From the course: Vibe Coding for Data Analysts
Join today to access over 25,200 courses taught by industry experts.
Creating explainer plots
From the course: Vibe Coding for Data Analysts
Creating explainer plots
- The best vibes come from relationships that you can explain. Getting strong predictions from sophisticated models will always make for happy analysts. What happens though, if we'd like to pull back the curtain and see our features effects? We can turn to explainer plots to get a good sense of how our features are working. Let's use our Vibe coding assistant to produce some SHAP plots. Load our XGBoost model and create SHAP plots. Let's move this over to our script and see what it's done. It looks like we have all of our packages in. Our model is loaded. Our data is loaded. Our explainer has been created, and our SHAP values. Let's try to get a summary plot. Amazing. What we have here is going to give us a really good idea of our feature importance. And we also have the impact on model output for each of our variables. This was pretty straightforward code for our assistant, but we cannot rely on our assistant to help us make sense of these plots. The relationships that we see and how…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.