feat: Implement upsert functionality for prediction data, updating existing records or inserting new ones, and remove the previous deletion of all predictions.

src/modules/equipment/Prediksi.py

@@ -239,57 +239,39 @@ class Prediksi:
             #             print(f"HTTP error occurred: {e}")
             #             return {}
 
-            # Menyiapkan data untuk batch insert
+            # Menyiapkan data untuk batch insert atau update
-            # print(f"Data to be inserted: {data}")
             records_to_insert = []
+            
+            check_existence_query = """
+            SELECT id FROM lcc_equipment_tr_data 
+            WHERE assetnum = %s AND tahun = %s AND is_actual = 0
+            """
+
+            update_query = """
+            UPDATE lcc_equipment_tr_data
+            SET 
+                rc_cm_material_cost = %s,
+                rc_cm_labor_cost = %s,
+                rc_pm_material_cost = %s,
+                rc_pm_labor_cost = %s,
+                rc_oh_material_cost = %s,
+                rc_oh_labor_cost = %s,
+                rc_predictive_material_cost = %s,
+                rc_predictive_labor_cost = %s,
+                updated_by = 'Sys', 
+                updated_at = NOW()
+            WHERE id = %s
+            """
+
             for _, row in data.iterrows():
-                max_seq = max_seq + 1
+                # Check if data exists
-                # (token already stored before defining fetch_api_data)
+                cursor.execute(check_existence_query, (equipment_id, int(row["year"])))
-                # maintain previous cm_interval between iterations using attribute on fetch_api_data
+                existing_record = cursor.fetchone()
-                # if not hasattr(fetch_api_data, "prev_cm"):
+
-                #     fetch_api_data.prev_cm = None
+                if existing_record:
-
+                    # Update existing record
-                # Update values from API (current year)
+                    record_id = existing_record[0]
-                # api_data = await fetch_api_data(equipment_id, row["year"])
+                    cursor.execute(update_query, (
-                # if api_data and "data" in api_data and isinstance(api_data["data"], list) and len(api_data["data"]) > 0:
-                #     try:
-                #         cur_cm = float(api_data["data"][0].get("num_fail", row.get("cm_interval", 1)))
-                #     except Exception:
-                #         cur_cm = float(row.get("cm_interval", 1)) if not pd.isna(row.get("cm_interval", None)) else 1.0
-                # else:
-                #     try:
-                #         val = float(row.get("cm_interval", 1))
-                #         cur_cm = val if val >= 1 else 1.0
-                #     except Exception:
-                #         cur_cm = 1.0
-
-                # Determine previous cm_interval: prefer stored prev_cm, otherwise try API for previous year, else fallback to cur_cm
-                # if fetch_api_data.prev_cm is not None:
-                #     prev_cm = float(fetch_api_data.prev_cm)
-                # else:
-                #     try:
-                #         api_prev = await fetch_api_data(equipment_id, int(row["year"]) - 1)
-                #         if api_prev and "data" in api_prev and isinstance(api_prev["data"], list) and len(api_prev["data"]) > 0:
-                #             prev_cm = float(api_prev["data"][0].get("num_fail", cur_cm))
-                #         else:
-                #             # attempt to use any available previous value from the row if present, otherwise fallback to current
-                #             prev_cm = float(row.get("cm_interval", cur_cm)) if not pd.isna(row.get("cm_interval", None)) else cur_cm
-                #     except Exception:
-                #         prev_cm = cur_cm
-
-                # compute difference: current year interval minus previous year interval
-                # try:
-                #     cm_interval_diff = float(cur_cm) - float(prev_cm)
-                # except Exception:
-                #     cm_interval_diff = 0.0
-
-                # append record using the difference for raw_cm_interval
-                records_to_insert.append(
-                    (
-                        str(uuid4()), # id
-                        int(max_seq), # seq
-                        int(row["year"]),
-                        equipment_id,
                         float(row.get("rc_cm_material_cost", 0)) if not pd.isna(row.get("rc_cm_material_cost", 0)) else 0.0,
                         float(row.get("rc_cm_labor_cost", 0)) if not pd.isna(row.get("rc_cm_labor_cost", 0)) else 0.0,
                         float(row.get("rc_pm_material_cost", 0)) if not pd.isna(row.get("rc_pm_material_cost", 0)) else 0.0,
@@ -298,16 +280,36 @@ class Prediksi:
                         float(row.get("rc_oh_labor_cost", 0)) if not pd.isna(row.get("rc_oh_labor_cost", 0)) else 0.0,
                         float(row.get("rc_predictive_material_cost", 0)) if not pd.isna(row.get("rc_predictive_material_cost", 0)) else 0.0,
                         float(row.get("rc_predictive_labor_cost", 0)) if not pd.isna(row.get("rc_predictive_labor_cost", 0)) else 0.0,
+                        record_id
+                    ))
+                else:
+                    max_seq = max_seq + 1
+                    # Prepare for insert
+                    records_to_insert.append(
+                        (
+                            str(uuid4()), # id
+                            int(max_seq), # seq
+                            int(row["year"]),
+                            equipment_id,
+                            float(row.get("rc_cm_material_cost", 0)) if not pd.isna(row.get("rc_cm_material_cost", 0)) else 0.0,
+                            float(row.get("rc_cm_labor_cost", 0)) if not pd.isna(row.get("rc_cm_labor_cost", 0)) else 0.0,
+                            float(row.get("rc_pm_material_cost", 0)) if not pd.isna(row.get("rc_pm_material_cost", 0)) else 0.0,
+                            float(row.get("rc_pm_labor_cost", 0)) if not pd.isna(row.get("rc_pm_labor_cost", 0)) else 0.0,
+                            float(row.get("rc_oh_material_cost", 0)) if not pd.isna(row.get("rc_oh_material_cost", 0)) else 0.0,
+                            float(row.get("rc_oh_labor_cost", 0)) if not pd.isna(row.get("rc_oh_labor_cost", 0)) else 0.0,
+                            float(row.get("rc_predictive_material_cost", 0)) if not pd.isna(row.get("rc_predictive_material_cost", 0)) else 0.0,
+                            float(row.get("rc_predictive_labor_cost", 0)) if not pd.isna(row.get("rc_predictive_labor_cost", 0)) else 0.0,
+                        )
                     )
-                )
-
-                # store current cm for next iteration
-                # fetch_api_data.prev_cm = cur_cm
 
-            # Eksekusi batch insert
+            # Eksekusi batch insert jika ada data baru
-            cursor.executemany(insert_query, records_to_insert)
+            if records_to_insert:
+                cursor.executemany(insert_query, records_to_insert)
+            
             connection.commit()
-            # print("Data proyeksi berhasil dimasukkan ke database.")
+            
+            # Recalculate total costs and update asset criticality
+            self.__update_data_lcc(equipment_id)
 
         except Exception as e:
             print(f"Error saat menyimpan data ke database: {e}")
@@ -930,7 +932,7 @@ class Prediksi:
             predictions_df = pd.DataFrame(predictions)
             # print(predictions_df)
             # Hapus data prediksi yang ada sebelumnya
-            self.__delete_predictions_from_db(assetnum)
+            # self.__delete_predictions_from_db(assetnum)
 
             # Insert hasil prediksi ke database
             try: