Upload app.py

app.py

@@ -21,22 +21,41 @@ from datasets import load_dataset
 
 #Read data training data.
 
-x1 = load_dataset("mertkarabacak/NCDB-GBM", data_files="6m_data_resampled.csv", use_auth_token = HF_TOKEN)
+x1 = load_dataset("mertkarabacak/NCDB-GBM", data_files="6m_data_train.csv", use_auth_token = HF_TOKEN)
 x1 = pd.DataFrame(x1['train'])
 x1 = x1.iloc[:, 1:]
 
-x2 = load_dataset("mertkarabacak/NCDB-GBM", data_files="12m_data_resampled.csv", use_auth_token = HF_TOKEN)
+x2 = load_dataset("mertkarabacak/NCDB-GBM", data_files="12m_data_train.csv", use_auth_token = HF_TOKEN)
 x2 = pd.DataFrame(x2['train'])
 x2 = x2.iloc[:, 1:]
 
-x3 = load_dataset("mertkarabacak/NCDB-GBM", data_files="18m_data_resampled.csv", use_auth_token = HF_TOKEN)
+x3 = load_dataset("mertkarabacak/NCDB-GBM", data_files="18m_data_train.csv", use_auth_token = HF_TOKEN)
 x3 = pd.DataFrame(x3['train'])
 x3 = x3.iloc[:, 1:]
 
-x4 = load_dataset("mertkarabacak/NCDB-GBM", data_files="24m_data_resampled.csv", use_auth_token = HF_TOKEN)
+x4 = load_dataset("mertkarabacak/NCDB-GBM", data_files="24m_data_train.csv", use_auth_token = HF_TOKEN)
 x4 = pd.DataFrame(x4['train'])
 x4 = x4.iloc[:, 1:]
 
+#Read validation data.
+
+x1_valid = load_dataset("mertkarabacak/NCDB-GBM", data_files="6m_data_valid.csv", use_auth_token = HF_TOKEN)
+x1_valid = pd.DataFrame(x1_valid['train'])
+x1_valid = x1_valid.iloc[:, 1:]
+
+x2_valid = load_dataset("mertkarabacak/NCDB-GBM", data_files="12m_data_valid.csv", use_auth_token = HF_TOKEN)
+x2_valid = pd.DataFrame(x2_valid['train'])
+x2_valid = x2_valid.iloc[:, 1:]
+
+x3_valid = load_dataset("mertkarabacak/NCDB-GBM", data_files="18m_data_valid.csv", use_auth_token = HF_TOKEN)
+x3_valid = pd.DataFrame(x3_valid['train'])
+x3_valid = x3_valid.iloc[:, 1:]
+
+x4_valid = load_dataset("mertkarabacak/NCDB-GBM", data_files="24m_data_valid.csv", use_auth_token = HF_TOKEN)
+x4_valid = pd.DataFrame(x4_valid['train'])
+x4_valid = x4_valid.iloc[:, 1:]
+
+
 #Define feature names.
 f1_names = list(x1.columns)
 f1_names = [f1.replace('__', ' - ') for f1 in f1_names]
@@ -56,17 +75,28 @@ f4_names = [f4.replace('_', ' ') for f4 in f4_names]
 
 
 #Prepare training data for the outcome 1.
-y1 = x1.pop('OUTCOME')
+y1_valid = x1_valid.pop('OUTCOME')
 
 #Prepare training data for the outcome 2.
-y2 = x2.pop('OUTCOME')
+y2_valid = x2_valid.pop('OUTCOME')
 
 #Prepare training data for the outcome 3.
-y3 = x3.pop('OUTCOME')
+y3_valid = x3_valid.pop('OUTCOME')
 
-#Prepare training data for the outcome 3.
-y4 = x4.pop('OUTCOME')
+#Prepare training data for the outcome 4.
+y4_valid = x4_valid.pop('OUTCOME')
+
+#Prepare validation data for the outcome 1.
+y1_valid = x1_valid.pop('OUTCOME')
+
+#Prepare validation data for the outcome 2.
+y2_valid = x2_valid.pop('OUTCOME')
 
+#Prepare validation data for the outcome 3.
+y3_valid = x3_valid.pop('OUTCOME')
+
+#Prepare validation data for the outcome 4.
+y4_valid = x4_valid.pop('OUTCOME')
 
 #Training models.
 
@@ -77,7 +107,7 @@ y1_model = tabpfn
 y1_model = y1_model.fit(x1, y1, overwrite_warning=True)
 
 y1_calib_model = CalibratedClassifierCV(y1_model, method='isotonic', cv='prefit')
-y1_calib_model = y1_calib_model.fit(x1, y1)
+y1_calib_model = y1_calib_model.fit(x1_valid, y1_valid)
 
 y1_explainer = shap.Explainer(y1_model.predict, x1)
 
@@ -89,7 +119,7 @@ y2_model = tabpfn
 y2_model = y2_model.fit(x2, y2, overwrite_warning=True)
 
 y2_calib_model = CalibratedClassifierCV(y2_model, method='isotonic', cv='prefit')
-y2_calib_model = y2_calib_model.fit(x2, y2)
+y2_calib_model = y2_calib_model.fit(x2_valid, y2_valid)
 
 y2_explainer = shap.Explainer(y2_model.predict, x2)
 
@@ -101,7 +131,7 @@ y3_model = tabpfn
 y3_model = y3_model.fit(x3, y3, overwrite_warning=True)
 
 y3_calib_model = CalibratedClassifierCV(y3_model, method='isotonic', cv='prefit')
-y3_calib_model = y3_calib_model.fit(x3, y3)
+y3_calib_model = y3_calib_model.fit(x3_valid, y3_valid)
 
 y3_explainer = shap.Explainer(y3_model.predict, x3)
 
@@ -113,7 +143,7 @@ y4_model = tabpfn
 y4_model = y4_model.fit(x4, y4, overwrite_warning=True)
 
 y4_calib_model = CalibratedClassifierCV(y4_model, method='isotonic', cv='prefit')
-y4_calib_model = y4_calib_model.fit(x4, y4)
+y4_calib_model = y4_calib_model.fit(x4_valid, y4_valid)
 
 y4_explainer = shap.Explainer(y4_model.predict, x4)
 
@@ -307,7 +337,7 @@ with gr.Blocks(title = "NCDB-Meningioma") as demo:
           </tr>
           <tr>
             <td>6-Month Mortality</td>
-            <td>Random Forest</td>
+            <td>TabPFN</td>
             <td>0.755 (0.733 - 0.777)</td>
             <td>0.767 (0.745 - 0.789)</td>
             <td>0.764 (0.742 - 0.786)</td>