It looks like you're applying matrix multiplication to a data.frame object. I'm not sure if that is legal R. What should work for sure is to extract the x1, x2, x3, etc. columns from the data.frame and explicitly construct a matrix from them (using cbind for example). Then, call matrix multiplication on this matrix and its transpose. Maybe it is as easy as coercing the data.frame into a matrix using as.matrix, as Martin suggested. I don't know. -Alexei On 02/19/2008 11:51 PM, Tim Ganser wrote:

Tim, As Martin and Alexei pointed out, the answer is that objects of class ?data.frame? can?t do matrix operations. as.matrix() should do the trick. Two further points: 1. When stuck on a problem like this, always check the helpfiles: ?"%*%" Yields: R Documentation Matrix Multiplication Multiplies two matrices, if they are conformable. If one argument is a vector, it will be promoted to either a row or column matrix to make the two arguments conformable. If both are vectors it will return the inner product. Usage a %*% b Arguments a, b numeric or complex matrices or vectors So you here you see under arguments that only numeric or complex matrices and vectors are "legal". As I said in section, about 95 percent of the error you get in R result from applying a command to an object of the wrong mode/class. Always check which mode/class is legal. 2. Another point: A less complicated way to get the transpose is t(X), which as ?"t" tells you actually works with both matrices and data.frames. Hope this helps, Jens From: gov2001-l-bounces at lists.fas.harvard.edu [mailto:gov2001-l-bounces at lists.fas.harvard.edu] On Behalf Of Tim Ganser Sent: Tuesday, February 19, 2008 11:51 PM To: gov2001-l at lists.fas.harvard.edu Subject: [gov2001-l] Problem with the matrix multiplication Dear all, ? I am trying to write my OLS function and I am running into a problem. It'd be great if you could take a look at what I am doing and let me know where I am making a mistake. ? The problem I get is the following: when I try to matrix-multiply x_transpose with x, I get the error message: "Error in x_trans %*% x : requires numeric matrix/vector arguments" ? Now, examining my x, using str(), I get the following info: 'data.frame':?? 1000 obs. of? 3 variables: ?$ x1: num?? 1.171? 1.419 -0.219 -0.907? 1.212 ... ?$ x2: num? 4.68 3.08 2.14 2.22 2.62 ... ?$ x3: num? 12.8 46.5 13.5 47.0 53.1 ... ? Examining my x_transpose, I get the following info: ?num [1:3, 1:1000]? 1.17? 4.68 12.84? 1.42? 3.08 ... ?- attr(*, "dimnames")=List of 2 ? ..$ : chr [1:3] "x1" "x2" "x3" ? ..$ : NULL By inspection, x_transpose contains all numbers. ? To be sure, I also transposed x, using the following for loop: x_trans <- matrix(NA, nrow = ncol(x), ncol = nrow(x)) for(i in 1:ncol(x)){ ?? for(j in 1:nrow(x)){ ????? x_trans[i,j] <- x[j,i] ?? } } Examining the result of this with str(), I get the following info: ??????? num [1:3, 1:1000]? 1.17? 4.68 12.84? 1.42? 3.08 ... ? Again, I get the same error message when attempting matrix multiplication. It would be great if someone could let me know what I am doing wrong as I have been stuck on this problem for the past couple of hours and simply can't figure it out. ? Thanks a lot. -Tim