Update app.py

app.py

@@ -6,8 +6,6 @@ from inference import (
     FEATURE_NAMES,
     REPORTING_OUTCOMES,
     OUTCOME_DESCRIPTIONS,
-    OUTCOMES,
-    SHAP_OUTCOMES,
     predict_with_comparison,
     create_all_shap_plots,
     icon_array,
@@ -144,4 +142,194 @@ DONOR_FEATURES   = ["DONORF", "GRAFTYPE", "HLA_FINAL",
 DISEASE_FEATURES = ["NACS2YR", "EXCHTFPR", "VOC2YPR", "VOCFRQPR", "SCATXRSN"]
 ALL_FEATURES     = PATIENT_FEATURES + DONOR_FEATURES + DISEASE_FEATURES
 
-# ------------------------------------------------------------
+# ---------------------------------------------------------------------------
+# Utility callbacks
+# ---------------------------------------------------------------------------
+
+def get_age_group(age):
+    if age is None or age == "":
+        return ""
+    try:
+        age = float(age)
+        if age <= 10:
+            return "<=10"
+        elif age <= 17:
+            return "11-17"
+        elif age <= 29:
+            return "18-29"
+        elif age <= 49:
+            return "30-49"
+        else:
+            return ">=50"
+    except (ValueError, TypeError):
+        return ""
+
+
+def vocfrqpr_from_voc2ypr(voc_status):
+    if voc_status == "No":
+        return gr.update(value="< 3/yr", interactive=False)
+    else:
+        return gr.update(value=None, interactive=True)
+
+
+def apply_grouped_preset(selected_value):
+    if not selected_value or selected_value in HEADER_VALUES:
+        return [gr.update(value=None)] + [gr.update()] * 6
+
+    preset = PUBLISHED_PRESETS.get(selected_value)
+    if not preset:
+        return [gr.update()] * 7
+
+    return [
+        gr.update(),
+        gr.update(value=preset["DONORF"]),
+        gr.update(value=preset["CONDGRPF"]),
+        gr.update(value=preset["CONDGRP_FINAL"]),
+        gr.update(value=preset["ATGF"]),
+        gr.update(value=preset["GVHD_FINAL"]),
+        gr.update(value=preset["HLA_FINAL"]),
+    ]
+
+# ---------------------------------------------------------------------------
+# Component factory
+# ---------------------------------------------------------------------------
+
+def make_component(name: str):
+    if name == "AGE":
+        return gr.Number(label="Age at transplant (years)", minimum=0, maximum=120)
+    elif name == "AGEGPFF":
+        return gr.Textbox(label="Age group (Auto-filled)", interactive=False)
+    elif name == "NACS2YR":
+        return gr.Number(
+            label="Number of Acute Chest Syndromes within 2 years pre-HCT",
+            minimum=0,
+        )
+    elif name == "SEX":
+        return gr.Dropdown(SEX_CHOICES, label="Sex")
+    elif name == "KPS":
+        return gr.Dropdown(KPS_CHOICES, label="Karnofsky/Lansky Performance Score at HCT")
+    elif name == "DONORF":
+        return gr.Dropdown(DONORF_CHOICES, label="Donor type")
+    elif name == "GRAFTYPE":
+        return gr.Dropdown(GRAFTYPE_CHOICES, label="Graft type")
+    elif name == "CONDGRPF":
+        return gr.Dropdown(CONDGRPF_CHOICES, label="Conditioning intensity")
+    elif name == "CONDGRP_FINAL":
+        return gr.Dropdown(CONDGRP_FINAL_CHOICES, label="Conditioning Regimen")
+    elif name == "ATGF":
+        return gr.Dropdown(ATGF_CHOICES, label="Serotherapy")
+    elif name == "GVHD_FINAL":
+        return gr.Dropdown(GVHD_FINAL_CHOICES, label="GVHD Prophylaxis")
+    elif name == "HLA_FINAL":
+        return gr.Dropdown(HLA_FINAL_CHOICES, label="Donor-Recipient HLA Matching")
+    elif name == "RCMVPR":
+        return gr.Dropdown(RCMVPR_CHOICES, label="Recipient CMV serostatus")
+    elif name == "EXCHTFPR":
+        return gr.Dropdown(EXCHTFPR_CHOICES, label="Exchange transfusion required?")
+    elif name == "VOC2YPR":
+        return gr.Dropdown(
+            VOC2YPR_CHOICES,
+            label="VOC requiring hospitalization within 2 years pre-HCT?",
+        )
+    elif name == "VOCFRQPR":
+        return gr.Dropdown(VOCFRQPR_CHOICES, label="Frequency of VOC hospitalizations")
+    elif name == "SCATXRSN":
+        return gr.Dropdown(SCATXRSN_CHOICES, label="Reason for Transplant")
+    else:
+        return gr.Textbox(label=name)
+
+# ---------------------------------------------------------------------------
+# Prediction callback
+# ---------------------------------------------------------------------------
+
+def predict_gradio(*values):
+    try:
+        user_vals = {f: v for f, v in zip(ALL_FEATURES, values)}
+
+        missing = []
+        for f, v in user_vals.items():
+            if v is None or v == "" or (isinstance(v, float) and pd.isna(v)):
+                missing.append(f)
+        if missing:
+            raise ValueError(
+                f"Please fill in all fields before predicting.\nMissing: {', '.join(missing)}"
+            )
+
+        calibrated, _ = predict_with_comparison(user_vals)
+        calibrated_probs, calibrated_intervals = calibrated
+
+        rows = []
+        for outcome in REPORTING_OUTCOMES:
+            desc       = OUTCOME_DESCRIPTIONS[outcome]
+            calib_prob = calibrated_probs[outcome]
+            ci_low_c, ci_high_c = calibrated_intervals[outcome]
+            rows.append({
+                "Outcome":     desc,
+                "Probability": f"{calib_prob * 100:.1f}%",
+                "95% CI":      f"[{ci_low_c * 100:.1f}% - {ci_high_c * 100:.1f}%]",
+            })
+        df = pd.DataFrame(rows)
+
+        shap_plots = create_all_shap_plots(user_vals, max_display=10)
+
+        icon_outcomes = ["DEAD", "GF", "AGVHD", "CGVHD", "VOCPSHI", "STROKEHI"]
+        icon_plots = {o: icon_array(calibrated_probs[o], o) for o in icon_outcomes}
+
+        return (
+            df,
+            icon_plots["DEAD"],
+            icon_plots["GF"],
+            icon_plots["AGVHD"],
+            icon_plots["CGVHD"],
+            icon_plots["VOCPSHI"],
+            icon_plots["STROKEHI"],
+            shap_plots["DEAD"],
+            shap_plots["GF"],
+            shap_plots["AGVHD"],
+            shap_plots["CGVHD"],
+            shap_plots["VOCPSHI"],
+            shap_plots["EFS"],
+            shap_plots["STROKEHI"],
+            shap_plots["OS"],
+        )
+
+    except Exception as e:
+        tb = traceback.format_exc()
+        print("=" * 60)
+        print("ERROR IN predict_gradio:")
+        print(tb)
+        print("=" * 60)
+        raise gr.Error(f"{type(e).__name__}: {str(e)}\n\nSee terminal for full traceback.")
+
+# ---------------------------------------------------------------------------
+# CSS
+# ---------------------------------------------------------------------------
+
+custom_css = """
+.predict-button {
+    background: linear-gradient(to right, #ff6b35, #ff8c42) !important;
+    border: none !important;
+    color: white !important;
+    font-weight: bold !important;
+    font-size: 16px !important;
+    padding: 12px !important;
+}
+.predict-button:hover {
+    background: linear-gradient(to right, #ff5722, #ff7b29) !important;
+}
+"""
+
+# ---------------------------------------------------------------------------
+# Gradio UI
+# ---------------------------------------------------------------------------
+
+with gr.Blocks(title="HCT Outcome Prediction Model", css=custom_css) as demo:
+    gr.Markdown(
+        """
+        # HCT Outcome Prediction Model
+
+        Enter patient, transplant, and disease characteristics to predict outcomes.
+        """
+    )
+
+    inputs_dict =