This is the first time that I've posted to this list, so if I'm doing
something wrong, please let me know. Also, if there is a searchable forum
where I can find help, that would be good too.
I'm doing a card sorting experiment, and I'm having problems imputing my
data into R for later analysis. This is what I get from each sorter:
Category Name Card numbers
Group 1: 1,3,5,7,9
Group 2: 2,4,6,8,10
I would like to use the adjusted Rand index to compare each sorter. As an
input the ARI needs a vector like this:
"Group 1" "Group 2" "Group 1" "Group 2"
"Group 1" "Group 2" "Group 1" "Group
2" "Group 1" "Group 2"
or even. 1 2 1 2 1 2 1 2 1 2
(Note: I am using the function adjustedRandIndex from the library mclust)
I also need to make a similarity matrix so that I can do cluster analysis on
the reviewers (which I am able to do).
I would like to be able to load in several of these vectors at once so that
I can create a big matrix with all of the data in it, and I currently do
this using read.table()
PROBLEM 1)
When I attempt to do the Adjusted Rand Index calculation I do this:
reviewer1 <- read.table(<filename>, sep=",") (they are csv
files)
.
Then I try to make a big matrix by doing this:
rset[1] <- reviewer1
rset[2] <- reviewer2 .
So when I try to do
adjustedRandIndex(rset[1],rset[2])
I get an error message:
Error in FUN(X, Y, ...) : comparison of these types is not implemented
And If I do this:
> rset[1]
[[1]]
[1] set1 set2 set1 set2 set1 set2 set1 set2 set1 set2
Levels: set1 set2
PROBLEM 2)
Some of my reviewers have put single cards into two piles (which was
allowed). However, this routine for the Adjusted Rand Index doesn't seem to
be able to handle that sort of category as an input. Is that a problem for
the Adj Rand Index in general? Is there a routine that can find the
Adjusted Rand Index for a different input?
Thanks,
-Steve Wolf
MSU--Lyman Briggs College
[[alternative HTML version deleted]]
