Sam Wong
2014-Aug-01 12:09 UTC
[R] Weighted Logistic Regression in R: Can R do what SAS PROC LOGISTIC does?
I wonder if anyone has written some additional R code to perform weighted logistic regression in the way of SAS PROC LGOISTIC WEIGHT statement. I want to weight the sample using a vector of probabilities generated from Dirichlet distribution. It is known that the R function glm has a WEIGHT option but it only takes a vector of integers which are the number of trials for each response. I wonder if anyone has written other R functions to perform the same job SAS PROC LGGISTIC WEIGHT statement does? Here is the post I read at stats.stackexchange.com Case weighted logistic regression - Cross Validated Case weighted logistic regression - Cross Validated I'm looking at a few logistic regression issues. ("regular" and "conditional"). View on stats.stackexchange.com Preview by Yahoo [[alternative HTML version deleted]]