ready

app.py

@@ -1,6 +1,7 @@
-import random
-
+# URL: https://huggingface.co/spaces/gradio/xgboost-income-prediction-with-explainability
+# imports
 import gradio as gr
+import random
 import matplotlib
 import matplotlib.pyplot as plt
 import pandas as pd
@@ -8,16 +9,13 @@ import shap
 import xgboost as xgb
 from datasets import load_dataset
 
-print(gr.__version__)
 
+# loading the model and setting up
 matplotlib.use("Agg")
-
 dataset = load_dataset("scikit-learn/adult-census-income")
-
 X_train = dataset["train"].to_pandas()
 _ = X_train.pop("fnlwgt")
 _ = X_train.pop("race")
-
 y_train = X_train.pop("income")
 y_train = (y_train == ">50K").astype(int)
 categorical_columns = [
@@ -30,12 +28,11 @@ categorical_columns = [
     "native.country",
 ]
 X_train = X_train.astype({col: "category" for col in categorical_columns})
-
-
 data = xgb.DMatrix(X_train, label=y_train, enable_categorical=True)
 model = xgb.train(params={"objective": "binary:logistic"}, dtrain=data)
 explainer = shap.TreeExplainer(model)
 
+# defining the two core fns
 
 def predict(*args):
     df = pd.DataFrame([args], columns=X_train.columns)
@@ -67,12 +64,17 @@ unique_occupation = sorted(X_train["occupation"].unique())
 unique_sex = sorted(X_train["sex"].unique())
 unique_country = sorted(X_train["native.country"].unique())
 
+# starting the block 
+
 with gr.Blocks() as demo:
+    # defining text on the page
     gr.Markdown("""
     **Income Classification with XGBoost 💰**:  This demo uses an XGBoost classifier predicts income based on demographic factors, along with Shapley value-based *explanations*. The [source code for this Gradio demo is here](https://huggingface.co/spaces/gradio/xgboost-income-prediction-with-explainability/blob/main/app.py).
     """)
+    # defining the layout
     with gr.Row():
         with gr.Column():
+            # defining the inputs
             age = gr.Slider(label="Age", minimum=17, maximum=90, step=1, randomize=True)
             work_class = gr.Dropdown(
                 label="Workclass",
@@ -128,11 +130,14 @@ with gr.Blocks() as demo:
                 value=lambda: random.choice(unique_country),
             )
         with gr.Column():
+            # defining the outputs
             label = gr.Label()
             plot = gr.Plot()
             with gr.Row():
+                # defining the buttons
                 predict_btn = gr.Button(value="Predict")
                 interpret_btn = gr.Button(value="Explain")
+            # defining the fn that will run when predict is clicked, what it will get as inputs, and which output it will update 
             predict_btn.click(
                 predict,
                 inputs=[
@@ -151,6 +156,7 @@ with gr.Blocks() as demo:
                 ],
                 outputs=[label],
             )
+            # defining the fn that will run when interpret is clicked, what it will get as inputs, and which output it will update 
             interpret_btn.click(
                 interpret,
                 inputs=[
@@ -170,4 +176,5 @@ with gr.Blocks() as demo:
                 outputs=[plot],
             )
 
+# launch
 demo.launch()