Re: [R] LSmeans and lsmeans
Thanks Bert. From: Bert Gunter [mailto:bgunter.4...@gmail.com] Sent: Tuesday, February 13, 2018 4:42 PM To: Pius Mwansa <pmwa...@shaw.ca> Cc: R-help <r-help@r-project.org> Subject: Re: [R] LSmeans and lsmeans A cursory reading indicates that they are identical; but others more knowledgeable than I need to confirm or deny this. -- Bert Bert Gunter "The trouble with having an open mind is that people keep coming along and sticking things into it." -- Opus (aka Berkeley Breathed in his "Bloom County" comic strip ) On Tue, Feb 13, 2018 at 3:38 PM, Pius Mwansa <pmwa...@shaw.ca <mailto:pmwa...@shaw.ca> > wrote: It is in the doBy package. Thanks From: Bert Gunter [mailto:bgunter.4...@gmail.com <mailto:bgunter.4...@gmail.com> ] Sent: Tuesday, February 13, 2018 4:32 PM To: Pius Mwansa <pmwa...@shaw.ca <mailto:pmwa...@shaw.ca> > Cc: R-help <r-help@r-project.org <mailto:r-help@r-project.org> > Subject: Re: [R] LSmeans and lsmeans Always cc the list unless there is good reason to keep your reply private. There is no LSmeans() function in the lsmeans package. Cheers, Bert Bert Gunter "The trouble with having an open mind is that people keep coming along and sticking things into it." -- Opus (aka Berkeley Breathed in his "Bloom County" comic strip ) On Tue, Feb 13, 2018 at 3:20 PM, Pius Mwansa <pmwa...@shaw.ca <mailto:pmwa...@shaw.ca> > wrote: They are in the lsmeans package. Pius From: Bert Gunter [mailto:bgunter.4...@gmail.com <mailto:bgunter.4...@gmail.com> ] Sent: Tuesday, February 13, 2018 4:16 PM To: Pius Mwansa <pmwa...@shaw.ca <mailto:pmwa...@shaw.ca> > Cc: R-help <r-help@r-project.org <mailto:r-help@r-project.org> > Subject: Re: [R] LSmeans and lsmeans In what packages? -- Bert Bert Gunter "The trouble with having an open mind is that people keep coming along and sticking things into it." -- Opus (aka Berkeley Breathed in his "Bloom County" comic strip ) On Tue, Feb 13, 2018 at 11:17 AM, Pius Mwansa <pmwa...@shaw.ca <mailto:pmwa...@shaw.ca> > wrote: Is there a difference between LSmeans and lsmeans functions in R? Thanks, Pius __ R-help@r-project.org <mailto:R-help@r-project.org> mailing list -- To UNSUBSCRIBE and more, see 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. [[alternative HTML version deleted]] __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see 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.
Re: [R] LSmeans and lsmeans
It is in the doBy package. Thanks From: Bert Gunter [mailto:bgunter.4...@gmail.com] Sent: Tuesday, February 13, 2018 4:32 PM To: Pius Mwansa <pmwa...@shaw.ca> Cc: R-help <r-help@r-project.org> Subject: Re: [R] LSmeans and lsmeans Always cc the list unless there is good reason to keep your reply private. There is no LSmeans() function in the lsmeans package. Cheers, Bert Bert Gunter "The trouble with having an open mind is that people keep coming along and sticking things into it." -- Opus (aka Berkeley Breathed in his "Bloom County" comic strip ) On Tue, Feb 13, 2018 at 3:20 PM, Pius Mwansa <pmwa...@shaw.ca <mailto:pmwa...@shaw.ca> > wrote: They are in the lsmeans package. Pius From: Bert Gunter [mailto:bgunter.4...@gmail.com <mailto:bgunter.4...@gmail.com> ] Sent: Tuesday, February 13, 2018 4:16 PM To: Pius Mwansa <pmwa...@shaw.ca <mailto:pmwa...@shaw.ca> > Cc: R-help <r-help@r-project.org <mailto:r-help@r-project.org> > Subject: Re: [R] LSmeans and lsmeans In what packages? -- Bert Bert Gunter "The trouble with having an open mind is that people keep coming along and sticking things into it." -- Opus (aka Berkeley Breathed in his "Bloom County" comic strip ) On Tue, Feb 13, 2018 at 11:17 AM, Pius Mwansa <pmwa...@shaw.ca <mailto:pmwa...@shaw.ca> > wrote: Is there a difference between LSmeans and lsmeans functions in R? Thanks, Pius __ R-help@r-project.org <mailto:R-help@r-project.org> mailing list -- To UNSUBSCRIBE and more, see 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. [[alternative HTML version deleted]] __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see 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.
Re: [R] LSmeans and lsmeans
A cursory reading indicates that they are identical; but others more knowledgeable than I need to confirm or deny this. -- Bert Bert Gunter "The trouble with having an open mind is that people keep coming along and sticking things into it." -- Opus (aka Berkeley Breathed in his "Bloom County" comic strip ) On Tue, Feb 13, 2018 at 3:38 PM, Pius Mwansa <pmwa...@shaw.ca> wrote: > It is in the doBy package. > > > > Thanks > > > > *From:* Bert Gunter [mailto:bgunter.4...@gmail.com] > *Sent:* Tuesday, February 13, 2018 4:32 PM > > *To:* Pius Mwansa <pmwa...@shaw.ca> > *Cc:* R-help <r-help@r-project.org> > *Subject:* Re: [R] LSmeans and lsmeans > > > > Always cc the list unless there is good reason to keep your reply private. > > There is no LSmeans() function in the lsmeans package. > > Cheers, > > Bert > > > Bert Gunter > > "The trouble with having an open mind is that people keep coming along and > sticking things into it." > -- Opus (aka Berkeley Breathed in his "Bloom County" comic strip ) > > > > On Tue, Feb 13, 2018 at 3:20 PM, Pius Mwansa <pmwa...@shaw.ca> wrote: > > They are in the lsmeans package. > > > > Pius > > > > *From:* Bert Gunter [mailto:bgunter.4...@gmail.com] > *Sent:* Tuesday, February 13, 2018 4:16 PM > *To:* Pius Mwansa <pmwa...@shaw.ca> > *Cc:* R-help <r-help@r-project.org> > *Subject:* Re: [R] LSmeans and lsmeans > > > > In what packages? > > -- Bert > > > Bert Gunter > > "The trouble with having an open mind is that people keep coming along and > sticking things into it." > -- Opus (aka Berkeley Breathed in his "Bloom County" comic strip ) > > > > On Tue, Feb 13, 2018 at 11:17 AM, Pius Mwansa <pmwa...@shaw.ca> wrote: > > Is there a difference between LSmeans and lsmeans functions in R? > > Thanks, > > Pius > > __ > R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see > 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. > > > > > [[alternative HTML version deleted]] __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see 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.
Re: [R] LSmeans and lsmeans
Always cc the list unless there is good reason to keep your reply private. There is no LSmeans() function in the lsmeans package. Cheers, Bert Bert Gunter "The trouble with having an open mind is that people keep coming along and sticking things into it." -- Opus (aka Berkeley Breathed in his "Bloom County" comic strip ) On Tue, Feb 13, 2018 at 3:20 PM, Pius Mwansa <pmwa...@shaw.ca> wrote: > They are in the lsmeans package. > > > > Pius > > > > *From:* Bert Gunter [mailto:bgunter.4...@gmail.com] > *Sent:* Tuesday, February 13, 2018 4:16 PM > *To:* Pius Mwansa <pmwa...@shaw.ca> > *Cc:* R-help <r-help@r-project.org> > *Subject:* Re: [R] LSmeans and lsmeans > > > > In what packages? > > -- Bert > > > Bert Gunter > > "The trouble with having an open mind is that people keep coming along and > sticking things into it." > -- Opus (aka Berkeley Breathed in his "Bloom County" comic strip ) > > > > On Tue, Feb 13, 2018 at 11:17 AM, Pius Mwansa <pmwa...@shaw.ca> wrote: > > Is there a difference between LSmeans and lsmeans functions in R? > > Thanks, > > Pius > > __ > R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see > 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. > > > [[alternative HTML version deleted]] __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see 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.
Re: [R] LSmeans and lsmeans
In what packages? -- Bert Bert Gunter "The trouble with having an open mind is that people keep coming along and sticking things into it." -- Opus (aka Berkeley Breathed in his "Bloom County" comic strip ) On Tue, Feb 13, 2018 at 11:17 AM, Pius Mwansawrote: > Is there a difference between LSmeans and lsmeans functions in R? > > Thanks, > > Pius > > __ > R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see > 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. > [[alternative HTML version deleted]] __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see 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.
[R] LSmeans and lsmeans
Is there a difference between LSmeans and lsmeans functions in R? Thanks, Pius __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see 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.
[R] ..lsmeans and coxme..
Dear list-eners, I run into the following problem when I want to get contrasts from a coxme model using the lsmeans package: A call to lsmeans on the coxme model throws the following error: Error in if (adjustSigma object$method == ML) V = V * object$dims$N/(object$dims$N - : missing value where TRUE/FALSE needed I give an example: library(coxme) library(lsmeans) fm - coxme(Surv(y, uncens) ~ trt + (trt | center) + strata(center), data=eortc) summary(fm) lsmeans(fm, ~ trt) traceback points to the internal function lsmeans:::lsm.basis.lme() Any ideas? Thanks in advance. -m [[alternative HTML version deleted]] __ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see 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.
[R] lsmeans in R
I need help with calculating lsmeans (adjusted means) of different terms in a linear model including the main effect and the interaction effect terms. I use lm to run the linear models...I previously noted from literature that that effects package can be used to generate lsmeans. But I tried to use it but could not figure out which option to use to get means. If anyone can give an example of how to get lsmeans using lm object, that will very helpful. Thanks, SUman [[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.
Re: [R] lsmeans in R
On 3/11/2009 2:45 AM, suman Duvvuru wrote: I need help with calculating lsmeans (adjusted means) of different terms in a linear model including the main effect and the interaction effect terms. I use lm to run the linear models...I previously noted from literature that that effects package can be used to generate lsmeans. But I tried to use it but could not figure out which option to use to get means. If anyone can give an example of how to get lsmeans using lm object, that will very helpful. This R-help thread from March 2007 should help: http://finzi.psych.upenn.edu/R/Rhelp02/archive/95809.html Thanks, SUman [[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. -- Chuck Cleland, Ph.D. NDRI, Inc. (www.ndri.org) 71 West 23rd Street, 8th floor New York, NY 10010 tel: (212) 845-4495 (Tu, Th) tel: (732) 512-0171 (M, W, F) fax: (917) 438-0894 __ 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.
Re: [R] lsmeans in R
Dear Suman, Chuck Cleland has already pointed you toward a reasonably complete discussion of the topic (thank you Chuck). To update my contribution to that discussion, the effects package now uses t-intervals for models with an estimated dispersion parameter (such as linear models) and will create displays for multinomial and proportional-odds logit models. (The latter isn't relevant to your intended application, of course.) Beyond that, I don't quite understand your question. Adjusted means are fitted values, and this is what the effects package gives you for linear models. Maybe you could explain in more detail what it is that you want and why you think that you're not getting it. Regards, John -Original Message- From: r-help-boun...@r-project.org [mailto:r-help-boun...@r-project.org] On Behalf Of suman Duvvuru Sent: March-11-09 2:45 AM To: r-help@r-project.org Subject: [R] lsmeans in R I need help with calculating lsmeans (adjusted means) of different terms in a linear model including the main effect and the interaction effect terms. I use lm to run the linear models...I previously noted from literature that that effects package can be used to generate lsmeans. But I tried to use it but could not figure out which option to use to get means. If anyone can give an example of how to get lsmeans using lm object, that will very helpful. Thanks, SUman [[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. __ 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.
Re: [R] lsmeans
I hope you'll forgive me for resurrecting this thread. My question refers to John Fox's comments in the discussion of lsmeans from https://stat.ethz.ch/pipermail/r-help/2008-June/164106.html John you said, It wouldn't be hard, however, to do the computations yourself, using the coefficient vector for the fixed effects and a suitably constructed model-matrix to compute the effects; you could also get standard errors by using the covariance matrix for the fixed effects. I've been able to make use of all of that except for the 'suitably constructed model-matrix' part. I've looked through some other threads on this topic, but am still a little in the dark as to what I'd need to do to construct a suitable matrix. I would like to use the least squares means to develop parameter estimates for a parametric ROC analysis, as described by Mithat Gonen's book (Analyzing Receiver Operating Characteristic Curves with SAS, 2007). Any suggestions on references that would explain how to go about constructing the suitable model matrix? Many Thanks Benjamin P Please consider the environment before printing this e-mail Cleveland Clinic is ranked one of the top hospitals in America by U.S. News World Report (2008). Visit us online at http://www.clevelandclinic.org for a complete listing of our services, staff and locations. Confidentiality Note: This message is intended for use\...{{dropped:13}} __ 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.
Re: [R] lsmeans
Dear Benjamin, In the absence of interactions, a suitably constructed model matrix could, for example, allow one predictor to range over its values while others are held to typical values (such as means). The effects package does this for linear and generalized linear models (and soon for proportional-odds and multinomial-logit models), and produces reasonable displays when there are interactions and other complex terms (such as regression splines or polynomials) in a model, but it doesn't have methods for mixed-effects models or survival-regression models. I hope this helps, John -- John Fox, Professor Department of Sociology McMaster University Hamilton, Ontario, Canada web: socserv.mcmaster.ca/jfox -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Nutter, Benjamin Sent: September-26-08 3:48 PM To: r-help@r-project.org Subject: Re: [R] lsmeans I hope you'll forgive me for resurrecting this thread. My question refers to John Fox's comments in the discussion of lsmeans from https://stat.ethz.ch/pipermail/r-help/2008-June/164106.html John you said, It wouldn't be hard, however, to do the computations yourself, using the coefficient vector for the fixed effects and a suitably constructed model-matrix to compute the effects; you could also get standard errors by using the covariance matrix for the fixed effects. I've been able to make use of all of that except for the 'suitably constructed model-matrix' part. I've looked through some other threads on this topic, but am still a little in the dark as to what I'd need to do to construct a suitable matrix. I would like to use the least squares means to develop parameter estimates for a parametric ROC analysis, as described by Mithat Gonen's book (Analyzing Receiver Operating Characteristic Curves with SAS, 2007). Any suggestions on references that would explain how to go about constructing the suitable model matrix? Many Thanks Benjamin P Please consider the environment before printing this e-mail Cleveland Clinic is ranked one of the top hospitals in America by U.S. News World Report (2008). Visit us online at http://www.clevelandclinic.org for a complete listing of our services, staff and locations. Confidentiality Note: This message is intended for use\...{{dropped:13}} __ 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. __ 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.
Re: [R] lsmeans
John Fox jfox at mcmaster.ca writes: Actually, the effects package does exactly what you suggest for continuous predictors. But not for lme. Dieter __ 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.
Re: [R] lsmeans
On 6/7/08, John Fox [EMAIL PROTECTED] wrote: Dear Dieter, I don't know whether I qualify as a master, but here's my brief take on the subject: First, I dislike the term least-squares means, which seems to me like nonsense. Second, what I prefer to call effect displays are just judiciously chosen regions of the response surface of a model, meant to clarify effects in complex models. For example, a two-way interaction is displayed by absorbing the constant and main-effect terms in the interaction (more generally, absorbing terms marginal to a particular term) and setting other terms to typical values. A table or graph of the resulting fitted values is, I would argue, easier to grasp than the coefficients, the interpretation of which can entail complicated mental arithmetic. I like that explanation, John. As I'm sure you are aware, the key phrase in what you wrote is setting other terms to typical values. That is, these are conditional cell means, yet they are almost universally misunderstood - even by statisticians who should know better - to be marginal cell means. A more subtle aspect of that phrase is the interpretation of typical. The user is not required to specify these typical values - they are calculated from the observed data. If there are no interactions with the other terms and if the values chosen for those other terms based on the observed data are indeed typical of the values for which we wish to make inferences with the model then these conditional cell means may tell us something about the marginal cell means. But if either of those conditions fails then these conditional means can be very different from the marginal means. I wouldn't have any problem at all with providing conditional cell means, especially if the user were required to specify the values at which to fix the other terms in the model, but that is not what people think they are getting. I don't want to encourage them in their delusions by letting them think i can evaluate marginal cell means as a single, conditional evaluation. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Dieter Menne Sent: June-07-08 4:36 AM To: [EMAIL PROTECTED] Subject: Re: [R] lsmeans John Fox jfox at mcmaster.ca writes: I intend at some point to extend the effects package to linear and generalized linear mixed-effects models, probably using lmer() rather than lme(), but as you discovered, it doesn't handle these models now. It wouldn't be hard, however, to do the computations yourself, using the coefficient vector for the fixed effects and a suitably constructed model-matrix to compute the effects; you could also get standard errors by using the covariance matrix for the fixed effects. Douglas Bates: https://stat.ethz.ch/pipermail/r-sig-mixed-models/2007q2/000222.html My big problem with lsmeans is that I have never been able to understand how they should be calculated and, more importantly, why one should want to calculate them. In other words, what do lsmeans represent and why should I be interested in these particular values? Truly Confused, torn apart by the Masters Dieter __ 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. __ 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. __ 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.
Re: [R] lsmeans
On Sun, Jun 8, 2008 at 12:58 PM, Douglas Bates [EMAIL PROTECTED] wrote: On 6/7/08, John Fox [EMAIL PROTECTED] wrote: Dear Dieter, I don't know whether I qualify as a master, but here's my brief take on the subject: First, I dislike the term least-squares means, which seems to me like nonsense. Second, what I prefer to call effect displays are just judiciously chosen regions of the response surface of a model, meant to clarify effects in complex models. For example, a two-way interaction is displayed by absorbing the constant and main-effect terms in the interaction (more generally, absorbing terms marginal to a particular term) and setting other terms to typical values. A table or graph of the resulting fitted values is, I would argue, easier to grasp than the coefficients, the interpretation of which can entail complicated mental arithmetic. I like that explanation, John. As I'm sure you are aware, the key phrase in what you wrote is setting other terms to typical values. That is, these are conditional cell means, yet they are almost universally misunderstood - even by statisticians who should know better - to be marginal cell means. A more subtle aspect of that phrase is the interpretation of typical. The user is not required to specify these typical values - they are calculated from the observed data. How does Searle's population marginal means fit in to this? The paper describes a PMM as expected value of an observed marginal mean as if there were one observation in every cell. - which was what I thought happened in the effects display. Is this a subtly on the definition of typical, or is that PMM's are only described for pure ANOVA's (i.e. no continuous variables in model)? Hadley -- http://had.co.nz/ __ 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.
Re: [R] lsmeans
Dear Doug, Your point is correct, of course, but if people are interested in computing marginal means (or marginal cell means), then they can do so simply and don't need a statistical model. I think that when such a model is fit, interest is typically in conditioning on the other explanatory variables. (Also see my responses to Hadley and Frank's points.) Regards, John -- John Fox, Professor Department of Sociology McMaster University Hamilton, Ontario, Canada web: socserv.mcmaster.ca/jfox -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Douglas Bates Sent: June-08-08 1:58 PM To: John Fox Cc: Dieter Menne; [EMAIL PROTECTED] Subject: Re: [R] lsmeans On 6/7/08, John Fox [EMAIL PROTECTED] wrote: Dear Dieter, I don't know whether I qualify as a master, but here's my brief take on the subject: First, I dislike the term least-squares means, which seems to me like nonsense. Second, what I prefer to call effect displays are just judiciously chosen regions of the response surface of a model, meant to clarify effects in complex models. For example, a two-way interaction is displayed by absorbing the constant and main-effect terms in the interaction (more generally, absorbing terms marginal to a particular term) and setting other terms to typical values. A table or graph of the resulting fitted values is, I would argue, easier to grasp than the coefficients, the interpretation of which can entail complicated mental arithmetic. I like that explanation, John. As I'm sure you are aware, the key phrase in what you wrote is setting other terms to typical values. That is, these are conditional cell means, yet they are almost universally misunderstood - even by statisticians who should know better - to be marginal cell means. A more subtle aspect of that phrase is the interpretation of typical. The user is not required to specify these typical values - they are calculated from the observed data. If there are no interactions with the other terms and if the values chosen for those other terms based on the observed data are indeed typical of the values for which we wish to make inferences with the model then these conditional cell means may tell us something about the marginal cell means. But if either of those conditions fails then these conditional means can be very different from the marginal means. I wouldn't have any problem at all with providing conditional cell means, especially if the user were required to specify the values at which to fix the other terms in the model, but that is not what people think they are getting. I don't want to encourage them in their delusions by letting them think i can evaluate marginal cell means as a single, conditional evaluation. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Dieter Menne Sent: June-07-08 4:36 AM To: [EMAIL PROTECTED] Subject: Re: [R] lsmeans John Fox jfox at mcmaster.ca writes: I intend at some point to extend the effects package to linear and generalized linear mixed-effects models, probably using lmer() rather than lme(), but as you discovered, it doesn't handle these models now. It wouldn't be hard, however, to do the computations yourself, using the coefficient vector for the fixed effects and a suitably constructed model-matrix to compute the effects; you could also get standard errors by using the covariance matrix for the fixed effects. Douglas Bates: https://stat.ethz.ch/pipermail/r-sig-mixed-models/2007q2/000222.html My big problem with lsmeans is that I have never been able to understand how they should be calculated and, more importantly, why one should want to calculate them. In other words, what do lsmeans represent and why should I be interested in these particular values? Truly Confused, torn apart by the Masters Dieter __ 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. __ 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. __ 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
Re: [R] lsmeans
Dear Hadley, Unfortunately, the term marginal gets used in two quite different ways, and Searle's population marginal means would, I believe, be more clearly called population conditional means or population partial means. This is more or less alternative terminology for least-squares means (to which Searle rightly objects). Regards, John -- John Fox, Professor Department of Sociology McMaster University Hamilton, Ontario, Canada web: socserv.mcmaster.ca/jfox -Original Message- From: hadley wickham [mailto:[EMAIL PROTECTED] Sent: June-08-08 2:52 PM To: Douglas Bates Cc: John Fox; Dieter Menne; [EMAIL PROTECTED] Subject: Re: [R] lsmeans On Sun, Jun 8, 2008 at 12:58 PM, Douglas Bates [EMAIL PROTECTED] wrote: On 6/7/08, John Fox [EMAIL PROTECTED] wrote: Dear Dieter, I don't know whether I qualify as a master, but here's my brief take on the subject: First, I dislike the term least-squares means, which seems to me like nonsense. Second, what I prefer to call effect displays are just judiciously chosen regions of the response surface of a model, meant to clarify effects in complex models. For example, a two-way interaction is displayed by absorbing the constant and main-effect terms in the interaction (more generally, absorbing terms marginal to a particular term) and setting other terms to typical values. A table or graph of the resulting fitted values is, I would argue, easier to grasp than the coefficients, the interpretation of which can entail complicated mental arithmetic. I like that explanation, John. As I'm sure you are aware, the key phrase in what you wrote is setting other terms to typical values. That is, these are conditional cell means, yet they are almost universally misunderstood - even by statisticians who should know better - to be marginal cell means. A more subtle aspect of that phrase is the interpretation of typical. The user is not required to specify these typical values - they are calculated from the observed data. How does Searle's population marginal means fit in to this? The paper describes a PMM as expected value of an observed marginal mean as if there were one observation in every cell. - which was what I thought happened in the effects display. Is this a subtly on the definition of typical, or is that PMM's are only described for pure ANOVA's (i.e. no continuous variables in model)? Hadley -- http://had.co.nz/ __ 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.
Re: [R] lsmeans
Well put Doug. I would add another condition, which I don't know how to state precisely. The settings for the other terms, which are usually marginal medians, modes, or means, must make sense when considered jointly. Frequently when all adjustment covariates are set to overall marginal means the resulting subject is very atypical. To me much of the problem is solved one one develops a liking for predicted values and differences in them. Maybe I'm still misunderstanding, but isn't that exactly what effects displays are? They're just some way to allow you to say, I'm interested in variables x, y and z, and I don't really care about the other variables in the model - what are some typical predictions? The effects package implements this idea for categorical x, y, and z, but the basic idea remains the same for continuous variables - except instead of using all the levels of the factor, you'd use a grid within the range of the data. Hadley -- http://had.co.nz/ __ 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.
Re: [R] lsmeans
Dear Hadley, Actually, the effects package does exactly what you suggest for continuous predictors. Regards, John -- John Fox, Professor Department of Sociology McMaster University Hamilton, Ontario, Canada web: socserv.mcmaster.ca/jfox -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of hadley wickham Sent: June-08-08 3:48 PM To: Frank E Harrell Jr Cc: John Fox; Douglas Bates; [EMAIL PROTECTED]; Dieter Menne Subject: Re: [R] lsmeans Well put Doug. I would add another condition, which I don't know how to state precisely. The settings for the other terms, which are usually marginal medians, modes, or means, must make sense when considered jointly. Frequently when all adjustment covariates are set to overall marginal means the resulting subject is very atypical. To me much of the problem is solved one one develops a liking for predicted values and differences in them. Maybe I'm still misunderstanding, but isn't that exactly what effects displays are? They're just some way to allow you to say, I'm interested in variables x, y and z, and I don't really care about the other variables in the model - what are some typical predictions? The effects package implements this idea for categorical x, y, and z, but the basic idea remains the same for continuous variables - except instead of using all the levels of the factor, you'd use a grid within the range of the data. Hadley -- http://had.co.nz/ __ 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. __ 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.
Re: [R] lsmeans
John Fox jfox at mcmaster.ca writes: I intend at some point to extend the effects package to linear and generalized linear mixed-effects models, probably using lmer() rather than lme(), but as you discovered, it doesn't handle these models now. It wouldn't be hard, however, to do the computations yourself, using the coefficient vector for the fixed effects and a suitably constructed model-matrix to compute the effects; you could also get standard errors by using the covariance matrix for the fixed effects. Douglas Bates: https://stat.ethz.ch/pipermail/r-sig-mixed-models/2007q2/000222.html My big problem with lsmeans is that I have never been able to understand how they should be calculated and, more importantly, why one should want to calculate them. In other words, what do lsmeans represent and why should I be interested in these particular values? Truly Confused, torn apart by the Masters Dieter __ 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.
Re: [R] lsmeans
Dear Dieter, I don't know whether I qualify as a master, but here's my brief take on the subject: First, I dislike the term least-squares means, which seems to me like nonsense. Second, what I prefer to call effect displays are just judiciously chosen regions of the response surface of a model, meant to clarify effects in complex models. For example, a two-way interaction is displayed by absorbing the constant and main-effect terms in the interaction (more generally, absorbing terms marginal to a particular term) and setting other terms to typical values. A table or graph of the resulting fitted values is, I would argue, easier to grasp than the coefficients, the interpretation of which can entail complicated mental arithmetic. Regards, John -- John Fox, Professor Department of Sociology McMaster University Hamilton, Ontario, Canada web: socserv.mcmaster.ca/jfox -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Dieter Menne Sent: June-07-08 4:36 AM To: [EMAIL PROTECTED] Subject: Re: [R] lsmeans John Fox jfox at mcmaster.ca writes: I intend at some point to extend the effects package to linear and generalized linear mixed-effects models, probably using lmer() rather than lme(), but as you discovered, it doesn't handle these models now. It wouldn't be hard, however, to do the computations yourself, using the coefficient vector for the fixed effects and a suitably constructed model-matrix to compute the effects; you could also get standard errors by using the covariance matrix for the fixed effects. Douglas Bates: https://stat.ethz.ch/pipermail/r-sig-mixed-models/2007q2/000222.html My big problem with lsmeans is that I have never been able to understand how they should be calculated and, more importantly, why one should want to calculate them. In other words, what do lsmeans represent and why should I be interested in these particular values? Truly Confused, torn apart by the Masters Dieter __ 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. __ 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.
Re: [R] lsmeans
On Sat, Jun 7, 2008 at 3:02 PM, John Fox [EMAIL PROTECTED] wrote: Dear Dieter, I don't know whether I qualify as a master, but here's my brief take on the subject: First, I dislike the term least-squares means, which seems to me like nonsense. Second, what I prefer to call effect displays are just judiciously chosen regions of the response surface of a model, meant to clarify effects in complex models. For example, a two-way interaction is displayed by absorbing the constant and main-effect terms in the interaction (more generally, absorbing terms marginal to a particular term) and setting other terms to typical values. A table or graph of the resulting fitted values is, I would argue, easier to grasp than the coefficients, the interpretation of which can entail complicated mental arithmetic. The other advantage is that the effects values are on the same scale as the original data, and it can be useful to supplement the pure effects with the original data. Hadley -- http://had.co.nz/ __ 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.
[R] lsmeans
Hello, I have the next function call: lme(fixed=Error ~ Temperature * Tumour ,random = ~1|ID, data=error_DB) which returns an lme object. I am interested on carrying out some kind of lsmeans on the data returned, but I cannot find any function to do this in R. I'have seen the effect() function, but it does not work with lme objects. Any idea? Best, Dani -- Daniel Valverde Saubí Grup de Biologia Molecular de Llevats Facultat de Veterinària de la Universitat Autònoma de Barcelona Edifici V, Campus UAB 08193 Cerdanyola del Vallès- SPAIN Centro de Investigación Biomédica en Red en Bioingeniería, Biomateriales y Nanomedicina (CIBER-BBN) Grup d'Aplicacions Biomèdiques de la RMN Facultat de Biociències Universitat Autònoma de Barcelona Edifici Cs, Campus UAB 08193 Cerdanyola del Vallès- SPAIN +34 93 5814126 __ 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.
Re: [R] lsmeans
Dear Dani, I intend at some point to extend the effects package to linear and generalized linear mixed-effects models, probably using lmer() rather than lme(), but as you discovered, it doesn't handle these models now. It wouldn't be hard, however, to do the computations yourself, using the coefficient vector for the fixed effects and a suitably constructed model-matrix to compute the effects; you could also get standard errors by using the covariance matrix for the fixed effects. I hope this helps, John On Fri, 06 Jun 2008 17:05:58 +0200 Dani Valverde [EMAIL PROTECTED] wrote: Hello, I have the next function call: lme(fixed=Error ~ Temperature * Tumour ,random = ~1|ID, data=error_DB) which returns an lme object. I am interested on carrying out some kind of lsmeans on the data returned, but I cannot find any function to do this in R. I'have seen the effect() function, but it does not work with lme objects. Any idea? Best, Dani -- Daniel Valverde Saubí Grup de Biologia Molecular de Llevats Facultat de Veterinària de la Universitat Autònoma de Barcelona Edifici V, Campus UAB 08193 Cerdanyola del Vallès- SPAIN Centro de Investigación Biomédica en Red en Bioingeniería, Biomateriales y Nanomedicina (CIBER-BBN) Grup d'Aplicacions Biomèdiques de la RMN Facultat de Biociències Universitat Autònoma de Barcelona Edifici Cs, Campus UAB 08193 Cerdanyola del Vallès- SPAIN +34 93 5814126 __ 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. John Fox, Professor Department of Sociology McMaster University Hamilton, Ontario, Canada http://socserv.mcmaster.ca/jfox/ __ 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.