[R] how to categorize continuous variable when useing regression
assume dependent variable y( continuous),independent variable x ( continuous),I try to categorize x with some interval,such that,those intervals would has most significant different effect on y. any one knows which method I should apply,I know it will cause the loss of information,but can I really do that?or by using what mehod ,I will keep the loss minimal,all I want just some key words,thanks in advance~ [[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] how to categorize continuous variable when useing regression
assume dependent variable y( continuous),independent variable x ( continuous),I try to categorize x with some interval,such that,those intervals would has most significant different effect on y. any one knows which method I should apply,I really need some hints,thanks so much~ [[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] who happenly read these two paper Mohsen Pourahmadi (biometrika1999, 2000)
http://biomet.oxfordjournals.org/cgi/reprint/86/3/677 biometrika1999 http://biomet.oxfordjournals.org/cgi/reprint/94/4/1006 biometrika2000 Hi All: I just want to try some luck. I am currenly working on my project,one part of my project is to reanalysis the kenward cattle data by using the method in Mohsen's paper,but I found I really can get the same or close output as he did,so,any one who have happenly read his paper,got any idea how he got the LS estimates of gamma. he assumed the autoregressive phi_tj satisfy the cubic function on lag of time,where t is from 1 to 11,j from 1 to t-1,his model is: phi_tj=gamma1+gamma2*(t-j)+gamma3*(t-j)^2+gamma4*(t-j)^3, #biometrika 1999,page685 at first moment ,I thougt my design matrix should be 1 1 1 1 1 2 4 8 1 3 9 27. but I found this is wrong,actually I should(I think) use a polynomial design matirx with level of 10,degree=3 here is my rocode~~ y1=c(1,0.9,0.98,1.06,0.83,1.00,0.41,0.93,1.01,0.86) y2=c(0.05,0.16,0,0.26,0.15,0.61,0.33,0.31,0.33) y3=c(-0.23,0,0.16,-0.03,0.22,-0.03,-0.17,-0.05) y4=c(0.04,-0.21,-0.04,-0.26,-0.03,-0.04,-0.05) y5=c(-0.02,-0.34,0.06,-0.22,-0.11,-0.31) y6=c(0.20,0.01,0.01,-0.26,0.01) y7=c(-0.06,-0.14,0.39,0.23) y8=c(0.21,0.10,0.09) y9=c(-0.24,-0.23) y10=c(0.13) y=c(y1,y2,y3,y4,y5,y6,y7,y8,y9,y10) ## autoregressive paramters ,table 1 biometrika 1999,page 685 om2=matrix(0,nrow=55,ncol=3) om1=poly(c(1:10),degree=3) ##polynomial design matirx with level =11,cubic om2[1:10,]=t(matrix(rep(om1[1,],10),ncol=10)) om2[11:19,]=t(matrix(rep(om1[2,],9),ncol=9)) om2[20:27,]=t(matrix(rep(om1[3,],8),ncol=8)) om2[28:34,]=t(matrix(rep(om1[4,],7),ncol=7)) om2[35:40,]=t(matrix(rep(om1[5,],6),ncol=6)) om2[41:45,]=t(matrix(rep(om1[6,],5),ncol=5)) om2[46:49,]=t(matrix(rep(om1[7,],4),ncol=4)) om2[50:52,]=t(matrix(rep(om1[8,],3),ncol=3)) om2[53:54,]=t(matrix(rep(om1[9,],2),ncol=2)) om2[55,]=t(matrix(rep(om1[10,],1),ncol=1)) ## so om2 is my orthogonal design matirx ## for example ,for y1=c(1,0.9,0.98,1.06,0.83,1.00,0.41,0.93,1.01,0.86), since t-j=1,so the design matrix should be the first row of om1,and repeat for 10 times. t1=om2[,1] t2=om2[,2] t3=om2[,3] fit-lm(y~(t1+t2+t3)) summary(fit) ### the estimate I got is Estimate Std. Error t value Pr(|t|) (Intercept) 0.091860.02980 3.083 0.0033 ** t1 -0.541430.10970 -4.936 8.94e-06 *** t2 0.489550.10417 4.700 2.01e-05 *** t3 0.560250.08771 6.387 5.05e-08 ** which is total different from the author's(0.18,-1.71,1.64,-1.11) I have been tried contact the author,but he only point out that the I might need to pay close attention to design matrix,but did not say how,I really run out of the idea how to construct another suitable design matrix ,since basing the paper,this should be the correct design matrix,so ,any one has any idea??? GREAT THANKS~~ [[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] who happenly read these two paper Mohsen Pourahmadi (biometrika1999, 2000)
http://biomet.oxfordjournals.org/cgi/reprint/86/3/677 biometrika1999 http://biomet.oxfordjournals.org/cgi/reprint/94/4/1006 biometrika2000 Hi All: I just want to try some luck. I am currenly working on my project,one part of my project is to reanalysis the kenward cattle data by using the method in Mohsen's paper,but I found I really can get the same or close output as he did,so,any one who have happenly read his paper,got any idea how he got the LS estimates of gamma. he assumed the autoregressive phi_tj satisfy the cubic function on lag of time,where t is from 1 to 11,j from 1 to t-1,his model is: phi_tj=gamma1+gamma2*(t-j)+gamma3*(t-j)^2+gamma4*(t-j)^3, #biometrika 1999,page685 at first moment ,I thougt my design matrix should be 1 1 1 1 1 2 4 8 1 3 9 27. but I found this is wrong,actually I should(I think) use a polynomial design matirx with level of 11,degree=3 here is my rocode~~ y1=c(1,0.9,0.98,1.06,0.83,1.00,0.41,0.93,1.01,0.86) y2=c(0.05,0.16,0,0.26,0.15,0.61,0.33,0.31,0.33) y3=c(-0.23,0,0.16,-0.03,0.22,-0.03,-0.17,-0.05) y4=c(0.04,-0.21,-0.04,-0.26,-0.03,-0.04,-0.05) y5=c(-0.02,-0.34,0.06,-0.22,-0.11,-0.31) y6=c(0.20,0.01,0.01,-0.26,0.01) y7=c(-0.06,-0.14,0.39,0.23) y8=c(0.21,0.10,0.09) y9=c(-0.24,-0.23) y10=c(0.13) y=c(y1,y2,y3,y4,y5,y6,y7,y8,y9,y10) ## autoregressive paramters ,table 1 biometrika 1999,page 685 om2=matrix(0,nrow=55,ncol=3) om1=poly(c(1:10),degree=3) ##polynomial design matirx with level =11,cubic om2[1:10,]=t(matrix(rep(om1[1,],10),ncol=10)) om2[11:19,]=t(matrix(rep(om1[2,],9),ncol=9)) om2[20:27,]=t(matrix(rep(om1[3,],8),ncol=8)) om2[28:34,]=t(matrix(rep(om1[4,],7),ncol=7)) om2[35:40,]=t(matrix(rep(om1[5,],6),ncol=6)) om2[41:45,]=t(matrix(rep(om1[6,],5),ncol=5)) om2[46:49,]=t(matrix(rep(om1[7,],4),ncol=4)) om2[50:52,]=t(matrix(rep(om1[8,],3),ncol=3)) om2[53:54,]=t(matrix(rep(om1[9,],2),ncol=2)) om2[55,]=t(matrix(rep(om1[10,],1),ncol=1)) ## so om2 is my orthogonal design matirx ## for example ,for y1=c(1,0.9,0.98,1.06,0.83,1.00,0.41,0.93,1.01,0.86), since t-j=1,so the design matrix should be the first row of om1,and repeat for 10 times. t1=om2[,1] t2=om2[,2] t3=om2[,3] fit-lm(y~(t1+t2+t3)) summary(fit) ### the estimate I got is Estimate Std. Error t value Pr(|t|) (Intercept) 0.091860.02980 3.083 0.0033 ** t1 -0.541430.10970 -4.936 8.94e-06 *** t2 0.489550.10417 4.700 2.01e-05 *** t3 0.560250.08771 6.387 5.05e-08 ** which is total different from the author's(0.18,-1.71,1.64,-1.11) I have been tried contact the author,but he only point out that the I might need to pay close attention to design matrix,but did not say how,I really run out of the idea how to construct another suitable design matrix ,since basing the paper,this should be the correct design matrix,so ,any one has any idea??? GREAT THANKS~~ [[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] Cholesky Decomposition in R
Hi everyone: I try to use r to do the Cholesky Decomposition,which is A=LDL',so far I only found how to decomposite A in to LL' by using chol(A),the function Cholesky(A) doesnt work,any one know other command to decomposte A in to LDL' My r code is: library(Matrix) A=matrix(c(1,1,1,1,5,5,1,5,14),nrow=3) chol(A) [,1] [,2] [,3] [1,]111 [2,]022 [3,]003 Cholesky(A) Error in function (classes, fdef, mtable) : unable to find an inherited method for function Cholesky, for signature matrix whatz wrong??? thanks~ [[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] anyone can help me with Cholesky Decomposition
Hi: what I want to do is decompose the a symmetric matrix A into this form A=LDL' hence TAT'=D,T is inverse of (L)and T is a lower trangular matrix,and D is dignoal matrix for one case A=1 1 1 1 5 5 1 5 14 T=inverse(L)= 1 0 0 -1 1 0 0 -1 1 D=(1,4,9) I tried to use chol(A),but it returns only trangular, anyone know the function can return both L and D? thank you very much~ [[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] how to omit NA without using ifelse
I have a 50*50 matrix,some entry are NAs,I want to replace these NA by 0,so can I use some syntax to do so other than using ifelse? I tried to use replace(a,NA,0),it didnt work~~(a is matrix name) Thanks~ [[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] how to create many variables at one time?
Hi: I need to create many variables at one time,how to do this in R? for eg ,X1,X2...X100? Thanks~ [[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] problem in R for Linear mixed model~
Dear R users: I just got confused some R code used in linear mixed model~ example,two factors,A, B,C,A is fixed ,B,C are random,and B is nested in C,if I wannt to use linear mixed model,are the following code correct for each case? case1:want to know random effect of B, case1-lme(y~A*B*C,random=~B|C) where B|C stand for what?,mean B is nested in C? case2: how to wirte random effect of C? case2-lme(y~A*B*C,random=~C)? this doesnt work out,it seem it must have somehing like #|$ case3.omitting the random effect for B from case1 case3-update(case1,random=~1|C),so I just type 1,so the random effect of B will be removed from the model,there only left random effect of c ,the random effect I removed ,which include both random intercept and slope ,correct?? case4:omitting the random intercept case4-update(case1,random=B-1|C) this code I got from some paper,it said by inputing B-1|C,then the random intercept is removed,so,if I want to remove random slode,I input B-2|C,it doesnt work out. case5 :how to know the both random effect of B,and C,I dont know how to wirtie this in R . And I am a little confused of these R code,especially the #|# part,what deos this syntax really mean in LME package, Thank you for your time~ [[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] problem in R for Linear mixed model~
Hi: Thank you very much,and for your case if I wirte as regress=lme(yield~nitro*Variety,data=Oats,random=~Variety|Block) what this mean? Regards 2008/6/22 Daniel Malter [EMAIL PROTECTED]: Hi, random=~1|B/C C is nested in B ##Example data=Oats regress=lme(yield~nitro*Variety,data=Oats,random=~1|Block/Variety) ##i.e. variety is nested in block summary(regress) ##End of example Best, Daniel - cuncta stricte discussurus - -Ursprüngliche Nachricht- Von: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Im Auftrag von Manli Yan Gesendet: Sunday, June 22, 2008 9:30 PM An: r-help@r-project.org Betreff: [R] problem in R for Linear mixed model~ Dear R users: I just got confused some R code used in linear mixed model~ example,two factors,A, B,C,A is fixed ,B,C are random,and B is nested in C,if I wannt to use linear mixed model,are the following code correct for each case? case1:want to know random effect of B, case1-lme(y~A*B*C,random=~B|C) where B|C stand for what?,mean B is nested in C? case2: how to wirte random effect of C? case2-lme(y~A*B*C,random=~C)? this doesnt work out,it seem it must have somehing like #|$ case3.omitting the random effect for B from case1 case3-update(case1,random=~1|C),so I just type 1,so the random effect of B will be removed from the model,there only left random effect of c ,the random effect I removed ,which include both random intercept and slope ,correct?? case4:omitting the random intercept case4-update(case1,random=B-1|C) this code I got from some paper,it said by inputing B-1|C,then the random intercept is removed,so,if I want to remove random slode,I input B-2|C,it doesnt work out. case5 :how to know the both random effect of B,and C,I dont know how to wirtie this in R . And I am a little confused of these R code,especially the #|# part,what deos this syntax really mean in LME package, Thank you for your time~ [[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.htmlhttp://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 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] try to find the MLE of a function
Yes,thats it,thank you guys so much~ 2008/6/18 Moshe Olshansky [EMAIL PROTECTED]: Hi, mle caused me some confusion too. Anyway, first of all, your x should not contain 0, so make it, let say, (1:10)/10. Secondly, f should not be the density, but -log(density), so in your case it should be f -function(theta) -sum(log(theta) + (theta-1)*log(x)) and finally you must specify a start value, i.e. your call should be mle(f,start=list(thetha=1)) --- On Thu, 19/6/08, Manli Yan [EMAIL PROTECTED] wrote: From: Manli Yan [EMAIL PROTECTED] Subject: [R] try to find the MLE of a function To: r-help@r-project.org r-help@r-project.org Received: Thursday, 19 June, 2008, 1:49 PM Hi everyone: I have a density function f(x|theta)=theta*x^(theta-1),where 0x1,0thetainfinite I want to pratice on R to find the MLE of this function,here is my code: x - (0:10)/10 f-function(theta) prod(theta*x^(theta-1)) mle(f) and r gave me :Error in eval(expr, envir, enclos) : argument is missing, with no default what mistake I just made?and how to add a constraint of theta0 in my function. Great thanks for your time~ [[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.htmlhttp://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 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] multiple multiplication in R
hi: just a very simple quesion,how to do multiple multiplication in R x-c(1,2,3,4,5,6) how to get 1*2*3*4*5*6? I checked the Arithmetic Operators in R,but did not found the operators for this function,anyway can do this except the loop? [[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] try to find the MLE of a function
Hi everyone: I have a density function f(x|theta)=theta*x^(theta-1),where 0x1,0thetainfinite I want to pratice on R to find the MLE of this function,here is my code: x - (0:10)/10 f-function(theta) prod(theta*x^(theta-1)) mle(f) and r gave me :Error in eval(expr, envir, enclos) : argument is missing, with no default what mistake I just made?and how to add a constraint of theta0 in my function. Great thanks for your time~ [[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] some R code of linear mixed model
Hello everyone: I have some quesions about the R code for linear mixed model,hope some one can give me some hints,thanks a lot~ Say: we have A B C three factors (i)A is fixed ,B and C are random,and B is nested in C,so the R code for this case would be: case1-lme(y~A+B+C+..,random=~B|C) where B|C actually mean B is nested in C, is it correct? (ii)and if I update the case as: updatedcase1-update(case1,random=~1|C), did I just remove the random effect of B, and only random effect of C left in the model? is it correct? (iii)A ,B,is fixed ,but C is random,no nested case,how am I gonna write the R code,since lme(y~A+B+C+.,random=~C) lead to error, (iv)for the case1,if I wirte this updatedcase2-update(case1,random=B-1|C) where,B-1|C,what this syntax stand for in R? thank you guys very much~ [[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] some r code for linear mixed model
Hello everyone: I have some quesions about the R code for linear mixed model,hope some one can give me some hints,thanks a lot~ Say: we have A B C three factors (i)A is fixed ,B and C are random,and B is nested in C,so the R code for this case would be: case1-lme(y~A+B+C+..,random=~B|C) where B|C actually mean B is nested in C, is it correct? (ii)and if I update the case as: updatedcase1-update(case1,random=~1|C), did I just remove the random effect of B, and only random effect of C left in the model? is it correct? (iii)A ,B,is fixed ,but C is random,no nested case,how am I gonna write the R code,since lme(y~A+B+C+.,random=~C) lead to error, (iv)for the case1,if I wirte this updatedcase2-update(case1,random=B-1|C) where,B-1|C,what this syntax stand for in R? thank you guys very much~ [[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] linear model with the repeated data type~
here is the data: y-c(5,2,3,7,9,0,1,4,5) id-c(1,1,6,6,7,8,15,15,19) t-c(50,56,50,56,50,50,50,60,50) table1-data.frame(y,id,t)//longitudinal data the above is only part of data. what I want to do is to use the linear model for each id ,then get the estimate value,like: fit1-lm(y~t,data=table1,subset=(id==1)) but ,you can see the variable id is quite irregular,they are not arranaged in order and many number missing,if I write a loop by using for,it will give me a lot NA, and for sure ,I dont want to type id=## for about 500 times so,how to get all the esimates for each id,and exclude the NA,then record those estimate in one table? great thanks ~~ [[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] create many variables at one time~
I need to create 100 variable ,whose name is id.1,id.2id.100 then I need to let a vector say id-c(id.1,id.2id.100) any easy way to do this? thanks a lot~ [[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] quite complicated case(the repeated data arranage~)
Hi everyone: I have been struggling with this repeated data type for whole afternoon,I sent two emails to server for help,many people kindly responded , hereby thank you so much,but since I dont want to write to much in email,so I divide the problem in parts,so far this seem did not work out very well,so this is my whole problem~ first I have example of data here: treatment-c(low,high,high,high,high,low,low,low,low) age-c(50,60,50,50,60,50,60,50,60) y-c(20,40,30,11,23,24,56,65,60) id-c(1,1,3,4,4,6,8,9,9) table1-cbind(treatment,id,age,y) *the actual data are way more than this*,the id is from 1~500,and not in regular ,some number missing~ all I want to do is put the cases to variable according the id for example when id =1 we have treatment1 age1 y low 5020 high 6040 this will generate a new matrix for this example I will have 6 new matrix,according to id. it is reasonable to do this in loop for,but the I met some problem: 1:how to automatically generate the new title such as treatment1 and age1 until treatment 500,age500 2:as you see,id is not strictly from 1 to 500,some time it jump from 15 to19,skip 16,17,18, if I write a loop,it will give me lot NA,certainly I need a way to avoid this and one more,say I have 100 vectors like x1x100,x1-c(1,3),x2-x(2,2),...x100-(number,number) I want to combine all this 100 vectors in one new vector say Xall which is Xall-cbind(x1,x2,...x100) //this need to type in 100 variables,take alot time so Xall will be x1x2 x3 .x100 1 2number 3 2number is here any easy way to do this,instead of inputting them one by one? *and Great Thanks for your time~~* [[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] sorting the data~
id-c(1,1,1,1,3,3,3,7,7,7,7,11,11,11) how to sort this kind of data to id:(1,1,1,1,2,2,2,3,3,3,3,4,4,4.) thanks~ [[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] sorting the data~
no,the id is variable of a table,such as: treatment id age response low 1 50 20 low 1 60 30 high5 50 30 high5 60 40 ... I want to rearranage the table according the id (increasing),since id is not strictly from 1~n,it is in increasing order but sometime jump through many number like 1 1 5 5,I like them to be 1 1 2 2~ 2008/6/4 Erik Iverson [EMAIL PROTECTED]: Are these the ranks of the data? help.search(rank) Manli Yan wrote: id-c(1,1,1,1,3,3,3,7,7,7,7,11,11,11) how to sort this kind of data to id:(1,1,1,1,2,2,2,3,3,3,3,4,4,4.) thanks~ [[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.htmlhttp://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 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] linear model in the repeated data type~
hi:lot thanks,how to use list to extract,I type allFit$coefficents,it came to nothing, such as I need to extract the estimates,how to do it by using list 2008/6/3 Austin, Matt [EMAIL PROTECTED]: How about library(nlme) allFits - lmList(y ~ t|id, data=table1, pool=FALSE) or allFits - by(table1, table1$id, function(x) lm(y ~ t, data=x)) Both ways store the results as a list, so you can access individual results using list extraction. --Matt -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Manli Yan Sent: Tuesday, June 03, 2008 9:07 PM To: r-help@r-project.org Subject: [R] linear model in the repeated data type~ here is the data: y-c(5,2,3,7,9,0,1,4,5) id-c(1,1,6,6,7,8,15,15,19) t-c(50,56,50,56,50,50,50,60,50) table1-data.frame(y,id,t)//longitudinal data what I want to do is to use the linear model for each id ,then get the estimate value,like: fit1-lm(y~t,data=table1,subset=(id==1)) but ,you can see the variable id is quite irregular,they are not arranaged in order and many number missing,if I write a loop by using for,it will give me a lot NA, and for sure ,I dont want to type id=## for about 500 times,any one know how to deal with it? [[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.htmlhttp://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 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] column and row
-fit1$coefficients y.control y.lowy.high (Intercept)19.628713 21.883999 20.023814 log(1 + (age - 45)/10) -7.383725 -6.017342 -5.084431 here is my outcome,I need one vector say b1=first row without the intercept ,like:(19.628713, 21.883999, 20.023814) and another b2=second row without the log(1 + (age - 45)/10),only the data,how to do this in R? [[alternative HTML version deleted]] __ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
[R] how to extract a specific group of data from a table?
a table with three varialbe:treatment (low,high),size(large,small),response y I want all response y with treatment=low and size=large, I try to write as : treatment-c(low,low,high,high) size-c(small,large,small,large) y-c(1,2,4,5) table1-data.frame(treatment,size,y) x=table1$y(treatment==low size==large) this does not work out~ thank you for the 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.
[R] linear model in the repeated data type~
here is the data: y-c(5,2,3,7,9,0,1,4,5) id-c(1,1,6,6,7,8,15,15,19) t-c(50,56,50,56,50,50,50,60,50) table1-data.frame(y,id,t)//longitudinal data what I want to do is to use the linear model for each id ,then get the estimate value,like: fit1-lm(y~t,data=table1,subset=(id==1)) but ,you can see the variable id is quite irregular,they are not arranaged in order and many number missing,if I write a loop by using for,it will give me a lot NA, and for sure ,I dont want to type id=## for about 500 times,any one know how to deal with it? [[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.