Hello folks, Let me first apologize: I'm not a professional nor a mathematician, just an ordinary guy, fooling around with the excellent R-package. I know the basic principles behind statistics, but haven't read anything more advanced than the ordinary first probability and statistics courses. Enough disclaimers? Good! I was examining the multinom-function (in the nnet-package) the other day and run a couple of tests. This is part of an output I got. multinom(formula = result ~., data = info) Coefficients: (Intercept) PH WH 1 -0.3974387 0.02201908 -0.0009618038 2 2.5183566 -0.07076967 -0.0596189836 The variable result I was trying to predict is a nominal variable, taking the values 0, 1 and 2. The independent variables are quite a few, some of them nominal, some not. I know about regression coefficients, intercepts, and all that jazz. I'm not sure, though, how I should use the coefficients and intercepts I got in the output. Why are there two different coefficients per variable? When should I use which? How should I interpret the output of the regression function? Should the output be rounded to the nearest integer, which would happen to be <=0, 1 or >=2 -> I can decide which the prediction is? Or am I missing some fine points in this particular field of math I'm (admittedly somewhat blindfolded) messing around in? I'm grateful for any help, or any pointers to good sources on the web. I don't have any problems with reading theory - it's just that I don't have any theory to read! Thanks a lot in advance, Mike __________________________________________________ Check out Yahoo! Shopping and Yahoo! Auctions for all of your unique holiday gifts! Buy at http://shopping.yahoo.com or bid at http://auctions.yahoo.com -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
multinom is support software for a book, Moderan Applied Statistics with S-PLUS by Venables & Ripley. You have read it and understood its examples, haven't you? On Wed, 12 Dec 2001, Hoodoo Gooroo wrote:> Hello folks, > > Let me first apologize: I'm not a professional nor a > mathematician, just an ordinary guy, fooling around > with the excellent R-package. I know the basic > principles behind statistics, but haven't read > anything more advanced than the ordinary first > probability and statistics courses. > > Enough disclaimers? Good! I was examining the > multinom-function (in the nnet-package) the other day > and run a couple of tests. This is part of an output I > got. > > multinom(formula = result ~., data = info) > > Coefficients: > (Intercept) PH WH > 1 -0.3974387 0.02201908 -0.0009618038 > 2 2.5183566 -0.07076967 -0.0596189836 > > The variable result I was trying to predict is a > nominal variable, taking the values 0, 1 and 2. The > independent variables are quite a few, some of them > nominal, some not. I know about regression > coefficients, intercepts, and all that jazz. I'm not > sure, though, how I should use the coefficients and > intercepts I got in the output. Why are there two > different coefficients per variable? When should I use > which? How should I interpret the output of the > regression function? Should the output be rounded to > the nearest integer, which would happen to be <=0, 1 > or >=2 -> I can decide which the prediction is? Or am > I missing some fine points in this particular field of > math I'm (admittedly somewhat blindfolded) messing > around in? > > I'm grateful for any help, or any pointers to good > sources on the web. I don't have any problems with > reading theory - it's just that I don't have any > theory to read! > > Thanks a lot in advance, > > Mike > > __________________________________________________ > > Check out Yahoo! Shopping and Yahoo! Auctions for all of > your unique holiday gifts! Buy at http://shopping.yahoo.com > or bid at http://auctions.yahoo.com > -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- > r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html > Send "info", "help", or "[un]subscribe" > (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch > _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._ >-- Brian D. Ripley, ripley at stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272860 (secr) Oxford OX1 3TG, UK Fax: +44 1865 272595 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Dear Mike, I have some lecture notes on logistic regression, available at <http://socserv.socsci.mcmaster.ca/jfox/Courses/soc740/chap15-overheads.pdf>, which briefly describe the multinomial logit model; look in the section on modeling polytomous data. You'd probably do well to read a text that covers the subject; J. Long, Regression Models for Categorical and Limited Dependent Variables (Sage, 1997) is quite accessible. John At 02:58 AM 12/12/2001 -0800, Hoodoo Gooroo wrote:>Hello folks, > >Let me first apologize: I'm not a professional nor a >mathematician, just an ordinary guy, fooling around >with the excellent R-package. I know the basic >principles behind statistics, but haven't read >anything more advanced than the ordinary first >probability and statistics courses. > >Enough disclaimers? Good! I was examining the >multinom-function (in the nnet-package) the other day >and run a couple of tests. This is part of an output I >got. > >multinom(formula = result ~., data = info) > >Coefficients: > (Intercept) PH WH >1 -0.3974387 0.02201908 -0.0009618038 >2 2.5183566 -0.07076967 -0.0596189836 > >The variable result I was trying to predict is a >nominal variable, taking the values 0, 1 and 2. The >independent variables are quite a few, some of them >nominal, some not. I know about regression >coefficients, intercepts, and all that jazz. I'm not >sure, though, how I should use the coefficients and >intercepts I got in the output. Why are there two >different coefficients per variable? When should I use >which? How should I interpret the output of the >regression function? Should the output be rounded to >the nearest integer, which would happen to be <=0, 1 >or >=2 -> I can decide which the prediction is? Or am >I missing some fine points in this particular field of >math I'm (admittedly somewhat blindfolded) messing >around in? > >I'm grateful for any help, or any pointers to good >sources on the web. I don't have any problems with >reading theory - it's just that I don't have any >theory to read! > >Thanks a lot in advance, > > Mike----------------------------------------------------- John Fox Department of Sociology McMaster University Hamilton, Ontario, Canada L8S 4M4 email: jfox at mcmaster.ca phone: 905-525-9140x23604 web: www.socsci.mcmaster.ca/jfox ----------------------------------------------------- -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._