The test is manova. I tried to use manova() function, I used the code below:fit
<- manova(Y ~ X)summary(fit, test="Wilks")but I get p values for
intercept and regression coefficient as in anova() function, not for the hull
model.
Date: Mon, 7 Feb 2011 00:57:43 -0800
Subject: Re: [R] FW: multivariate regression
From: djmuser@gmail.com
To: denizsigirli@hotmail.com
CC: r-help@r-project.org
Hi:
You don't state the test for which you want the p-value, and to reiterate
what Dr. Ligges asked in response to your earlier post, how do you propose to
define a single R^2 measure? One may be able to answer your question re an
overall significance test using the anova() function:
>
Y<-matrix(c(3,5,6,3,4,2,4,5,3,2,3,5,6,3,4,2,4,5,3,2,3,5,6,3,4,2,4,5,3,2),
nrow = 10, ncol=3, byrow=TRUE)
> X1<-matrix(c(42,54,67,76,45,76,54,87,34,65), nrow = 10, ncol=1,
byrow=TRUE)X2<-matrix(c(38,21,67,76,45,76,54,87,34,65), nrow = 10, ncol=1,
byrow=TRUE)
> m <- lm(Y~X)
> anova(m) # Default is Pillai's trace
Analysis of Variance Table
Df Pillai approx F num Df den Df Pr(>F)
(Intercept) 1 0.97219 69.917 3 6 4.656e-05 ***
X 1 0.36415 1.145 3 6 0.4041
Residuals 8
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
> anova(m, test = 'Wilks') # Wilks' lambda
Analysis of Variance Table
Df Wilks approx F num Df den Df Pr(>F)
(Intercept) 1 0.02781 69.917 3 6 4.656e-05 ***
X 1 0.63585 1.145 3 6 0.4041
Residuals 8
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Roy's maximum root test and the Lawley-Hotelling statistic can also be
applied by using 'Roy' or 'Hotelling' as the value of the test =
argument of anova.lm().
HTH,
Dennis
On Sun, Feb 6, 2011 at 11:08 PM, Deniz SIGIRLI <denizsigirli@hotmail.com>
wrote:
#I have got 3 dependent variables:
Y<-matrix(c(3,5,6,3,4,2,4,5,3,2,3,5,6,3,4,2,4,5,3,2,3,5,6,3,4,2,4,5,3,2),
nrow = 10, ncol=3, byrow=TRUE)
#I've got one independent variable:
X<-matrix(c(42,54,67,76,45,76,54,87,34,65), nrow = 10, ncol=1, byrow=TRUE)
summary(lm(Y~X))
and the result is as below:
Response Y1 :
Call:
lm(formula = Y1 ~ X)
Residuals:
Min 1Q Median 3Q Max
-1.5040 -0.8838 -0.3960 1.1174 2.1162
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 4.43507 1.70369 2.603 0.0315 *
X -0.01225 0.02742 -0.447 0.6668
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Residual standard error: 1.401 on 8 degrees of freedom
Multiple R-squared: 0.02435, Adjusted R-squared: -0.09761
F-statistic: 0.1997 on 1 and 8 DF, p-value: 0.6668
Response Y2 :
Call:
lm(formula = Y2 ~ X)
Residuals:
Min 1Q Median 3Q Max
-1.4680 -0.8437 -0.2193 0.9050 1.9960
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 1.37994 1.50111 0.919 0.385
X 0.03867 0.02416 1.601 0.148
Residual standard error: 1.235 on 8 degrees of freedom
Multiple R-squared: 0.2426, Adjusted R-squared: 0.1479
F-statistic: 2.562 on 1 and 8 DF, p-value: 0.1481
Response Y3 :
Call:
lm(formula = Y3 ~ X)
Residuals:
Min 1Q Median 3Q Max
-1.7689 -0.7316 -0.1943 1.1448 2.0933
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 4.38913 1.70626 2.572 0.033 *
X -0.01149 0.02746 -0.418 0.687
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Residual standard error: 1.403 on 8 degrees of freedom
Multiple R-squared: 0.0214, Adjusted R-squared: -0.1009
F-statistic: 0.175 on 1 and 8 DF, p-value: 0.6867
There are 3 F statistics, R2 and p-values. But I want just one R2 and pvalue for
my multivariate regression model.
> Date: Fri, 4 Feb 2011 08:23:39 -0500
> From: jsorkin@grecc.umaryland.edu
> To: denizsigirli@hotmail.com; r-help@r-project.org
> Subject: Re: [R] multivariate regression
>
> Please help us help you. Follow the posting rules and send us a copy of
your code and output.
> John
> John Sorkin
> Chief Biostatistics and Informatics
> Univ. of Maryland School of Medicine
> Division of Gerontology and Geriatric Medicine
> JSorkin@grecc.umaryland.edu
> -----Original Message-----
> From: Deniz SIGIRLI <denizsigirli@hotmail.com>
> To: <r-help@r-project.org>
>
> Sent: 2/4/2011 7:54:56 AM
> Subject: [R] multivariate regression
>
>
> How can I run multivariate linear regression in R (I have got 3 dependent
variables and only 1 independent variable)? I tried lm function, but it gave
different R2 and p values for every dependent variable. I need one R2 and p
value for the model.
> [[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.
>
> Confidentiality Statement:
> This email message, including any attachments, is for ...{{dropped:22}}
