I am working with a dataset that has a lot of missing values. Because of this, I am implementing multiple imputation, by which I get M different complete datasets. After this, I train M different CoxPHFitter()'s, which I wish to pool into one model. To do this, I extract all the parameters from the models and combine them using Rubins rules. However, once I obtain the pooled coefficients, I can't find a way to place them back into the model. Simply setting the params_ of the model effectively does nothing. Is it even possible to do so?
I am working with a dataset that has a lot of missing values. Because of this, I am implementing multiple imputation, by which I get M different complete datasets. After this, I train M different
CoxPHFitter()'s, which I wish to pool into one model. To do this, I extract all the parameters from the models and combine them using Rubins rules. However, once I obtain the pooled coefficients, I can't find a way to place them back into the model. Simply setting theparams_of the model effectively does nothing. Is it even possible to do so?