Closed Blanch-Font closed 5 years ago
Hello,
I revised the calculation of the ATE weights for multinomial treatments of glm models. I expected the weight was sum_{i 1:n_treatmetn}(1/ps[i]), but the weights is 1/ps[,1]. Is that correct?
Hello,
I revised the calculation of the ATE weights for multinomial treatments of glm models. I expected the weight was sum_{i 1:n_treatmetn}(1/ps[i]), but the weights is 1/ps[,1]. Is that correct?