6.8 Interpreting models with Bambi

We have been using bmb.interpret_plot_predictions a lot in this chapter. But that’s not the only tool that Bambi offers us to help us understand models. One of them is bmb.interpret_plot_comparisons. This tool helps us answer the question, ”What is the expected predictive difference when we compare two values of a given variable while keeping all the rest at constant values?”.

Let’s use model_int from the previous section, so we don’t need to fit a new model. We use the following code block to generate Figure 6.15:

Code 6.20

bmb.interpret.plot_comparisons(model_int, idata_int, 
                               contrast={"bill_depth":[1.4, 1.8]}, 
                ...