""" BIGQUERY ANALYSIS LAYER - FOOD SECURITY AGGREGATION Semua agregasi pakai norm_value dari _get_norm_value_df() UPDATED: Simpan 6 tabel ke fs_asean_gold (layer='gold'): - agg_pillar_composite - agg_pillar_by_country - agg_framework_by_country - agg_framework_asean (+ kolom performance_status: 'Good'/'Bad', threshold=60) - agg_narrative_overview (bilingual: narrative_en, narrative_id) - agg_narrative_pillar (bilingual: narrative_en, narrative_id) Narrative style: - Plain text, tanpa markdown bold (**) - Interpretatif: membaca tren, gap, anomali, konsistensi dari data nyata - Bilingual: narrative_en (Inggris) + narrative_id (Indonesia) - Granularity: per tahun (Overview & Pillar) """ import pandas as pd import numpy as np from datetime import datetime import logging import json import sys as _sys from scripts.bigquery_config import get_bigquery_client from scripts.bigquery_helpers import ( log_update, load_to_bigquery, read_from_bigquery, setup_logging, save_etl_metadata, ) from google.cloud import bigquery # ============================================================================= # KONSTANTA GLOBAL # ============================================================================= DIRECTION_INVERT_KEYWORDS = frozenset({ "negative", "lower_better", "lower_is_better", "inverse", "neg", }) DIRECTION_POSITIVE_KEYWORDS = frozenset({ "positive", "higher_better", "higher_is_better", }) NORMALIZE_FRAMEWORKS_JOINTLY = False PERFORMANCE_THRESHOLD = 60.0 SDG_ONLY_KEYWORDS: frozenset = frozenset([ "prevalence of undernourishment (percent) (3-year average)", "number of people undernourished (million) (3-year average)", "prevalence of severe food insecurity in the total population (percent) (3-year average)", "prevalence of severe food insecurity in the male adult population (percent) (3-year average)", "prevalence of severe food insecurity in the female adult population (percent) (3-year average)", "prevalence of moderate or severe food insecurity in the total population (percent) (3-year average)", "prevalence of moderate or severe food insecurity in the male adult population (percent) (3-year average)", "prevalence of moderate or severe food insecurity in the female adult population (percent) (3-year average)", "number of severely food insecure people (million) (3-year average)", "number of severely food insecure male adults (million) (3-year average)", "number of severely food insecure female adults (million) (3-year average)", "number of moderately or severely food insecure people (million) (3-year average)", "number of moderately or severely food insecure male adults (million) (3-year average)", "number of moderately or severely food insecure female adults (million) (3-year average)", "percentage of children under 5 years of age who are stunted (modelled estimates) (percent)", "number of children under 5 years of age who are stunted (modeled estimates) (million)", "percentage of children under 5 years affected by wasting (percent)", "number of children under 5 years affected by wasting (million)", "percentage of children under 5 years of age who are overweight (modelled estimates) (percent)", "number of children under 5 years of age who are overweight (modeled estimates) (million)", "prevalence of anemia among women of reproductive age (15-49 years) (percent)", "number of women of reproductive age (15-49 years) affected by anemia (million)", ]) _SDG_ONLY_LOWER: frozenset = frozenset(k.lower() for k in SDG_ONLY_KEYWORDS) _FIES_DETECTION_LOWER: frozenset = frozenset([ "prevalence of severe food insecurity in the total population (percent) (3-year average)", "prevalence of moderate or severe food insecurity in the total population (percent) (3-year average)", "number of severely food insecure people (million) (3-year average)", "number of moderately or severely food insecure people (million) (3-year average)", ]) # ============================================================================= # WINDOWS CP1252 SAFE LOGGING # ============================================================================= class _SafeStreamHandler(logging.StreamHandler): def emit(self, record): try: super().emit(record) except UnicodeEncodeError: try: msg = self.format(record) self.stream.write( msg.encode("utf-8", errors="replace").decode("ascii", errors="replace") + self.terminator ) self.flush() except Exception: self.handleError(record) # ============================================================================= # HELPERS # ============================================================================= def _should_invert(direction: str, logger=None, context: str = "") -> bool: d = str(direction).lower().strip() if d in DIRECTION_INVERT_KEYWORDS: return True if d in DIRECTION_POSITIVE_KEYWORDS: return False if logger: logger.warning( f" [DIRECTION WARNING] Unknown direction '{direction}' " f"{'(' + context + ')' if context else ''}. Defaulting to positive (no invert)." ) return False def global_minmax(series: pd.Series, lo: float = 1.0, hi: float = 100.0) -> pd.Series: values = series.dropna().values if len(values) == 0: return pd.Series(np.nan, index=series.index) v_min, v_max = values.min(), values.max() if v_min == v_max: return pd.Series((lo + hi) / 2.0, index=series.index) result = np.full(len(series), np.nan) not_nan = series.notna() raw = series[not_nan].values result[not_nan.values] = lo + (raw - v_min) / (v_max - v_min) * (hi - lo) return pd.Series(result, index=series.index) def add_yoy(df: pd.DataFrame, group_cols: list, score_col: str) -> pd.DataFrame: df = df.sort_values(group_cols + ["year"]).reset_index(drop=True) if group_cols: df["year_over_year_change"] = df.groupby(group_cols)[score_col].diff() else: df["year_over_year_change"] = df[score_col].diff() return df def safe_int(series: pd.Series, fill: int = 0, col_name: str = "", logger=None) -> pd.Series: n_nan = series.isna().sum() if n_nan > 0 and logger: logger.warning( f" [NaN WARNING] Kolom '{col_name}' punya {n_nan} NaN -> di-fill dengan {fill}" ) return series.fillna(fill).astype(int) def check_and_dedup(df: pd.DataFrame, key_cols: list, context: str = "", logger=None) -> pd.DataFrame: dupes = df.duplicated(subset=key_cols, keep=False) if dupes.any(): n_dupes = dupes.sum() if logger: logger.warning( f" [DEDUP WARNING] {context}: {n_dupes} duplikat rows pada {key_cols}. " f"Di-aggregate dengan mean." ) numeric_cols = df.select_dtypes(include=[np.number]).columns.tolist() agg_dict = { c: ("mean" if c in numeric_cols else "first") for c in df.columns if c not in key_cols } df = df.groupby(key_cols, as_index=False).agg(agg_dict) return df def _performance_status(score) -> str: if score is None or (isinstance(score, float) and np.isnan(score)): return "N/A" return "Good" if score >= PERFORMANCE_THRESHOLD else "Bad" def _fmt_score(score) -> str: if score is None or (isinstance(score, float) and np.isnan(score)): return "N/A" return f"{score:.2f}" def _fmt_delta(delta) -> str: if delta is None or (isinstance(delta, float) and np.isnan(delta)): return "N/A" sign = "+" if delta >= 0 else "" return f"{sign}{delta:.2f}" # ============================================================================= # NARRATIVE CONDITION DETECTORS (shared) # ============================================================================= def _detect_series_trend(scores: list) -> str: """ Deteksi tren dari list skor berurutan. Return: 'improving_consistent' | 'improving_slowing' | 'deteriorating' | 'fluctuating' """ if len(scores) < 3: return "insufficient" x = np.arange(len(scores)) slope = np.polyfit(x, scores, 1)[0] cv = np.std(scores) / (np.mean(scores) + 1e-9) if cv > 0.20: return "fluctuating" mid = len(scores) // 2 slope1 = np.polyfit(np.arange(mid), scores[:mid], 1)[0] if mid > 1 else slope slope2 = np.polyfit(np.arange(len(scores) - mid), scores[mid:], 1)[0] if (len(scores) - mid) > 1 else slope if slope > 0: slowing = slope2 < slope1 return "improving_slowing" if slowing else "improving_consistent" else: return "deteriorating" def _detect_country_gap(scores_by_country_year: pd.DataFrame, score_col: str) -> str: """ Deteksi apakah std antar negara melebar atau menyempit dari waktu ke waktu. scores_by_country_year: df dengan kolom [year, country_id, score_col] """ std_by_year = ( scores_by_country_year.groupby("year")[score_col] .std().dropna() ) if len(std_by_year) < 3: return "unknown" years = sorted(std_by_year.index) stds = [std_by_year[y] for y in years] slope = np.polyfit(np.arange(len(stds)), stds, 1)[0] mean_s = np.mean(stds) if abs(slope) < 0.02 * mean_s: return "stable" return "widening" if slope > 0 else "narrowing" def _find_anomaly_year(values_by_year: dict) -> tuple: """ Cari tahun dengan perubahan YoY paling ekstrem. values_by_year: {year: score} Return: (year, 'drop' | 'rise') atau (None, None) """ years = sorted(values_by_year.keys()) deltas = {} for i in range(1, len(years)): y0, y1 = years[i-1], years[i] v0, v1 = values_by_year.get(y0), values_by_year.get(y1) if v0 is not None and v1 is not None and not (pd.isna(v0) or pd.isna(v1)): deltas[y1] = v1 - v0 if not deltas: return None, None threshold = 1.5 * np.std(list(deltas.values())) min_y = min(deltas, key=deltas.get) max_y = max(deltas, key=deltas.get) if abs(deltas[min_y]) > threshold and deltas[min_y] < 0: return min_y, "drop" if abs(deltas[max_y]) > threshold and deltas[max_y] > 0: return max_y, "rise" return None, None # ============================================================================= # NARRATIVE BUILDER — OVERVIEW (per tahun) # ============================================================================= def _build_overview_narrative( year: int, score: float, performance_status: str, yoy_val, n_mdg: int, n_sdg: int, ranking_list: list, most_improved_country, most_improved_delta, most_declined_country, most_declined_delta, historical_scores: dict, # {year: score} semua tahun sebelumnya country_scores_all: pd.DataFrame, # df [year, country_name, framework_score_1_100] ) -> tuple: """ Narasi overview per tahun — interpretatif, plain text, bilingual. Return: (narrative_en, narrative_id) """ sentences_en = [] sentences_id = [] # ---- 1. Status tahun ini vs threshold ---- perf_word_en = "good" if performance_status == "Good" else "below target" perf_word_id = "baik" if performance_status == "Good" else "di bawah target" s1_en = ( f"In {year}, ASEAN food security scored {_fmt_score(score)} out of 100 " f"({perf_word_en}), covering {n_mdg + n_sdg} indicators " f"({n_mdg} MDGs and {n_sdg} SDGs)." ) s1_id = ( f"Pada tahun {year}, skor ketahanan pangan ASEAN mencapai {_fmt_score(score)} dari 100 " f"({perf_word_id}), mencakup {n_mdg + n_sdg} indikator " f"({n_mdg} MDGs dan {n_sdg} SDGs)." ) sentences_en.append(s1_en) sentences_id.append(s1_id) # ---- 2. Kondisi YoY tahun ini ---- if yoy_val is not None and not pd.isna(yoy_val): if abs(yoy_val) < 0.5: s2_en = f"The score was relatively stable compared to the previous year." s2_id = f"Skor relatif stabil dibandingkan tahun sebelumnya." elif yoy_val > 0: s2_en = f"This represents an improvement of {abs(yoy_val):.2f} points from the previous year." s2_id = f"Ini merupakan peningkatan sebesar {abs(yoy_val):.2f} poin dari tahun sebelumnya." else: s2_en = f"This represents a decline of {abs(yoy_val):.2f} points from the previous year." s2_id = f"Ini merupakan penurunan sebesar {abs(yoy_val):.2f} poin dari tahun sebelumnya." sentences_en.append(s2_en) sentences_id.append(s2_id) # ---- 3. Tren historis (baca dari semua data yang ada) ---- hist_years = sorted(historical_scores.keys()) hist_scores = [historical_scores[y] for y in hist_years if not pd.isna(historical_scores.get(y, np.nan))] if len(hist_scores) >= 3: trend = _detect_series_trend(hist_scores) if trend == "improving_consistent": s3_en = f"The overall trajectory since {hist_years[0]} has been consistently upward." s3_id = f"Trajektori keseluruhan sejak {hist_years[0]} menunjukkan tren yang konsisten meningkat." elif trend == "improving_slowing": s3_en = f"While the long-term trend since {hist_years[0]} is positive, the pace of improvement has slowed in recent years." s3_id = f"Meskipun tren jangka panjang sejak {hist_years[0]} positif, laju perbaikan melambat dalam beberapa tahun terakhir." elif trend == "deteriorating": s3_en = f"The overall trend since {hist_years[0]} shows a declining trajectory that warrants attention." s3_id = f"Tren keseluruhan sejak {hist_years[0]} menunjukkan trajektori yang menurun dan perlu perhatian." elif trend == "fluctuating": s3_en = f"Progress since {hist_years[0]} has been uneven, with scores fluctuating across years." s3_id = f"Kemajuan sejak {hist_years[0]} tidak merata, dengan skor yang berfluktuasi antar tahun." else: s3_en = "" s3_id = "" if s3_en: sentences_en.append(s3_en) sentences_id.append(s3_id) # ---- 4. Gap antar negara ---- if not country_scores_all.empty: gap_trend = _detect_country_gap( country_scores_all[country_scores_all["year"] <= year], "framework_score_1_100" ) if gap_trend == "widening": s4_en = "The performance gap among ASEAN member states has widened over time, indicating unequal progress." s4_id = "Kesenjangan performa antar negara anggota ASEAN semakin melebar, mengindikasikan kemajuan yang tidak merata." elif gap_trend == "narrowing": s4_en = "The performance gap among ASEAN member states has narrowed, reflecting more balanced regional development." s4_id = "Kesenjangan performa antar negara anggota ASEAN semakin menyempit, mencerminkan pembangunan regional yang lebih merata." elif gap_trend == "stable": s4_en = "The performance gap among ASEAN member states has remained relatively stable." s4_id = "Kesenjangan performa antar negara anggota ASEAN relatif stabil." else: s4_en = "" s4_id = "" if s4_en: sentences_en.append(s4_en) sentences_id.append(s4_id) # ---- 5. Top dan bottom country tahun ini ---- if ranking_list and len(ranking_list) >= 2: top = ranking_list[0] bottom = ranking_list[-1] s5_en = ( f"In {year}, {top['country_name']} led the region with a score of " f"{_fmt_score(top['score'])}, while {bottom['country_name']} ranked last " f"at {_fmt_score(bottom['score'])}." ) s5_id = ( f"Pada tahun {year}, {top['country_name']} memimpin kawasan dengan skor " f"{_fmt_score(top['score'])}, sementara {bottom['country_name']} berada di " f"posisi terbawah dengan skor {_fmt_score(bottom['score'])}." ) sentences_en.append(s5_en) sentences_id.append(s5_id) # ---- 6. Most improved / declined country ---- if most_improved_country and most_declined_country: if most_improved_country != most_declined_country: s6_en = ( f"{most_improved_country} showed the biggest improvement " f"({_fmt_delta(most_improved_delta)} pts), " f"while {most_declined_country} experienced the largest decline " f"({_fmt_delta(most_declined_delta)} pts)." ) s6_id = ( f"{most_improved_country} mencatat peningkatan terbesar " f"({_fmt_delta(most_improved_delta)} poin), " f"sementara {most_declined_country} mengalami penurunan terbesar " f"({_fmt_delta(most_declined_delta)} poin)." ) sentences_en.append(s6_en) sentences_id.append(s6_id) narrative_en = " ".join(s for s in sentences_en if s) narrative_id = " ".join(s for s in sentences_id if s) return narrative_en, narrative_id # ============================================================================= # NARRATIVE BUILDER — PILLAR (per tahun per pilar) # ============================================================================= def _build_pillar_narrative( year: int, pillar_name: str, pillar_score: float, rank_in_year: int, n_pillars: int, yoy_val, top_country: str, top_country_score, bot_country: str, bot_country_score, pillar_scores_history: dict, # {year: score} untuk pilar ini all_pillar_scores_year: pd.DataFrame, # df [pillar_name, pillar_score_1_100] tahun ini country_pillar_all: pd.DataFrame, # df [year, country_id, pillar_country_score_1_100] pilar ini ) -> tuple: """ Narasi pillar per tahun — interpretatif, plain text, bilingual. Return: (narrative_en, narrative_id) """ sentences_en = [] sentences_id = [] # ---- 1. Posisi pilar tahun ini ---- rank_suffix = {1: "st", 2: "nd", 3: "rd"}.get(rank_in_year, "th") perf_word_en = "good" if pillar_score >= PERFORMANCE_THRESHOLD else "below target" perf_word_id = "baik" if pillar_score >= PERFORMANCE_THRESHOLD else "di bawah target" s1_en = ( f"In {year}, the {pillar_name} pillar ranked {rank_in_year}{rank_suffix} out of " f"{n_pillars} pillars with a score of {_fmt_score(pillar_score)} ({perf_word_en})." ) s1_id = ( f"Pada tahun {year}, pilar {pillar_name} menempati peringkat {rank_in_year} dari " f"{n_pillars} pilar dengan skor {_fmt_score(pillar_score)} ({perf_word_id})." ) sentences_en.append(s1_en) sentences_id.append(s1_id) # ---- 2. YoY pilar ini ---- if yoy_val is not None and not pd.isna(yoy_val): if abs(yoy_val) < 0.5: s2_en = "Performance was relatively stable compared to the previous year." s2_id = "Performa relatif stabil dibandingkan tahun sebelumnya." elif yoy_val > 0: s2_en = f"This is an improvement of {abs(yoy_val):.2f} points from the previous year." s2_id = f"Ini merupakan peningkatan {abs(yoy_val):.2f} poin dari tahun sebelumnya." else: s2_en = f"This marks a decline of {abs(yoy_val):.2f} points from the previous year." s2_id = f"Ini menandai penurunan {abs(yoy_val):.2f} poin dari tahun sebelumnya." sentences_en.append(s2_en) sentences_id.append(s2_id) # ---- 3. Tren historis pilar ini ---- hist_years = sorted(pillar_scores_history.keys()) hist_scores = [ pillar_scores_history[y] for y in hist_years if not pd.isna(pillar_scores_history.get(y, np.nan)) ] if len(hist_scores) >= 3: trend = _detect_series_trend(hist_scores) if trend == "improving_consistent": s3_en = f"This pillar has shown consistent improvement since {hist_years[0]}." s3_id = f"Pilar ini menunjukkan perbaikan yang konsisten sejak {hist_years[0]}." elif trend == "improving_slowing": s3_en = f"While the pillar improved since {hist_years[0]}, the pace has slowed in recent years." s3_id = f"Meskipun pilar ini membaik sejak {hist_years[0]}, lajunya melambat dalam beberapa tahun terakhir." elif trend == "deteriorating": s3_en = f"This pillar has shown a declining trend since {hist_years[0]}, requiring targeted intervention." s3_id = f"Pilar ini menunjukkan tren penurunan sejak {hist_years[0]}, memerlukan intervensi yang terarah." elif trend == "fluctuating": s3_en = f"Performance in this pillar has been inconsistent since {hist_years[0]}, with no clear trend." s3_id = f"Performa pilar ini tidak konsisten sejak {hist_years[0]}, tanpa tren yang jelas." else: s3_en = "" s3_id = "" if s3_en: sentences_en.append(s3_en) sentences_id.append(s3_id) # ---- 4. Gap antar negara dalam pilar ini ---- if not country_pillar_all.empty: gap_trend = _detect_country_gap( country_pillar_all[country_pillar_all["year"] <= year], "pillar_country_score_1_100" ) if gap_trend == "widening": s4_en = "Country disparities within this pillar have widened over time." s4_id = "Kesenjangan antar negara dalam pilar ini semakin melebar seiring waktu." elif gap_trend == "narrowing": s4_en = "Country disparities within this pillar have narrowed, indicating more balanced progress." s4_id = "Kesenjangan antar negara dalam pilar ini menyempit, mengindikasikan kemajuan yang lebih merata." else: s4_en = "" s4_id = "" if s4_en: sentences_en.append(s4_en) sentences_id.append(s4_id) # ---- 5. Top/bottom country dalam pilar ini ---- if top_country and bot_country and top_country != bot_country: s5_en = ( f"{top_country} performed best in this pillar ({_fmt_score(top_country_score)}), " f"while {bot_country} had the lowest score ({_fmt_score(bot_country_score)})." ) s5_id = ( f"{top_country} memiliki performa terbaik dalam pilar ini ({_fmt_score(top_country_score)}), " f"sementara {bot_country} memiliki skor terendah ({_fmt_score(bot_country_score)})." ) sentences_en.append(s5_en) sentences_id.append(s5_id) # ---- 6. Posisi relatif pilar ini vs pilar lain ---- if not all_pillar_scores_year.empty and len(all_pillar_scores_year) > 1: sorted_pillars = all_pillar_scores_year.sort_values("pillar_score_1_100", ascending=False) strongest = sorted_pillars.iloc[0] weakest = sorted_pillars.iloc[-1] if strongest["pillar_name"] != pillar_name and weakest["pillar_name"] != pillar_name: s6_en = ( f"Across all pillars in {year}, {strongest['pillar_name']} scored highest " f"({_fmt_score(strongest['pillar_score_1_100'])}) and {weakest['pillar_name']} " f"scored lowest ({_fmt_score(weakest['pillar_score_1_100'])})." ) s6_id = ( f"Di antara semua pilar pada tahun {year}, {strongest['pillar_name']} mendapat skor " f"tertinggi ({_fmt_score(strongest['pillar_score_1_100'])}) dan {weakest['pillar_name']} " f"mendapat skor terendah ({_fmt_score(weakest['pillar_score_1_100'])})." ) sentences_en.append(s6_en) sentences_id.append(s6_id) narrative_en = " ".join(s for s in sentences_en if s) narrative_id = " ".join(s for s in sentences_id if s) return narrative_en, narrative_id # ============================================================================= # MAIN CLASS # ============================================================================= class FoodSecurityAggregator: def __init__(self, client: bigquery.Client): self.client = client self.logger = logging.getLogger(self.__class__.__name__) self.logger.propagate = False self.load_metadata = { "agg_pillar_composite": {"rows_loaded": 0, "status": "pending", "start_time": None, "end_time": None}, "agg_pillar_by_country": {"rows_loaded": 0, "status": "pending", "start_time": None, "end_time": None}, "agg_framework_by_country": {"rows_loaded": 0, "status": "pending", "start_time": None, "end_time": None}, "agg_framework_asean": {"rows_loaded": 0, "status": "pending", "start_time": None, "end_time": None}, "agg_narrative_overview": {"rows_loaded": 0, "status": "pending", "start_time": None, "end_time": None}, "agg_narrative_pillar": {"rows_loaded": 0, "status": "pending", "start_time": None, "end_time": None}, } self.df = None self.sdgs_start_year = None self._ind_year_framework: pd.DataFrame = None # ========================================================================= # STEP 1: Load data # ========================================================================= def load_data(self): self.logger.info("=" * 70) self.logger.info("STEP 1: LOAD DATA from fs_asean_gold") self.logger.info("=" * 70) self.df = read_from_bigquery( self.client, "fact_asean_food_security_selected", layer='gold' ) self.logger.info(f" fact_asean_food_security_selected : {len(self.df):,} rows") required_cols = { "country_id", "country_name", "indicator_id", "indicator_name", "direction", "pillar_id", "pillar_name", "time_id", "year", "value", } missing_cols = required_cols - set(self.df.columns) if missing_cols: raise ValueError( f"Kolom berikut tidak ditemukan: {missing_cols}" ) n_null_dir = self.df["direction"].isna().sum() if n_null_dir > 0: self.logger.warning(f" [DIRECTION] {n_null_dir} rows NULL -> diisi 'positive'") self.df["direction"] = self.df["direction"].fillna("positive") self.logger.info(f" Rows : {len(self.df):,}") self.logger.info(f" Countries : {self.df['country_id'].nunique()}") self.logger.info(f" Indicators: {self.df['indicator_id'].nunique()}") self.logger.info( f" Years : {int(self.df['year'].min())} - {int(self.df['year'].max())}" ) # ========================================================================= # STEP 1b: Detect sdgs_start_year + assign framework # ========================================================================= def _detect_sdgs_start_year(self) -> int: fies_rows = self.df[ self.df["indicator_name"].str.lower().str.strip().isin(_FIES_DETECTION_LOWER) ] if not fies_rows.empty: sdgs_start = int(fies_rows["year"].min()) self.logger.info(f" [FIES explicit] sdgs_start_year = {sdgs_start}") return sdgs_start ind_min_year = ( self.df.groupby("indicator_id")["year"] .min().reset_index().rename(columns={"year": "min_year"}) ) unique_years = sorted(ind_min_year["min_year"].unique()) if len(unique_years) == 1: return int(unique_years[0]) + 9999 gaps = [ (unique_years[i+1] - unique_years[i], unique_years[i], unique_years[i+1]) for i in range(len(unique_years) - 1) ] gaps.sort(reverse=True) _, y_before, y_after = gaps[0] self.logger.info(f" [Fallback gap] sdgs_start_year = {y_after}") return int(y_after) def _assign_framework_labels(self): self.logger.info("\n" + "=" * 70) self.logger.info("STEP 1b: ASSIGN FRAMEWORK LABELS") self.logger.info(f" sdgs_start_year = {self.sdgs_start_year}") self.logger.info("=" * 70) df = self.df.copy() df["_is_sdg_kw"] = df["indicator_name"].str.lower().str.strip().isin(_SDG_ONLY_LOWER) df["framework"] = "MDGs" mask_sdgs = df["_is_sdg_kw"] & (df["year"] >= self.sdgs_start_year) df.loc[mask_sdgs, "framework"] = "SDGs" df = df.drop(columns=["_is_sdg_kw"]) self.df = df self._ind_year_framework = ( self.df[["indicator_id", "year", "framework"]] .drop_duplicates() .reset_index(drop=True) ) fw_dist = self.df["framework"].value_counts() self.logger.info("\n Framework distribution (rows):") for fw, cnt in fw_dist.items(): self.logger.info(f" {fw:<6}: {cnt:,} rows") def _count_framework_indicators(self, year: int, framework: str) -> int: mask = ( (self._ind_year_framework["year"] == year) & (self._ind_year_framework["framework"] == framework) ) return int(self._ind_year_framework.loc[mask, "indicator_id"].nunique()) # ========================================================================= # CORE HELPER: normalisasi raw value per indikator # ========================================================================= def _get_norm_value_df(self) -> pd.DataFrame: if "framework" not in self.df.columns: raise ValueError("Kolom 'framework' tidak ada.") norm_parts = [] for ind_id, grp in self.df.groupby("indicator_id"): grp = grp.copy() direction = str(grp["direction"].iloc[0]) do_invert = _should_invert(direction, self.logger, context=f"indicator_id={ind_id}") valid_mask = grp["value"].notna() n_valid = valid_mask.sum() if n_valid < 2: grp["norm_value"] = np.nan norm_parts.append(grp) continue raw = grp.loc[valid_mask, "value"].values v_min, v_max = raw.min(), raw.max() normed = np.full(len(grp), np.nan) if v_min == v_max: normed[valid_mask.values] = 0.5 else: normed[valid_mask.values] = (raw - v_min) / (v_max - v_min) if do_invert: normed = np.where(np.isnan(normed), np.nan, 1.0 - normed) grp["norm_value"] = normed norm_parts.append(grp) return pd.concat(norm_parts, ignore_index=True) # ========================================================================= # METADATA BUILDER # ========================================================================= def _build_etl_metadata( self, table_name: str, rows_loaded: int, start_time: datetime, end_time: datetime, status: str, error_msg: str = None, ) -> dict: duration = (end_time - start_time).total_seconds() if (start_time and end_time) else 0.0 return { "source_class" : "FoodSecurityAggregator", "table_name" : table_name, "execution_timestamp": start_time or end_time, "duration_seconds" : round(duration, 4), "rows_fetched" : rows_loaded, "rows_transformed" : rows_loaded, "rows_loaded" : rows_loaded, "completeness_pct" : 100.0 if status == "success" else 0.0, "config_snapshot" : json.dumps({ "layer" : "gold", "write_disposition" : "WRITE_TRUNCATE", "normalize_frameworks_jointly": NORMALIZE_FRAMEWORKS_JOINTLY, "performance_threshold" : PERFORMANCE_THRESHOLD, "status" : status, }), "validation_metrics" : json.dumps({ "status" : status, "error_msg": error_msg or "", }), } # ========================================================================= # STEP 2: agg_pillar_composite # ========================================================================= def calc_pillar_composite(self) -> pd.DataFrame: table_name = "agg_pillar_composite" self.load_metadata[table_name]["start_time"] = datetime.now() self.logger.info("\n" + "=" * 70) self.logger.info(f"STEP 2: {table_name} -> [Gold] fs_asean_gold") self.logger.info("=" * 70) try: df_normed = self._get_norm_value_df() df = ( df_normed .groupby(["pillar_id", "pillar_name", "year"]) .agg( pillar_norm =("norm_value", "mean"), n_indicators=("indicator_id", "nunique"), n_countries =("country_id", "nunique"), ) .reset_index() ) df["pillar_score_1_100"] = global_minmax(df["pillar_norm"]) df["rank_in_year"] = ( df.groupby("year")["pillar_score_1_100"] .rank(method="min", ascending=False) .astype(int) ) df = add_yoy(df, ["pillar_id"], "pillar_score_1_100") df["pillar_id"] = df["pillar_id"].astype(int) df["year"] = df["year"].astype(int) df["n_indicators"] = safe_int(df["n_indicators"], col_name="n_indicators", logger=self.logger) df["n_countries"] = safe_int(df["n_countries"], col_name="n_countries", logger=self.logger) df["rank_in_year"] = df["rank_in_year"].astype(int) df["pillar_norm"] = df["pillar_norm"].astype(float) df["pillar_score_1_100"] = df["pillar_score_1_100"].astype(float) schema = [ bigquery.SchemaField("pillar_id", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("pillar_name", "STRING", mode="REQUIRED"), bigquery.SchemaField("year", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("pillar_norm", "FLOAT", mode="REQUIRED"), bigquery.SchemaField("n_indicators", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("n_countries", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("pillar_score_1_100", "FLOAT", mode="REQUIRED"), bigquery.SchemaField("rank_in_year", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("year_over_year_change", "FLOAT", mode="NULLABLE"), ] rows = load_to_bigquery( self.client, df, table_name, layer='gold', write_disposition="WRITE_TRUNCATE", schema=schema ) self._finalize(table_name, rows) return df except Exception as e: self._fail(table_name, e) raise # ========================================================================= # STEP 3: agg_pillar_by_country # ========================================================================= def calc_pillar_by_country(self) -> pd.DataFrame: table_name = "agg_pillar_by_country" self.load_metadata[table_name]["start_time"] = datetime.now() self.logger.info("\n" + "=" * 70) self.logger.info(f"STEP 3: {table_name} -> [Gold] fs_asean_gold") self.logger.info("=" * 70) try: df_normed = self._get_norm_value_df() df = ( df_normed .groupby(["country_id", "country_name", "pillar_id", "pillar_name", "year"]) .agg(pillar_country_norm=("norm_value", "mean")) .reset_index() ) df["pillar_country_score_1_100"] = global_minmax(df["pillar_country_norm"]) df["rank_in_pillar_year"] = ( df.groupby(["pillar_id", "year"])["pillar_country_score_1_100"] .rank(method="min", ascending=False) .astype(int) ) df = add_yoy(df, ["country_id", "pillar_id"], "pillar_country_score_1_100") df["country_id"] = df["country_id"].astype(int) df["pillar_id"] = df["pillar_id"].astype(int) df["year"] = df["year"].astype(int) df["rank_in_pillar_year"] = df["rank_in_pillar_year"].astype(int) df["pillar_country_norm"] = df["pillar_country_norm"].astype(float) df["pillar_country_score_1_100"] = df["pillar_country_score_1_100"].astype(float) schema = [ bigquery.SchemaField("country_id", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("country_name", "STRING", mode="REQUIRED"), bigquery.SchemaField("pillar_id", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("pillar_name", "STRING", mode="REQUIRED"), bigquery.SchemaField("year", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("pillar_country_norm", "FLOAT", mode="REQUIRED"), bigquery.SchemaField("pillar_country_score_1_100", "FLOAT", mode="REQUIRED"), bigquery.SchemaField("rank_in_pillar_year", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("year_over_year_change", "FLOAT", mode="NULLABLE"), ] rows = load_to_bigquery( self.client, df, table_name, layer='gold', write_disposition="WRITE_TRUNCATE", schema=schema ) self._finalize(table_name, rows) return df except Exception as e: self._fail(table_name, e) raise # ========================================================================= # STEP 4: agg_framework_by_country # ========================================================================= def _calc_country_composite_inmemory(self) -> pd.DataFrame: df_normed = self._get_norm_value_df() df = ( df_normed .groupby(["country_id", "country_name", "year"]) .agg( composite_score=("norm_value", "mean"), n_indicators =("indicator_id", "nunique"), ) .reset_index() ) df["score_1_100"] = global_minmax(df["composite_score"]) df["rank_in_asean"] = ( df.groupby("year")["score_1_100"] .rank(method="min", ascending=False) .astype(int) ) df = add_yoy(df, ["country_id"], "score_1_100") df["country_id"] = df["country_id"].astype(int) df["year"] = df["year"].astype(int) df["n_indicators"] = safe_int(df["n_indicators"], col_name="n_indicators", logger=self.logger) df["composite_score"] = df["composite_score"].astype(float) df["score_1_100"] = df["score_1_100"].astype(float) df["rank_in_asean"] = df["rank_in_asean"].astype(int) return df def calc_framework_by_country(self) -> pd.DataFrame: table_name = "agg_framework_by_country" self.load_metadata[table_name]["start_time"] = datetime.now() self.logger.info("\n" + "=" * 70) self.logger.info(f"STEP 4: {table_name} -> [Gold] fs_asean_gold") self.logger.info("=" * 70) try: country_composite = self._calc_country_composite_inmemory() df_normed = self._get_norm_value_df() parts = [] agg_total = ( country_composite[[ "country_id", "country_name", "year", "score_1_100", "n_indicators", "composite_score" ]] .copy() .rename(columns={ "score_1_100" : "framework_score_1_100", "composite_score": "framework_norm", }) ) agg_total["framework"] = "Total" parts.append(agg_total) pre_sdgs_rows = country_composite[country_composite["year"] < self.sdgs_start_year].copy() if not pre_sdgs_rows.empty: mdgs_pre = ( pre_sdgs_rows[[ "country_id", "country_name", "year", "score_1_100", "n_indicators", "composite_score" ]] .copy() .rename(columns={ "score_1_100" : "framework_score_1_100", "composite_score": "framework_norm", }) ) mdgs_pre["framework"] = "MDGs" parts.append(mdgs_pre) mdgs_indicator_ids = set( self._ind_year_framework[self._ind_year_framework["framework"] == "MDGs"]["indicator_id"] ) if mdgs_indicator_ids: df_mdgs_mixed = df_normed[ (df_normed["indicator_id"].isin(mdgs_indicator_ids)) & (df_normed["year"] >= self.sdgs_start_year) ].copy() if not df_mdgs_mixed.empty: agg_mdgs_mixed = ( df_mdgs_mixed .groupby(["country_id", "country_name", "year"]) .agg( framework_norm=("norm_value", "mean"), n_indicators =("indicator_id", "nunique"), ) .reset_index() ) if not NORMALIZE_FRAMEWORKS_JOINTLY: agg_mdgs_mixed["framework_score_1_100"] = global_minmax(agg_mdgs_mixed["framework_norm"]) agg_mdgs_mixed["framework"] = "MDGs" parts.append(agg_mdgs_mixed) sdgs_indicator_ids = set( self._ind_year_framework[self._ind_year_framework["framework"] == "SDGs"]["indicator_id"] ) if sdgs_indicator_ids: df_sdgs = df_normed[ (df_normed["indicator_id"].isin(sdgs_indicator_ids)) & (df_normed["year"] >= self.sdgs_start_year) ].copy() if not df_sdgs.empty: agg_sdgs = ( df_sdgs .groupby(["country_id", "country_name", "year"]) .agg( framework_norm=("norm_value", "mean"), n_indicators =("indicator_id", "nunique"), ) .reset_index() ) if not NORMALIZE_FRAMEWORKS_JOINTLY: agg_sdgs["framework_score_1_100"] = global_minmax(agg_sdgs["framework_norm"]) agg_sdgs["framework"] = "SDGs" parts.append(agg_sdgs) df = pd.concat(parts, ignore_index=True) if NORMALIZE_FRAMEWORKS_JOINTLY: mixed_mask = (df["framework"].isin(["MDGs", "SDGs"])) & (df["year"] >= self.sdgs_start_year) if mixed_mask.any(): df.loc[mixed_mask, "framework_score_1_100"] = global_minmax(df.loc[mixed_mask, "framework_norm"]) df = check_and_dedup(df, ["country_id", "framework", "year"], context=table_name, logger=self.logger) df["rank_in_framework_year"] = ( df.groupby(["framework", "year"])["framework_score_1_100"] .rank(method="min", ascending=False) .astype(int) ) df = add_yoy(df, ["country_id", "framework"], "framework_score_1_100") df["country_id"] = df["country_id"].astype(int) df["year"] = df["year"].astype(int) df["n_indicators"] = safe_int(df["n_indicators"], col_name="n_indicators", logger=self.logger) df["rank_in_framework_year"] = safe_int(df["rank_in_framework_year"], col_name="rank_in_framework_year", logger=self.logger) df["framework_norm"] = df["framework_norm"].astype(float) df["framework_score_1_100"] = df["framework_score_1_100"].astype(float) self._validate_mdgs_equals_total(df, level="country") schema = [ bigquery.SchemaField("country_id", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("country_name", "STRING", mode="REQUIRED"), bigquery.SchemaField("year", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("framework", "STRING", mode="REQUIRED"), bigquery.SchemaField("n_indicators", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("framework_norm", "FLOAT", mode="REQUIRED"), bigquery.SchemaField("framework_score_1_100", "FLOAT", mode="REQUIRED"), bigquery.SchemaField("rank_in_framework_year", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("year_over_year_change", "FLOAT", mode="NULLABLE"), ] rows = load_to_bigquery( self.client, df, table_name, layer='gold', write_disposition="WRITE_TRUNCATE", schema=schema ) self._finalize(table_name, rows) return df except Exception as e: self._fail(table_name, e) raise # ========================================================================= # STEP 5: agg_framework_asean # ========================================================================= def calc_framework_asean(self) -> pd.DataFrame: table_name = "agg_framework_asean" self.load_metadata[table_name]["start_time"] = datetime.now() self.logger.info("\n" + "=" * 70) self.logger.info(f"STEP 5: {table_name} -> [Gold] fs_asean_gold") self.logger.info("=" * 70) try: df_normed = self._get_norm_value_df() country_composite = self._calc_country_composite_inmemory() country_norm = ( df_normed .groupby(["country_id", "country_name", "year"])["norm_value"] .mean().reset_index() .rename(columns={"norm_value": "country_norm"}) ) asean_overall = ( country_norm.groupby("year") .agg( asean_norm =("country_norm", "mean"), std_norm =("country_norm", "std"), n_countries=("country_norm", "count"), ) .reset_index() ) asean_overall["asean_score_1_100"] = global_minmax(asean_overall["asean_norm"]) asean_comp = ( country_composite.groupby("year")["composite_score"] .mean().reset_index() .rename(columns={"composite_score": "asean_composite"}) ) asean_overall = asean_overall.merge(asean_comp, on="year", how="left") parts = [] def _n_ind(year_val, framework_val): return self._count_framework_indicators(year_val, framework_val) total_cols = asean_overall[[ "year", "asean_score_1_100", "asean_norm", "std_norm", "n_countries" ]].copy().rename(columns={ "asean_score_1_100": "framework_score_1_100", "asean_norm" : "framework_norm", "n_countries" : "n_countries_with_data", }) total_cols["n_indicators"] = total_cols["year"].apply( lambda y: int(self._ind_year_framework[ self._ind_year_framework["year"] == y ]["indicator_id"].nunique()) ) total_cols["framework"] = "Total" parts.append(total_cols) pre_sdgs = asean_overall[asean_overall["year"] < self.sdgs_start_year].copy() if not pre_sdgs.empty: mdgs_pre = pre_sdgs[[ "year", "asean_score_1_100", "asean_norm", "std_norm", "n_countries" ]].copy().rename(columns={ "asean_score_1_100": "framework_score_1_100", "asean_norm" : "framework_norm", "n_countries" : "n_countries_with_data", }) mdgs_pre["n_indicators"] = mdgs_pre["year"].apply(lambda y: _n_ind(y, "MDGs")) mdgs_pre["framework"] = "MDGs" parts.append(mdgs_pre) mdgs_indicator_ids = set( self._ind_year_framework[self._ind_year_framework["framework"] == "MDGs"]["indicator_id"] ) if mdgs_indicator_ids: df_mdgs_mixed = df_normed[ (df_normed["indicator_id"].isin(mdgs_indicator_ids)) & (df_normed["year"] >= self.sdgs_start_year) ].copy() if not df_mdgs_mixed.empty: cn = ( df_mdgs_mixed .groupby(["country_id", "year"])["norm_value"].mean() .reset_index().rename(columns={"norm_value": "country_norm"}) ) asean_mdgs = cn.groupby("year").agg( framework_norm =("country_norm", "mean"), std_norm =("country_norm", "std"), n_countries_with_data=("country_id", "count"), ).reset_index() asean_mdgs["n_indicators"] = asean_mdgs["year"].apply(lambda y: _n_ind(y, "MDGs")) if not NORMALIZE_FRAMEWORKS_JOINTLY: asean_mdgs["framework_score_1_100"] = global_minmax(asean_mdgs["framework_norm"]) asean_mdgs["framework"] = "MDGs" parts.append(asean_mdgs) sdgs_indicator_ids = set( self._ind_year_framework[self._ind_year_framework["framework"] == "SDGs"]["indicator_id"] ) if sdgs_indicator_ids: df_sdgs = df_normed[ (df_normed["indicator_id"].isin(sdgs_indicator_ids)) & (df_normed["year"] >= self.sdgs_start_year) ].copy() if not df_sdgs.empty: cn = ( df_sdgs .groupby(["country_id", "year"])["norm_value"].mean() .reset_index().rename(columns={"norm_value": "country_norm"}) ) asean_sdgs = cn.groupby("year").agg( framework_norm =("country_norm", "mean"), std_norm =("country_norm", "std"), n_countries_with_data=("country_id", "count"), ).reset_index() asean_sdgs["n_indicators"] = asean_sdgs["year"].apply(lambda y: _n_ind(y, "SDGs")) if not NORMALIZE_FRAMEWORKS_JOINTLY: asean_sdgs["framework_score_1_100"] = global_minmax(asean_sdgs["framework_norm"]) asean_sdgs["framework"] = "SDGs" parts.append(asean_sdgs) df = pd.concat(parts, ignore_index=True) if NORMALIZE_FRAMEWORKS_JOINTLY: mixed_mask = (df["framework"].isin(["MDGs", "SDGs"])) & (df["year"] >= self.sdgs_start_year) if mixed_mask.any(): df.loc[mixed_mask, "framework_score_1_100"] = global_minmax(df.loc[mixed_mask, "framework_norm"]) df = check_and_dedup(df, ["framework", "year"], context=table_name, logger=self.logger) df = add_yoy(df, ["framework"], "framework_score_1_100") df["performance_status"] = df["framework_score_1_100"].apply(_performance_status) df["year"] = df["year"].astype(int) df["n_indicators"] = safe_int(df["n_indicators"], col_name="n_indicators", logger=self.logger) df["n_countries_with_data"] = safe_int(df["n_countries_with_data"], col_name="n_countries_with_data", logger=self.logger) for col in ["framework_norm", "std_norm", "framework_score_1_100"]: df[col] = df[col].astype(float) self._validate_mdgs_equals_total(df, level="asean") schema = [ bigquery.SchemaField("year", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("framework", "STRING", mode="REQUIRED"), bigquery.SchemaField("n_indicators", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("n_countries_with_data", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("framework_norm", "FLOAT", mode="REQUIRED"), bigquery.SchemaField("std_norm", "FLOAT", mode="NULLABLE"), bigquery.SchemaField("framework_score_1_100", "FLOAT", mode="REQUIRED"), bigquery.SchemaField("year_over_year_change", "FLOAT", mode="NULLABLE"), bigquery.SchemaField("performance_status", "STRING", mode="REQUIRED"), ] rows = load_to_bigquery( self.client, df, table_name, layer='gold', write_disposition="WRITE_TRUNCATE", schema=schema ) self._finalize(table_name, rows) return df except Exception as e: self._fail(table_name, e) raise # ========================================================================= # STEP 6: agg_narrative_overview # ========================================================================= def calc_narrative_overview( self, df_framework_asean: pd.DataFrame, df_framework_by_country: pd.DataFrame, ) -> pd.DataFrame: table_name = "agg_narrative_overview" self.load_metadata[table_name]["start_time"] = datetime.now() self.logger.info("\n" + "=" * 70) self.logger.info(f"STEP 6: {table_name} -> [Gold] fs_asean_gold") self.logger.info(" Narrative: interpretatif, plain text, bilingual EN/ID") self.logger.info("=" * 70) try: asean_total = ( df_framework_asean[df_framework_asean["framework"] == "Total"] .sort_values("year") .reset_index(drop=True) ) score_by_year = dict(zip(asean_total["year"].astype(int), asean_total["framework_score_1_100"].astype(float))) status_by_year = dict(zip(asean_total["year"].astype(int), asean_total["performance_status"].astype(str))) country_total = df_framework_by_country[df_framework_by_country["framework"] == "Total"].copy() records = [] for _, row in asean_total.iterrows(): yr = int(row["year"]) score = float(row["framework_score_1_100"]) perf_status = str(row["performance_status"]) yoy = row["year_over_year_change"] yoy_val = float(yoy) if pd.notna(yoy) else None n_mdg = self._count_framework_indicators(yr, "MDGs") n_sdg = self._count_framework_indicators(yr, "SDGs") n_total_ind = int( self._ind_year_framework[ self._ind_year_framework["year"] == yr ]["indicator_id"].nunique() ) prev_score = score_by_year.get(yr - 1, None) prev_status = status_by_year.get(yr - 1, "N/A") yoy_pct = ( (yoy_val / prev_score * 100) if (yoy_val is not None and prev_score is not None and prev_score != 0) else None ) yr_country = ( country_total[country_total["year"] == yr] .sort_values("rank_in_framework_year") .reset_index(drop=True) ) ranking_list = [] for _, cr in yr_country.iterrows(): cr_yoy = cr.get("year_over_year_change", None) ranking_list.append({ "rank": int(cr["rank_in_framework_year"]), "country_name": str(cr["country_name"]), "score": round(float(cr["framework_score_1_100"]), 2), "yoy_change": round(float(cr_yoy), 2) if pd.notna(cr_yoy) else None, }) country_ranking_json = json.dumps(ranking_list, ensure_ascii=False) yr_country_yoy = yr_country.dropna(subset=["year_over_year_change"]) if not yr_country_yoy.empty: best_idx = yr_country_yoy["year_over_year_change"].idxmax() worst_idx = yr_country_yoy["year_over_year_change"].idxmin() most_improved_country = str(yr_country_yoy.loc[best_idx, "country_name"]) most_improved_delta = round(float(yr_country_yoy.loc[best_idx, "year_over_year_change"]), 2) most_declined_country = str(yr_country_yoy.loc[worst_idx, "country_name"]) most_declined_delta = round(float(yr_country_yoy.loc[worst_idx, "year_over_year_change"]), 2) else: most_improved_country = most_declined_country = None most_improved_delta = most_declined_delta = None # Semua data skor negara untuk gap analysis country_scores_all = country_total[["year", "country_id", "framework_score_1_100"]].copy() narrative_en, narrative_id = _build_overview_narrative( year = yr, score = score, performance_status = perf_status, yoy_val = yoy_val, n_mdg = n_mdg, n_sdg = n_sdg, ranking_list = ranking_list, most_improved_country = most_improved_country, most_improved_delta = most_improved_delta, most_declined_country = most_declined_country, most_declined_delta = most_declined_delta, historical_scores = score_by_year, country_scores_all = country_scores_all, ) records.append({ "year": yr, "n_mdg_indicators": n_mdg, "n_sdg_indicators": n_sdg, "n_total_indicators": n_total_ind, "asean_total_score": round(score, 2), "performance_status": perf_status, "yoy_change": yoy_val, "yoy_change_pct": round(yoy_pct, 2) if yoy_pct is not None else None, "country_ranking_json": country_ranking_json, "most_improved_country": most_improved_country, "most_improved_delta": most_improved_delta, "most_declined_country": most_declined_country, "most_declined_delta": most_declined_delta, "narrative_en": narrative_en, "narrative_id": narrative_id, }) df = pd.DataFrame(records) df["year"] = df["year"].astype(int) df["n_mdg_indicators"] = df["n_mdg_indicators"].astype(int) df["n_sdg_indicators"] = df["n_sdg_indicators"].astype(int) df["n_total_indicators"] = df["n_total_indicators"].astype(int) df["asean_total_score"] = df["asean_total_score"].astype(float) df["performance_status"] = df["performance_status"].astype(str) df["narrative_en"] = df["narrative_en"].astype(str) df["narrative_id"] = df["narrative_id"].astype(str) for col in ["yoy_change", "yoy_change_pct", "most_improved_delta", "most_declined_delta"]: df[col] = pd.to_numeric(df[col], errors="coerce").astype(float) self.logger.info("\n Sample narrative_en (year 1):") self.logger.info(f" {df.iloc[0]['narrative_en'][:300]}") self.logger.info("\n Sample narrative_id (year 1):") self.logger.info(f" {df.iloc[0]['narrative_id'][:300]}") schema = [ bigquery.SchemaField("year", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("n_mdg_indicators", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("n_sdg_indicators", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("n_total_indicators", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("asean_total_score", "FLOAT", mode="REQUIRED"), bigquery.SchemaField("performance_status", "STRING", mode="REQUIRED"), bigquery.SchemaField("yoy_change", "FLOAT", mode="NULLABLE"), bigquery.SchemaField("yoy_change_pct", "FLOAT", mode="NULLABLE"), bigquery.SchemaField("country_ranking_json", "STRING", mode="REQUIRED"), bigquery.SchemaField("most_improved_country", "STRING", mode="NULLABLE"), bigquery.SchemaField("most_improved_delta", "FLOAT", mode="NULLABLE"), bigquery.SchemaField("most_declined_country", "STRING", mode="NULLABLE"), bigquery.SchemaField("most_declined_delta", "FLOAT", mode="NULLABLE"), bigquery.SchemaField("narrative_en", "STRING", mode="REQUIRED"), bigquery.SchemaField("narrative_id", "STRING", mode="REQUIRED"), ] rows = load_to_bigquery( self.client, df, table_name, layer='gold', write_disposition="WRITE_TRUNCATE", schema=schema, ) self._finalize(table_name, rows) return df except Exception as e: self._fail(table_name, e) raise # ========================================================================= # STEP 7: agg_narrative_pillar # ========================================================================= def calc_narrative_pillar( self, df_pillar_composite: pd.DataFrame, df_pillar_by_country: pd.DataFrame, ) -> pd.DataFrame: table_name = "agg_narrative_pillar" self.load_metadata[table_name]["start_time"] = datetime.now() self.logger.info("\n" + "=" * 70) self.logger.info(f"STEP 7: {table_name} -> [Gold] fs_asean_gold") self.logger.info(" Narrative: interpretatif, plain text, bilingual EN/ID") self.logger.info("=" * 70) try: records = [] years = sorted(df_pillar_composite["year"].unique()) # Precompute history per pillar pillar_history = {} for p_id, grp in df_pillar_composite.groupby("pillar_id"): pillar_history[p_id] = dict( zip(grp["year"].astype(int), grp["pillar_score_1_100"].astype(float)) ) for yr in years: yr_pillars = ( df_pillar_composite[df_pillar_composite["year"] == yr] .sort_values("rank_in_year") .reset_index(drop=True) ) yr_country_pillar = df_pillar_by_country[df_pillar_by_country["year"] == yr] for _, prow in yr_pillars.iterrows(): p_id = int(prow["pillar_id"]) p_name = str(prow["pillar_name"]) p_score = float(prow["pillar_score_1_100"]) p_rank = int(prow["rank_in_year"]) p_yoy = prow["year_over_year_change"] p_yoy_val = float(p_yoy) if pd.notna(p_yoy) else None p_country = ( yr_country_pillar[yr_country_pillar["pillar_id"] == p_id] .sort_values("rank_in_pillar_year") .reset_index(drop=True) ) if not p_country.empty: top_country = str(p_country.iloc[0]["country_name"]) top_country_score = round(float(p_country.iloc[0]["pillar_country_score_1_100"]), 2) bot_country = str(p_country.iloc[-1]["country_name"]) bot_country_score = round(float(p_country.iloc[-1]["pillar_country_score_1_100"]), 2) else: top_country = bot_country = None top_country_score = bot_country_score = None # Data historis hanya sampai tahun ini hist_up_to_yr = { y: s for y, s in pillar_history.get(p_id, {}).items() if y <= yr } # Data negara-pilar ini semua tahun (untuk gap analysis) country_pillar_all = df_pillar_by_country[ df_pillar_by_country["pillar_id"] == p_id ][["year", "country_id", "pillar_country_score_1_100"]].copy() narrative_en, narrative_id = _build_pillar_narrative( year = yr, pillar_name = p_name, pillar_score = p_score, rank_in_year = p_rank, n_pillars = len(yr_pillars), yoy_val = p_yoy_val, top_country = top_country, top_country_score = top_country_score, bot_country = bot_country, bot_country_score = bot_country_score, pillar_scores_history = hist_up_to_yr, all_pillar_scores_year= yr_pillars[["pillar_name", "pillar_score_1_100"]].copy(), country_pillar_all = country_pillar_all, ) records.append({ "year": yr, "pillar_id": p_id, "pillar_name": p_name, "pillar_score": round(p_score, 2), "rank_in_year": p_rank, "yoy_change": p_yoy_val, "top_country": top_country, "top_country_score": top_country_score, "bottom_country": bot_country, "bottom_country_score": bot_country_score, "narrative_en": narrative_en, "narrative_id": narrative_id, }) df = pd.DataFrame(records) df["year"] = df["year"].astype(int) df["pillar_id"] = df["pillar_id"].astype(int) df["rank_in_year"] = df["rank_in_year"].astype(int) df["narrative_en"] = df["narrative_en"].astype(str) df["narrative_id"] = df["narrative_id"].astype(str) for col in ["pillar_score", "yoy_change", "top_country_score", "bottom_country_score"]: df[col] = pd.to_numeric(df[col], errors="coerce").astype(float) self.logger.info("\n Sample narrative_en (first row):") self.logger.info(f" {df.iloc[0]['narrative_en'][:300]}") self.logger.info("\n Sample narrative_id (first row):") self.logger.info(f" {df.iloc[0]['narrative_id'][:300]}") schema = [ bigquery.SchemaField("year", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("pillar_id", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("pillar_name", "STRING", mode="REQUIRED"), bigquery.SchemaField("pillar_score", "FLOAT", mode="REQUIRED"), bigquery.SchemaField("rank_in_year", "INTEGER", mode="REQUIRED"), bigquery.SchemaField("yoy_change", "FLOAT", mode="NULLABLE"), bigquery.SchemaField("top_country", "STRING", mode="NULLABLE"), bigquery.SchemaField("top_country_score", "FLOAT", mode="NULLABLE"), bigquery.SchemaField("bottom_country", "STRING", mode="NULLABLE"), bigquery.SchemaField("bottom_country_score", "FLOAT", mode="NULLABLE"), bigquery.SchemaField("narrative_en", "STRING", mode="REQUIRED"), bigquery.SchemaField("narrative_id", "STRING", mode="REQUIRED"), ] rows = load_to_bigquery( self.client, df, table_name, layer='gold', write_disposition="WRITE_TRUNCATE", schema=schema, ) self._finalize(table_name, rows) return df except Exception as e: self._fail(table_name, e) raise # ========================================================================= # HELPERS # ========================================================================= def _validate_mdgs_equals_total(self, df: pd.DataFrame, level: str = ""): self.logger.info(f"\n Validasi MDGs < {self.sdgs_start_year} == Total [{level}]:") group_by = ["year"] if level.startswith("asean") else ["country_id", "year"] mdgs_pre = df[(df["framework"] == "MDGs") & (df["year"] < self.sdgs_start_year)][group_by + ["framework_score_1_100"]].rename(columns={"framework_score_1_100": "mdgs_score"}) total_pre = df[(df["framework"] == "Total") & (df["year"] < self.sdgs_start_year)][group_by + ["framework_score_1_100"]].rename(columns={"framework_score_1_100": "total_score"}) if mdgs_pre.empty and total_pre.empty: self.logger.info(f" -> Tidak ada data pre-{self.sdgs_start_year} (skip)") return if mdgs_pre.empty or total_pre.empty: self.logger.warning(f" -> [WARNING] Salah satu kosong") return check = mdgs_pre.merge(total_pre, on=group_by) max_diff = (check["mdgs_score"] - check["total_score"]).abs().max() status = "OK (identik)" if max_diff < 0.01 else f"MISMATCH! max_diff={max_diff:.6f}" self.logger.info(f" -> {status} (n_checked={len(check)})") def _finalize(self, table_name: str, rows_loaded: int): end_time = datetime.now() start_time = self.load_metadata[table_name].get("start_time") self.load_metadata[table_name].update({ "rows_loaded": rows_loaded, "status" : "success", "end_time" : end_time, }) log_update(self.client, "DW", table_name, "full_load", rows_loaded) try: save_etl_metadata( self.client, self._build_etl_metadata( table_name = table_name, rows_loaded = rows_loaded, start_time = start_time, end_time = end_time, status = "success", ) ) except Exception as meta_err: self.logger.warning(f" [METADATA WARNING] {table_name}: {meta_err}") self.logger.info(f" [OK] {table_name}: {rows_loaded:,} rows -> [Gold] fs_asean_gold") def _fail(self, table_name: str, error: Exception): end_time = datetime.now() start_time = self.load_metadata[table_name].get("start_time") error_msg = str(error) self.load_metadata[table_name].update({"status": "failed", "end_time": end_time}) log_update(self.client, "DW", table_name, "full_load", 0, "failed", error_msg) try: save_etl_metadata( self.client, self._build_etl_metadata( table_name = table_name, rows_loaded = 0, start_time = start_time, end_time = end_time, status = "failed", error_msg = error_msg, ) ) except Exception as meta_err: self.logger.warning(f" [METADATA WARNING] {table_name}: {meta_err}") self.logger.error(f" [FAIL] {table_name}: {error_msg}") # ========================================================================= # RUN # ========================================================================= def run(self): start = datetime.now() self.logger.info("\n" + "=" * 70) self.logger.info("FOOD SECURITY AGGREGATION — 6 TABLES -> fs_asean_gold") self.logger.info(f" Performance threshold: {PERFORMANCE_THRESHOLD}") self.logger.info(f" Narrative style : interpretive, plain text, bilingual EN/ID") self.logger.info("=" * 70) self.load_data() self.sdgs_start_year = self._detect_sdgs_start_year() self._assign_framework_labels() df_pillar_composite = self.calc_pillar_composite() df_pillar_by_country = self.calc_pillar_by_country() df_framework_by_country = self.calc_framework_by_country() df_framework_asean = self.calc_framework_asean() self.calc_narrative_overview( df_framework_asean = df_framework_asean, df_framework_by_country = df_framework_by_country, ) self.calc_narrative_pillar( df_pillar_composite = df_pillar_composite, df_pillar_by_country = df_pillar_by_country, ) duration = (datetime.now() - start).total_seconds() total_rows = sum(m["rows_loaded"] for m in self.load_metadata.values()) self.logger.info("\n" + "=" * 70) self.logger.info("SELESAI") self.logger.info("=" * 70) self.logger.info(f" Durasi : {duration:.2f}s") self.logger.info(f" Total rows : {total_rows:,}") for tbl, meta in self.load_metadata.items(): icon = "[OK]" if meta["status"] == "success" else "[FAIL]" self.logger.info(f" {icon} {tbl:<35} {meta['rows_loaded']:>10,}") # ============================================================================= # AIRFLOW TASK # ============================================================================= def run_aggregation(): from scripts.bigquery_config import get_bigquery_client client = get_bigquery_client() agg = FoodSecurityAggregator(client) agg.run() total = sum(m["rows_loaded"] for m in agg.load_metadata.values()) print(f"Aggregation completed: {total:,} total rows loaded") # ============================================================================= # MAIN # ============================================================================= if __name__ == "__main__": import io if _sys.stdout.encoding and _sys.stdout.encoding.lower() not in ("utf-8", "utf8"): _sys.stdout = io.TextIOWrapper(_sys.stdout.buffer, encoding="utf-8", errors="replace") if _sys.stderr.encoding and _sys.stderr.encoding.lower() not in ("utf-8", "utf8"): _sys.stderr = io.TextIOWrapper(_sys.stderr.buffer, encoding="utf-8", errors="replace") print("=" * 70) print("FOOD SECURITY AGGREGATION -> fs_asean_gold") print(f" NORMALIZE_FRAMEWORKS_JOINTLY : {NORMALIZE_FRAMEWORKS_JOINTLY}") print(f" PERFORMANCE_THRESHOLD : {PERFORMANCE_THRESHOLD}") print("=" * 70) logger = setup_logging() for handler in logger.handlers: handler.__class__ = _SafeStreamHandler client = get_bigquery_client() agg = FoodSecurityAggregator(client) agg.run() print("\n" + "=" * 70) print("[OK] SELESAI") print("=" * 70)