Hi, I'm used the pscl library to fit a zero inflated poisson model. My data has 25,327 rows (no missing values.) zeroinfl(y ~ x) works fine. I get coefficients, etc. Now, I want to see the fitted values and do some comparisons to the truth. So I used: predict(mod, type="count) The resulting vector only contains 25,213 values. So, somehow, the predict function skipped predictions for several of my data points. The problem is I don't know why and I don't know which values were skipped. Any suggestions? -- Noah Silverman, M.S. UCLA Department of Statistics 8117 Math Sciences Building Los Angeles, CA 90095