Pram, pd and colSums.

[matthias-da]

From Miranda Faranna:

I have regarding the "pd" parameter in the "pram" function. I initially thought it represented the minimum value that probabilities on the diagonal of the transition matrix could assume. However, I have found that some probabilities are lower than the specified number.

Additionally, I wanted to bring to your attention that the SdcMicro help manual for the "pram" function establishes certain criteria for the transition matrix. One of these criteria is that the rowSums and colSums of the matrix need to equal one. I have confirmed that indeed the rowSums equal one, but it appears that the colSums do not necessarily need to. I trust this information proves useful to you, and I would appreciate it if you could verify this.