try waterfall plot explainability

app.py

@@ -4,6 +4,7 @@ import joblib
 import pandas as pd
 import matplotlib.pyplot as plt
 import numpy as np
+import shap
 
 project = hopsworks.login(project="SonyaStern_Lab1")
 fs = project.get_feature_store()
@@ -52,6 +53,9 @@ with gr.Blocks() as demo:
         with gr.Column():
             output = gr.Text(label="Model prediction")
             plot = gr.Plot()
+    with gr.Row():
+        with gr.Accordion("See model explanability"):
+            waterfall_plot = gr.Plot()
 
     def submit_inputs(
         age_input,
@@ -105,6 +109,11 @@ with gr.Blocks() as demo:
         ax.set_xticks(indices + bar_width / 2)
         ax.set_xticklabels(categories)
 
+        ## explainability plots
+        explainer = shap.Explainer(model)
+        shap_values = explainer(df)
+        shap_waterfall_plot = shap.plots.waterfall(shap_values[0])
+
         ## save user's data in hopsworks
         if consent_input == True:
             user_data_fg = fs.get_or_create_feature_group(
@@ -117,7 +126,7 @@ with gr.Blocks() as demo:
             user_data_df["diabetes"] = existent_info_input
             user_data_fg.insert(user_data_df)
             print("inserted new user data to hopsworks", user_data_df)
-        return res, fig
+        return res, fig, shap_waterfall_plot
 
     btn.click(
         submit_inputs,
@@ -129,7 +138,7 @@ with gr.Blocks() as demo:
             existent_info_input,
             consent_input,
         ],
-        outputs=[output, plot],
+        outputs=[output, plot, waterfall_plot],
     )
 
 demo.launch()

requirements.txt

@@ -4,4 +4,5 @@ joblib
 pandas
 scikit-learn==1.1.1
 matplotlib
-numpy
+numpy
+shap