Skip to content
Prev 310693 / 398506 Next

Tukey test for subgroups in a data frame

arun 
HI,
Try this:
dat1<-read.table(text="
? date name? value
? 10/5/2008????? L???? 26
? 10/5/2008????? L???? 30
? 10/5/2008????? L???? 54
? 10/5/2008????? L???? 25
? 10/5/2008???? M???? 18
? 10/5/2008???? M???? 21
? 10/5/2008???? M???? 29
? 10/5/2008???? M???? 17
? 11/5/2008????? L???? 32
? 11/5/2008????? L???? 35
? 11/5/2008????? L???? 44
? 11/5/2008????? L???? 28
? 11/5/2008???? M???? 19
? 11/5/2008???? M???? 22
? 11/5/2008???? M???? 18
? 11/5/2008???? M???? 16
? 12/5/2008????? L???? 24
? 12/5/2008????? L???? 33
? 12/5/2008????? L???? 35
? 12/5/2008????? L???? 28
? 12/5/2008???? M???? 29
? 12/5/2008???? M???? 23
? 12/5/2008???? M???? 39
? 12/5/2008???? M???? 42
? ",sep="",header=TRUE,stringsAsFactors=FALSE)
?dat1$date<-as.Date(dat1$date,format="%m/%d/%Y")
dat1<-within(dat1,{name<-factor(name)})
res<- lapply(split(dat1,dat1$date),function(x) TukeyHSD(aov(value~name,data=x),"name",ordered=TRUE))
?res
#$`2008-10-05`
?# Tukey multiple comparisons of means
? #? 95% family-wise confidence level
?? # factor levels have been ordered

#Fit: aov(formula = value ~ name, data = x)

#$name
? #? diff?????? lwr????? upr???? p adj
#L-M 12.5 -5.501854 30.50185 0.1402197


#$`2008-11-05`
?# Tukey multiple comparisons of means
? #? 95% family-wise confidence level
?? # factor levels have been ordered

#Fit: aov(formula = value ~ name, data = x)

#$name
?# ? diff???? lwr????? upr???? p adj
#L-M?? 16 7.13522 24.86478 0.0044882


----------------------------------------
A.K.

----- Original Message -----
From: raz <barvazduck at gmail.com>
To: r-help at r-project.org
Cc: 
Sent: Tuesday, November 13, 2012 11:27 AM
Subject: [R] Tukey test for subgroups in a data frame

Hello,

I have a data frame with the following columns: "date","name","value"
the name is the same for each date
I would like to get TukeyHSD p-value for the differences of "value" between
"name"s in each "date" separately I tried different ANOVA (aov()) but can
only get either tukey by "name" or by "data" but not "name by date"

any suggestions?

Thanks

Thread (3 messages)

raz Tukey test for subgroups in a data frame Nov 13 arun Tukey test for subgroups in a data frame Nov 13 arun Tukey test for subgroups in a data frame Nov 13