Hi Juvin,
The error "dim(X) must have a positive length" usually shows when you are
passing a vector to "apply", ie.
apply(1:5,2,mean)
#Error in apply(1:5, 2, mean) : dim(X) must have a positive length
Also, if your dataset originally has "1206" columns, it is not clear why you
n
igues
To: Rui Barradas
Cc: r-help@r-project.org
Sent: Wednesday, January 16, 2013 11:23 PM
Subject: Re: [R] Mean calculation by two variables
Hello Rui!
Thanks a lot for your help!
Sorry for make mistake with the factor's name, was inattention during data
preparation for the asking
Hello Rui!
Thanks a lot for your help!
Sorry for make mistake with the factor's name, was inattention during data
preparation for the asking mail.
Unfortunately, it not simple mean calculation, because same individual
could have more than one row, if it eat more than a food item. In example
data
Hello,
If you want to calculate the mean weight by Specie and Food.item, you
can use ?aggregate.
In what follows, I've named your EXAMPLE ex and your SOLUTION sol. Note
that the result is different from sol.
(No exemplo não há Peixe, é Fish, e os números são outros.)
aggregate(Weight ~ Specie +
Hello,
Or ?tapply.
ave will return a vector with the length of the input, tapply just one
value per group.
ave(test$Score, test$Name, FUN = mean) # 24 values
tapply(test$Score, test$Name, FUN = mean) # 3 values
Hope this helps,
Rui Barradas
Em 25-09-2012 15:54, Bert Gunter escreveu:
Nic
Nico:
1. Thanks for the example.
2. Please read -- or at least peruse -- the tutorial, An Introduction
to R (or other tutorial that you might find on the Web, there are
many) to familiarize yourself with R's features.
3. This is an example of what the apply() family of functions can do,
so you m
Hello,
I was hunting for something else and came across your old message, and
I noticed that it had never received an answer (at least on the list).
I really tried to figure out what might be going on just by looking
at your code, but I could not get enough information from your email
and your co
7 matches
Mail list logo
