Pretty simple, just call in the HurdleLogNormal class. It runs like any scikit-learn regression.
model = HurdleLogNormal().fit(X, y)
Access the logistic coefficients model.logistic.coef_ model.logistic.intercept_
Access the regression coefficients model.linear.named_step["regressor"].coef_ model.linear.named_step["regressor"].intercept_
As a note, I used Ridge regression to help with cases where there was weird numerical instability (i think?). Working on creating utility functions that make it easier to access the coefficients / logFC for instances where people might use this as a drop-in for MAST.