Re: [R] Inconsistent linear model calculations
For the record, v6 represented the averages for columns v3-v5 because I could not quickly find the r instruction to calculate the average. Similiarly v7 represents square power of v2. At the time it was quicker than trawling through five or so user manuals for the formula syntax. __ 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] Inconsistent linear model calculations
Hi Maybe you shall reconsider your decision to avoid manuals especially R-intro.pdf, where you can learn basic operations and data structures. [EMAIL PROTECTED] napsal dne 16.05.2008 09:52:22: For the record, v6 represented the averages for columns v3-v5 because I could not quickly find the r instruction to calculate the average. Did you try ?mean. You could get to it in seconds if you used help.search(average) and then from ave page there is direct link to mean. Similiarly v7 represents square power of v2. At the time it was In many programs n^m means to m'th power of n. Least but not last the problem **was not** in R but in your data, as was suggested originally by D.Bates. Regards Petr quicker than trawling through five or so user manuals for the formula syntax. __ 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] Inconsistent linear model calculations
On Thu, 2008-05-15 at 14:50 +0100, e-letter wrote: 2008/5/15 Douglas Bates [EMAIL PROTECTED]: Did you happen to notice the part at the bottom of every message about provide commented, minimal, self-contained, reproducible code? Sorry, don't understand what that statement means in my context. Considering that the result you quote from 251 has 2 coefficients and the result from 171 has 3 coefficients one might contemplate the possibility that you are fitting different models or perhaps using different data. However we can't verify anything about the causes because we have no data regarding the problem. data below: yabcd e f 1 300 39.87 39.85 39.90 39.87333 9 2 400 45.16 45.23 45.17 45.18667 16 3 500 50.72 51.03 50.90 50.88333 25 4 600 56.85 56.80 57.02 56.89000 36 5 700 63.01 63.09 63.14 63.08000 49 6 800 69.52 59.68 69.63 66.27667 64 so substitute a in my generic equation below for data column e (e.g. 39.87...) and b for f (e.g. 9e4). Perhaps you are Brian Ripley's evil twin trying to provoke him. I would say that references to a nonexistent version 171 are deliberately provocative. No idea and couldn't care less who br is: R : Copyright 2003, The R Development Core Team Version 1.7.1 (2003-06-16) Get your facts correct before spouting further rubbish. I'm sorry e-letter, but that is just not an acceptable attitude for this list. I could explain Doug's BR comment (it was just a little bit of fun), but I guess from your answer you haven't been on the list for very long - and you might come to regret caring less who BR is - the assistance he has given me and countless other UseR's other the years has been invaluable. I'm amazed people still bothered to help you after this response - just goes to show that most people on this list are incredibly helpful and are doing their bit to help people use a wonderful piece of software; furthering the cause. There really isn't an R version 171 or even 251; the printed output from the R session *you* quote shows that. Glad you got your problem sorted in the end. G -- %~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~% Dr. Gavin Simpson [t] +44 (0)20 7679 0522 ECRC, UCL Geography, [f] +44 (0)20 7679 0565 Pearson Building, [e] gavin.simpsonATNOSPAMucl.ac.uk Gower Street, London [w] http://www.ucl.ac.uk/~ucfagls/ UK. WC1E 6BT. [w] http://www.freshwaters.org.uk %~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~%~% __ 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] Inconsistent linear model calculations
Did you happen to notice the part at the bottom of every message about provide commented, minimal, self-contained, reproducible code? Considering that the result you quote from 251 has 2 coefficients and the result from 171 has 3 coefficients one might contemplate the possibility that you are fitting different models or perhaps using different data. However we can't verify anything about the causes because we have no data regarding the problem. Perhaps you are Brian Ripley's evil twin trying to provoke him. I would say that references to a nonexistent version 171 are deliberately provocative. (251 could be considered a lazy person's attempt at 2.5.1 but 171 which presumably came after 251 doesn't make any sense to me.) On Thu, May 15, 2008 at 7:05 AM, e-letter [EMAIL PROTECTED] wrote: Readers, Using version 251 I tried the following command: lm(y~a+b,data=datafile) Resulting in, inter alia: ... coefficients (intercept) a 1.2 3.4 Packages installed: acepack ace() and avas() for selecting regression transformations adlift An adaptive lifting scheme algorithm akima Interpolation of irregularly spaced data alr3Methods and data to accompany Applied Linear Regression 3rd edition car Companion to Applied Regression codaOutput analysis and diagnostics for MCMC drc Analysis of dose-response curves EbayesThreshEmpirical Bayes thresholding and related methods emplik Empirical likelihood ratio for censored/truncated data gllmGeneralised log-linear model glmcFitting Generalized Linear Models Subject to Constraints glmmML Generalized linear models with clustering glmpath L1 Regularization Path for Generalized Linear Models and Cox Proportional Hazards Model HydroMe Estimation of Soil Hydraulic Parameters from Experimental Data JGR JGR - Java Gui for R lmtest Testing Linear Regression Models locfit Local Regression, Likelihood and Density Estimation. mvtnorm Multivariate Normal and t Distributions NISTnls Nonlinear least squares examples from NIST nlmeLinear and Nonlinear Mixed Effects Models nlstoolsTools for nonlinear regression diagnostics nlt A nondecimated lifting transform for signal denoising nlts(non)linear time series analysis nnlsThe Lawson-Hanson algorithm for non-negative least squares (NNLS) plotrix Various plotting functions quantregQuantile Regression Rcmdr ** No title available (pre-2.0.0 install?) ** sandwichRobust Covariance Matrix Estimators SparseM Sparse Linear Algebra sspline Smoothing Splines on the Sphere zoo Z's ordered observations Packages in library '/usr/lib/R/library': baseThe R Base Package bootBootstrap R (S-Plus) Functions (Canty) class Functions for Classification cluster Cluster Analysis Extended Rousseeuw et al. codetools Code Analysis Tools for R datasetsThe R Datasets Package foreign Read Data Stored by Minitab, S, SAS, SPSS, Stata, Systat, dBase, ... graphicsThe R Graphics Package grDevices The R Graphics Devices and Support for Colours and Fonts gridThe Grid Graphics Package KernSmooth Functions for kernel smoothing for Wand Jones (1995) lattice Lattice Graphics MASSMain Package of Venables and Ripley's MASS methods Formal Methods and Classes mgcvGAMs with GCV smoothness estimation and GAMMs by REML/PQL nlmeLinear and Nonlinear Mixed Effects Models nnetFeed-forward Neural Networks and Multinomial Log-Linear Models quadprogFunctions to solve Quadratic Programming lattice Lattice Graphics MASSMain Package of Venables and Ripley's MASS methods Formal Methods and Classes mgcvGAMs with GCV smoothness estimation and GAMMs by REML/PQL nlmeLinear and Nonlinear Mixed Effects Models nnet
Re: [R] Inconsistent linear model calculations
Hi, To get help with this problem, you will have to create an example that others can duplicate. That is why each message to the list says (at the bottom): PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. Before you do any of that, please run some textbook examples using lm(), which will tell you whether lm() is working or not. It is ***extremely*** unlikely that for a function as stable and heavily- used as lm(), the same data and the same command produced different results on different computers, different versions of R, or different OSs. Your list of packages installed is not informative, because these packages are unlikely to affect your running of lm(). To tell the list which version of R you are running please include the output of sessionInfo(). You will find that many people are eager to help beginners, as long as they give enough information about the problem they encountered. _ Professor Michael Kubovy University of Virginia Department of Psychology USPS: P.O.Box 400400Charlottesville, VA 22904-4400 Parcels:Room 102Gilmer Hall McCormick RoadCharlottesville, VA 22903 Office:B011+1-434-982-4729 Lab:B019+1-434-982-4751 Fax:+1-434-982-4766 WWW:http://www.people.virginia.edu/~mk9y/ On May 15, 2008, at 8:05 AM, e-letter wrote: Readers, Using version 251 I tried the following command: lm(y~a+b,data=datafile) Resulting in, inter alia: ... coefficients (intercept) a 1.2 3.4 Packages installed: snip Why do I get different results when entering the same equation command? The correct answer was obtained using the older version of the software, so I want to replicate using the new version, which is installed onto a new pc that I am using. Any advice. Yours, [EMAIL PROTECTED] [[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] Inconsistent linear model calculations
2008/5/15 Douglas Bates [EMAIL PROTECTED]: Did you happen to notice the part at the bottom of every message about provide commented, minimal, self-contained, reproducible code? Sorry, don't understand what that statement means in my context. Considering that the result you quote from 251 has 2 coefficients and the result from 171 has 3 coefficients one might contemplate the possibility that you are fitting different models or perhaps using different data. However we can't verify anything about the causes because we have no data regarding the problem. data below: yabcd e f 1 300 39.87 39.85 39.90 39.87333 9 2 400 45.16 45.23 45.17 45.18667 16 3 500 50.72 51.03 50.90 50.88333 25 4 600 56.85 56.80 57.02 56.89000 36 5 700 63.01 63.09 63.14 63.08000 49 6 800 69.52 59.68 69.63 66.27667 64 so substitute a in my generic equation below for data column e (e.g. 39.87...) and b for f (e.g. 9e4). Perhaps you are Brian Ripley's evil twin trying to provoke him. I would say that references to a nonexistent version 171 are deliberately provocative. No idea and couldn't care less who br is: R : Copyright 2003, The R Development Core Team Version 1.7.1 (2003-06-16) Get your facts correct before spouting further rubbish. __ 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] Inconsistent linear model calculations
e-letter inpost at gmail.com writes: 2008/5/15 Douglas Bates bates at stat.wisc.edu: Did you happen to notice the part at the bottom of every message about provide commented, minimal, self-contained, reproducible code? Sorry, don't understand what that statement means in my context. It means that we need to be able to see exactly what you're trying to do -- you haven't given us enough information. Considering that the result you quote from 251 has 2 coefficients and the result from 171 has 3 coefficients one might contemplate the possibility that you are fitting different models or perhaps using different data. However we can't verify anything about the causes because we have no data regarding the problem. in version 2.7.0 (both of the versions you are using are very old): data1 = read.table(textConnection(yabcd e f\ + 1 300 39.87 39.85 39.90 39.87333 9\ + 2 400 45.16 45.23 45.17 45.18667 16\ + 3 500 50.72 51.03 50.90 50.88333 25\ + 4 600 56.85 56.80 57.02 56.89000 36\ + 5 700 63.01 63.09 63.14 63.08000 49\ + 6 800 69.52 59.68 69.63 66.27667 64), +header=TRUE) I wasn't sure from your answers whether you really meant to use a and b or whether you meant to use e and f lm(y~e+f,data=data1) Call: lm(formula = y ~ e + f, data = data1) Coefficients: (Intercept)ef -4.097e+001.234e-012.929e-06 lm(y~a+b,data=data1) Call: lm(formula = y ~ a + b, data = data1) Coefficients: (Intercept)ab -2.000e+001.000e-02 -1.294e-15 Repeating what Doug Bates said above: the result that you quote for 2.5.1, with only two estimates for a three-parameter model (intercept plus two covariates), is surprising indeed, and strongly suggests a user error. I would also be very surprised if anything had changed in the answers I quote above between 2.5.1 and the current version of R. Please remember how much you paid for R, and for this advice, and adjust your tone accordingly ... cheers Ben Bolker __ 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] Inconsistent linear model calculations
On 15/05/2008, Michael Kubovy [EMAIL PROTECTED] wrote: Hi, Your reply doesn't help you with your lm() problem. You need to show the problem with lm() as well. Your original message was unclear on what is wrong. The problem I have with lm() is that when I enter this command, I obtain different answers! So here is a todo list, before you do anything else: (1) Get the most recent version of R (2.7.0). No thank you. I rely on my urpmi repository. (2) Make sure you issue the command update.packages() at least once a week You must be joking! Waste my time every week to perform a basic task such as interpolation, or least fit squares linear regression? I am a novice seeking only to perform rudimentary statistics. (3) Find a textbook example of a linear model. Run it using lm(). (a) If it comes out right, then nothing is wrong with the function, and you've been confused somewhere else. Well, I sought a statistician who showed me how to do this task using mathematica. Despite advice otherwise, I am interested to learn how to perform this basic task using r because it's free and and use linux. The output from the lm() command using version 171 yielded the same result as obtained from the statistician who helped me. In contrast enabling the version 251 did not yield the same result, hence this original post. (b) If it comes out wrong, read the help page for lm(), get a textbook or read some of the introductory materials on CRAN. lm() is never wrong (this is not necessarily true for all R functions, but it's true for most). (4) Before you ask the list for help, read http://www.R-project.org/posting-guide.html carefully, and follow the rules religiously. Done, but not religiously. ;) (5) Generally do not reply to the person who helped you, but to the list. You can cc the person who helped you, but that is rarely appropriate (avoid cluttering people's mailboxes). (6) Do *not* be rude (as in your recent message). You will be shunned. Ripley and Bates are two of the most distinguished statisticians alive. Accept their criticism gracefully. Telling a user the software version is nonexistent is equally rude; I've little respect for such typical techie petulance. (7) Your most recent message doesn't help enough. Write it so that if I copy and paste your code into R, I can reproduce the result you're worried about. I didn't know you can copy data directly into r; why not simply copy the text to your text editor and save in csv format? Yours, __ 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] Inconsistent linear model calculations
On Thu, May 15, 2008 at 3:05 PM, e-letter [EMAIL PROTECTED] wrote: Using version 251 I tried the following command: lm(y~a+b,data=datafile) Resulting in, inter alia: ... coefficients (intercept) a 1.2 3.4 [...] When using version 171 I entered the same command: lm(y~a+b,data=datafile) Resulting in, inter alia: ... coefficients (intercept) a b 5.6 7.8 9.01 Why do I get different results when entering the same equation command? The correct answer was obtained using the older version of the software, so I want to replicate using the new version, which is installed onto a new pc that I am using. Any advice. First check if your data frame is the same on both occasions. (It definitely wasn't the data set you provided in a later message - so it sounds possible you've made another error.) Secondly, try to make sure you're copying real output here. You should have three coefficients if you ask for three. Third, check your lm function: lm%in%ls() Maybe an evil twin of yours has modified it, for instance, this way: lm- function(...) cat(coefficients\ (intercept) \ta\ 1.2 \t\t3.4\n) [[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] Inconsistent linear model calculations
On 16/05/2008, at 7:23 AM, e-letter wrote:
Well, I sought a statistician who showed me how to do this task using
mathematica. Despite advice otherwise,
You are certainly getting bad advice!
I am interested to learn how to
perform this basic task using r because it's free and and use linux.
That's ``R'' not ``r''.
The output from the lm() command using version 171 yielded the same
result as obtained from the statistician who helped me. In contrast
enabling the version 251 did not yield the same result, hence this
original post.
As was previously pointed out:
***THERE IS NOT SUCH THING AS VERSION 251 OF R!!!***
Nor is there a version ``171''.
You can find out what version you are actually using by
typing ``version''. For example, when I type ``version''
I get the following:
version
_
platform i386-apple-darwin8.10.1
arch i386
os darwin8.10.1
system i386, darwin8.10.1
status
major 2
minor 7.0
year 2008
month 04
day22
svn rev45424
language R
version.string R version 2.7.0 (2008-04-22)
(``Arrows'' - added.) So the version of R that I am
running is 2.7.0.
cheers,
Rolf Turner
##
Attention:\ This e-mail message is privileged and confid...{{dropped:9}}
__
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] Inconsistent linear model calculations
Below is direct copy from command terminals of both pcs (mandrake 92
with r 171; mandriva 2008 with r 251, respectively).
R : Copyright 2003, The R Development Core Team
Version 1.7.1 (2003-06-16)
R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type `license()' or `licence()' for distribution details.
R is a collaborative project with many contributors.
Type `contributors()' for more information.
Type `demo()' for some demos, `help()' for on-line help, or
`help.start()' for a HTML browser interface to help.
Type `q()' to quit R.
dubious-read.table('/path/to/file/dodgy.csv')
dubious
V1 V2V3V4V5 V6 V7
1 1 300 39.87 39.85 39.90 39.87333 9
2 2 400 45.16 45.23 45.17 45.18667 16
3 3 500 50.72 51.03 50.90 50.88333 25
4 4 600 56.85 56.80 57.02 56.89000 36
5 5 700 63.01 63.09 63.14 63.08000 49
6 6 800 69.52 69.82 69.63 69.65667 64
lm(V6~V2+V7,data=dubious)
Call:
lm(formula = V6 ~ V2 + V7, data = dubious)
Coefficients:
(Intercept) V2 V7
2.553e+014.332e-021.480e-05
R version 2.5.1 (2007-06-27)
Copyright (C) 2007 The R Foundation for Statistical Computing
ISBN 3-900051-07-0
R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.
Natural language support but running in an English locale
R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.
Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.
dubious-read.table('/path/to/file/dodgy.csv')
dubious
V1 V2V3V4V5 V6 V7
1 1 300 39.87 39.85 39.90 39.87333 9
2 2 400 45.16 45.23 45.17 45.18667 16
3 3 500 50.72 51.03 50.90 50.88333 25
4 4 600 56.85 56.80 57.02 56.89000 36
5 5 700 63.01 63.09 63.14 63.08000 49
6 6 800 69.52 59.68 69.63 66.27667 64
lm(V6~V2+V7,data=dubious)
Call:
lm(formula = V6 ~ V2 + V7, data = dubious)
Coefficients:
(Intercept) V2 V7
1.937e+017.168e-02 -1.537e-05
Below is the csv file itself:
1 300 39.87 39.85 39.90 39.87333 9
2 400 45.16 45.23 45.17 45.18667 16
3 500 50.72 51.03 50.90 50.88333 25
4 600 56.85 56.80 57.02 56.89000 36
5 700 63.01 63.09 63.14 63.08000 49
6 800 69.52 59.68 69.63 66.27667 64
Enjoy! :)
__
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] Inconsistent linear model calculations
On May 15, 2008, at 5:37 PM, e-letter wrote:
Below is direct copy from command terminals of both pcs (mandrake 92
with r 171; mandriva 2008 with r 251, respectively).
R : Copyright 2003, The R Development Core Team
Version 1.7.1 (2003-06-16)
R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type `license()' or `licence()' for distribution details.
R is a collaborative project with many contributors.
Type `contributors()' for more information.
Type `demo()' for some demos, `help()' for on-line help, or
`help.start()' for a HTML browser interface to help.
Type `q()' to quit R.
dubious-read.table('/path/to/file/dodgy.csv')
dubious
V1 V2V3V4V5 V6 V7
1 1 300 39.87 39.85 39.90 39.87333 9
2 2 400 45.16 45.23 45.17 45.18667 16
3 3 500 50.72 51.03 50.90 50.88333 25
4 4 600 56.85 56.80 57.02 56.89000 36
5 5 700 63.01 63.09 63.14 63.08000 49
6 6 800 69.52 69.82 69.63 69.65667 64
This is exactly why we asked you for a reproducible example and full
code. That last entry in the V6 column is 69.65667 in this case, but
66.27667 in the other cases. So you clearly are working with two
slightly different dodgy files, and consequently two slightly
different dubious data sets, and lm rightfully produces two slightly
different accurate coefficients.
Haris Skiadas
Department of Mathematics and Computer Science
Hanover College
lm(V6~V2+V7,data=dubious)
Call:
lm(formula = V6 ~ V2 + V7, data = dubious)
Coefficients:
(Intercept) V2 V7
2.553e+014.332e-021.480e-05
R version 2.5.1 (2007-06-27)
Copyright (C) 2007 The R Foundation for Statistical Computing
ISBN 3-900051-07-0
R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.
Natural language support but running in an English locale
R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.
Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.
dubious-read.table('/path/to/file/dodgy.csv')
dubious
V1 V2V3V4V5 V6 V7
1 1 300 39.87 39.85 39.90 39.87333 9
2 2 400 45.16 45.23 45.17 45.18667 16
3 3 500 50.72 51.03 50.90 50.88333 25
4 4 600 56.85 56.80 57.02 56.89000 36
5 5 700 63.01 63.09 63.14 63.08000 49
6 6 800 69.52 59.68 69.63 66.27667 64
lm(V6~V2+V7,data=dubious)
Call:
lm(formula = V6 ~ V2 + V7, data = dubious)
Coefficients:
(Intercept) V2 V7
1.937e+017.168e-02 -1.537e-05
Below is the csv file itself:
1 300 39.87 39.85 39.90 39.87333 9
2 400 45.16 45.23 45.17 45.18667 16
3 500 50.72 51.03 50.90 50.88333 25
4 600 56.85 56.80 57.02 56.89000 36
5 700 63.01 63.09 63.14 63.08000 49
6 800 69.52 59.68 69.63 66.27667 64
Enjoy! :)
__
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] Inconsistent linear model calculations
On 16/05/2008, at 9:44 AM, Charilaos Skiadas wrote:
This is exactly why we asked you for a reproducible example and
full code. That last entry in the V6 column is 69.65667 in this
case, but 66.27667 in the other cases. So you clearly are working
with two slightly different dodgy files, and consequently two
slightly different dubious data sets, and lm rightfully produces
two slightly different accurate coefficients.
Well spotted!!! I stared and stared and was flummoxed as to what
was going on.
Good eyes!
cheers,
Rolf Turner
##
Attention:\ This e-mail message is privileged and confid...{{dropped:9}}
__
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] Inconsistent linear model calculations
Thank you to all you eagle eyes; amendment made accordingly and solved. Not sure how the difference occurred... [[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] Inconsistent linear model calculations
Also, it's worth pointing out the reason for the numerical instability
of the parameter estimates: the predictors are nearly collinear.
(dubious - read.table('clipboard'))
V1 V2V3V4V5 V6 V7
1 1 300 39.87 39.85 39.90 39.87333 9
2 2 400 45.16 45.23 45.17 45.18667 16
3 3 500 50.72 51.03 50.90 50.88333 25
4 4 600 56.85 56.80 57.02 56.89000 36
5 5 700 63.01 63.09 63.14 63.08000 49
6 6 800 69.52 59.68 69.63 66.27667 64
round(cor(dubious), 3)
V1V2V3V4V5V6V7
V1 1.000 1.000 0.999 0.951 0.999 0.997 0.991
V2 1.000 1.000 0.999 0.951 0.999 0.997 0.991
V3 0.999 0.999 1.000 0.942 1.000 0.995 0.996
V4 0.951 0.951 0.942 1.000 0.943 0.970 0.912
V5 0.999 0.999 1.000 0.943 1.000 0.995 0.995
V6 0.997 0.997 0.995 0.970 0.995 1.000 0.983
V7 0.991 0.991 0.996 0.912 0.995 0.983 1.000
Note that the correlation for V2 and V7 is about .991
Kingsford Jones
On Thu, May 15, 2008 at 3:01 PM, e-letter [EMAIL PROTECTED] wrote:
Thank you to all you eagle eyes; amendment made accordingly and solved. Not
sure how the difference occurred...
[[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.
