Update app.py

app.py

@@ -67,17 +67,14 @@ def build_train_test_for_window(df: pd.DataFrame, start_sel: date, end_sel: date
     Train = all history strictly before start_sel (business days).
     Test  = start_sel..end_sel inclusive (business days).
     """
-    # ensure B-day indexes exist
-    idx = df.index
-    # align selection to nearest business days inside universe
+    # align selection to universe
     start_sel = clip_to_universe(start_sel)
     end_sel   = clip_to_universe(end_sel)
     if start_sel > end_sel:
         raise ValueError("Selected start date must be <= end date.")
 
-    # Slice by label (dates) safely
-    train_mask = idx < pd.Timestamp(start_sel)
-    test_mask  = (idx >= pd.Timestamp(start_sel)) & (idx <= pd.Timestamp(end_sel))
+    train_mask = df.index < pd.Timestamp(start_sel)
+    test_mask  = (df.index >= pd.Timestamp(start_sel)) & (df.index <= pd.Timestamp(end_sel))
 
     train_s = df.loc[train_mask, 'Close']
     test_s  = df.loc[test_mask,  'Close']
@@ -114,11 +111,11 @@ def arima_roll_eval(train_s: pd.Series, test_s: pd.Series, order=(2,1,2)):
     preds, trues = [], []
     for i in range(len(test_s)):
         fcast = res.forecast(1)
-        fval = float(np.asarray(fcast).reshape(-1)[0])  # robust scalar, avoids FutureWarning
+        fval = float(np.asarray(fcast).reshape(-1)[0])  # robust scalar
         preds.append(fval)
         trues.append(float(test_s.iloc[i]))
         try:
-            res = res.append(test_s.iloc[i:i+1])  # update state with the realized value
+            res = res.append(test_s.iloc[i:i+1])  # update state with realized value
         except Exception:
             res = ARIMA(pd.concat([train_s, test_s.iloc[:i+1]]), order=order,
                         enforce_stationarity=False, enforce_invertibility=False).fit()
@@ -212,7 +209,7 @@ def lstm_roll_eval(train_s: pd.Series, test_s: pd.Series):
     return np.array(trues), np.array(preds)
 
 # =========================
-# Metrics & formatting
+# Metrics & plotting
 # =========================
 def metrics_df(y_true, y_pred) -> pd.DataFrame:
     rmse = math.sqrt(mean_squared_error(y_true, y_pred))
@@ -233,7 +230,7 @@ def plot_series(dates, actual, pred, title):
 # =========================
 # Main callback
 # =========================
-def run_evaluation(ticker_choice: str, start_sel: str, end_sel: str):
+def run_evaluation(ticker_choice: str, start_sel_txt: str, end_sel_txt: str):
     """
     Evaluate pretrained ARIMA + LSTM on [start_sel, end_sel] inside the fixed universe.
     Show metrics, two plots, and table (Date, Actual, ARIMA, LSTM).
@@ -243,9 +240,12 @@ def run_evaluation(ticker_choice: str, start_sel: str, end_sel: str):
         if not ticker:
             raise ValueError("Please choose a stock.")
 
-        # Parse user-selected dates and clip to universe
-        s_sel = clip_to_universe(datetime.strptime(start_sel, "%Y-%m-%d").date())
-        e_sel = clip_to_universe(datetime.strptime(end_sel,   "%Y-%m-%d").date())
+        # Parse user-entered dates (text boxes) -> YYYY-MM-DD
+        def parse_date(s: str) -> date:
+            return datetime.strptime(s.strip(), "%Y-%m-%d").date()
+
+        s_sel = clip_to_universe(parse_date(start_sel_txt))
+        e_sel = clip_to_universe(parse_date(end_sel_txt))
         if s_sel > e_sel:
             raise ValueError("Start date must be before or equal to end date.")
         # Fetch universe data and build window train/test
@@ -291,7 +291,7 @@ def run_evaluation(ticker_choice: str, start_sel: str, end_sel: str):
         kpi = (
             f"**Ticker:** {ticker}  |  **Window:** {s_sel} → {e_sel}  \n"
             f"**ARIMA {order}** — RMSE: {ar_tbl['RMSE'][0]:.2f} | MAE: {ar_tbl['MAE'][0]:.2f} | MAPE: {ar_tbl['MAPE (%)'][0]:.2f}%  \n"
-            f"**LSTM (lookback={lookback})** — RMSE: {lstm_tbl['RMSE'][0]:.2f} | MAE: {lstm_tbl['MAE'][0]:.2f} | MAPE: {lstm_tbl['MAPE (%)'][0]:.2f}%"
+            f"**LSTM (lookback={lookback})** — RMSE: {lstm_tbl['RMSE'][0]:.2f} | MAE: {lstm_tbl['MAE'][0]:.2f} | MAPE: {lstm_tbl['MAPE (%)'][0]::.2f}%"
         )
 
         return kpi, fig_arima, fig_lstm, table
@@ -316,9 +316,18 @@ with gr.Blocks(title="DataSynthis_ML_JobTask", theme=gr.themes.Soft()) as demo:
     with gr.Row():
         ticker_choice = gr.Dropdown(choices=POPULAR, value="AAPL",
                                     label="Choose a stock")
+
     with gr.Row():
-        start_sel = gr.Date(label="Start date (inside universe)", value=str(UNIVERSE_START))
-        end_sel   = gr.Date(label="End date (inside universe)",   value=str(UNIVERSE_END))
+        start_sel = gr.Textbox(
+            label="Start date (YYYY-MM-DD, inside universe)",
+            value=str(UNIVERSE_START),
+            placeholder="YYYY-MM-DD"
+        )
+        end_sel = gr.Textbox(
+            label="End date (YYYY-MM-DD, inside universe)",
+            value=str(UNIVERSE_END),
+            placeholder="YYYY-MM-DD"
+        )
 
     run_btn = gr.Button("Run Evaluation", variant="primary")