ID: <40D89A1D.EEB8EC99 at gene.com> >Organization: Genentech, Inc. >X-Mailer: Mozilla 4.78 [en]C-CCK-MCD (Windows NT 5.0; U) >X-Accept-Language: en >References: <BAY9-F42Xng77T4kfjD0003cf53 at hotmail.com> >Return-Path: bgunter at gene.com >X-OriginalArrivalTime: 22 Jun 2004 20:45:15.0723 (UTC) >FILETIME=[D27351B0:01C45899] > >devshruti pahuja wrote: > > > Hi All > > > > I received a raw data set with one record per tennis player (both male >and > > female) and then i cured it by aggregation i.e by 4 age groups, 2 gender > > levels and 6 income levels. Gender and Income are categorical variables. > > Please advise me how to use 'R' to model this data set (Actually, i want >to > > know the right regression technique and steps to do that, including >removing > > outliers at the begining and looking for hidden interactions). > > > >--- in which case, shouldn't we also sign and hand in the homework for you? > >-- Bert >
Perhaps you have not read the posting guide, as the footer suggested? There's a good reason why Bert said what he said. To tell you the truth, I was going to reply with something like: I'd like to know how to get R to cook dinner for me. Can anyone help? You really should learn how to ask questions in ways that will compell people to respond with useful replies. You have not told us the goal of the analysis, the purpose for the model, the context the data came from, etc., not to say that the question you asked is so general that it has little to do with R, other than the fact that you want to do it with R, and R-help is not really the place to ask general stat questions. You shouldn't expect free stat consulting from a software mailing list. Given the little info that you provided, I can recommend Prof. Harrell's book "Regression Modelling Strategies". You should read it over and over until you can't get it out of your head. Andy> From: devshruti pahuja > > Please ..... this is not a homework problem and i don't want > spoon feeding > but just the right way to approach the problem or the > reference from regular > texts. I'm doing this for my own learning and understanding > of regression > modeling. > > I'm familiar with most of the techniques of regression > modeling but had a > problem with my last project which had 17 predictors and i > was able to > unravel very few interactions despite using most of the > techniques that were > there in some standard textbooks. I want an insight from > those who have > modelled fuzzy data --> logical tables and then tried to get > some behavioral > insights. > > I don't mind coding my own R functions and possess a sound > familiarity with > Numerical Linear Algebra and Matrix theory. Secondly, i don't > want to dig > deep into books to find little useful. Any references that > you provide will > be pretty helpful > > Please advise > > -Dev > > > >From: Berton Gunter <gunter.berton at gene.com> > >To: devshruti pahuja <devshruti at hotmail.com> > >Subject: Re: [R] Regression Modeling query > >Date: Tue, 22 Jun 2004 13:44:13 -0700 > >MIME-Version: 1.0 > >X-Sender: "Berton Gunter" <bgunter at smtp.gene.com> > >Received: from compton.gene.com ([192.12.78.250]) by > mc8-f14.hotmail.com > >with Microsoft SMTPSVC(5.0.2195.6824); Tue, 22 Jun 2004 > 13:45:15 -0700 > >Received: from gene.com (dhcp164-33.gene.com [128.137.164.33])by > >compton.gene.com (Switch-3.1.4/Switch-3.1.0) with ESMTP id > >i5MKiDr3021487for <devshruti at hotmail.com>; Tue, 22 Jun 2004 > 13:44:13 -0700 > >(PDT) > >X-Message-Info: JGTYoYF78jHuGzi9i5wyM1lYLSA9WFC7 > >Message-ID: <40D89A1D.EEB8EC99 at gene.com> > >Organization: Genentech, Inc. > >X-Mailer: Mozilla 4.78 [en]C-CCK-MCD (Windows NT 5.0; U) > >X-Accept-Language: en > >References: <BAY9-F42Xng77T4kfjD0003cf53 at hotmail.com> > >Return-Path: bgunter at gene.com > >X-OriginalArrivalTime: 22 Jun 2004 20:45:15.0723 (UTC) > >FILETIME=[D27351B0:01C45899] > > > >devshruti pahuja wrote: > > > > > Hi All > > > > > > I received a raw data set with one record per tennis > player (both male > >and > > > female) and then i cured it by aggregation i.e by 4 age > groups, 2 gender > > > levels and 6 income levels. Gender and Income are > categorical variables. > > > Please advise me how to use 'R' to model this data set > (Actually, i want > >to > > > know the right regression technique and steps to do that, > including > >removing > > > outliers at the begining and looking for hidden interactions). > > > > > > >--- in which case, shouldn't we also sign and hand in the > homework for you? > > > >-- Bert > > > > ______________________________________________ > R-help at stat.math.ethz.ch mailing list > https://www.stat.math.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide! > http://www.R-project.org/posting-guide.html > >