izu/airflow-coolify
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

new narrative teks

Browse Source
This commit is contained in:
Debby
2026-04-22 16:02:05 +07:00
parent 40528766bd
commit f9d013f8e6
2 changed files with 206 additions and 464 deletions
Download Patch File Download Diff File Expand all files Collapse all files

406
scripts/bigquery_aggraget_fact_selected_layer.py
View File

@@ -44,22 +44,24 @@ Output Schema (agg_indicator_norm):
agg_narrative_indicator
=============================================================================
Tujuan:
Menghasilkan narasi otomatis 1 paragraf per indikator per tahun di level ASEAN
(rata-rata seluruh negara ASEAN), dijalankan otomatis setelah agg_indicator_norm
selesai dalam pipeline yang sama.
Menghasilkan narasi otomatis 1 paragraf per indikator (level ASEAN,
merangkum seluruh periode + seluruh negara), dijalankan otomatis setelah
agg_indicator_norm selesai dalam pipeline yang sama.
Granularity:
year x indicator_id (level ASEAN, bukan per negara)
indicator_id (all years, all ASEAN countries)
Output Schema (agg_narrative_indicator):
year, indicator_id, indicator_name, unit, direction,
indicator_id, indicator_name, unit, direction,
pillar_name, framework,
avg_value, -- rata-rata value ASEAN
avg_norm_score_1_100, -- rata-rata norm_score_1_100 ASEAN
year_min, year_max, n_countries,
avg_value_first, avg_value_last,
avg_norm_score_1_100,
performance, -- Good | Bad | null
yoy_avg_value, -- perubahan avg_value vs tahun sebelumnya
n_countries, -- jumlah negara yang punya data tahun ini
narrative -- 1 paragraf narasi otomatis
n_yoy_total, n_yoy_positive,
best_yoy_from, best_yoy_to,
country_worst, country_best,
narrative
"""
import pandas as pd
@@ -114,10 +116,8 @@ SDG_ONLY_KEYWORDS: frozenset = frozenset([
"number of women of reproductive age (15-49 years) affected by anemia (million)",
])
# Lowercase set untuk matching case-insensitive
_SDG_ONLY_LOWER: frozenset = frozenset(k.lower() for k in SDG_ONLY_KEYWORDS)
# FIES-specific keywords untuk deteksi sdgs_start_year
_FIES_DETECTION_KEYWORDS: 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)",
@@ -133,12 +133,11 @@ DIRECTION_POSITIVE_KEYWORDS = frozenset({
"positive", "higher_better", "higher_is_better",
})
# Threshold performance label
_PERFORMANCE_THRESHOLD: float = 60.0
# =============================================================================
# PURE HELPERS — agg_indicator_norm
# PURE HELPERS
# =============================================================================
def _should_invert(direction: str, logger=None, context: str = "") -> bool:
@@ -171,12 +170,8 @@ def global_minmax(series: pd.Series, lo: float = 1.0, hi: float = 100.0) -> pd.S
def _compute_yoy(df: pd.DataFrame) -> pd.DataFrame:
"""
Hitung YoY untuk satu grup (indicator_id, country_id) yang sudah di-sort by year.
Kolom yang ditambahkan:
yoy_value : value - value_prev
yoy_norm_value : norm_value - norm_value_prev
Baris pertama tiap grup selalu null (tidak ada tahun sebelumnya).
Kolom yang ditambahkan: yoy_value, yoy_norm_value.
Baris pertama tiap grup selalu null.
"""
df = df.sort_values("year").copy()
@@ -199,70 +194,22 @@ def _compute_yoy(df: pd.DataFrame) -> pd.DataFrame:
return df
# =============================================================================
# PURE HELPERS — agg_narrative_indicator
# =============================================================================
def _is_lower_better(direction: str) -> bool:
return str(direction).lower().strip() in DIRECTION_INVERT_KEYWORDS
def _format_value(value: float, unit: str) -> str:
"""Format nilai dengan unit yang sesuai."""
if pd.isna(value):
return "N/A"
unit = str(unit).strip() if unit else ""
if abs(value) >= 1000:
formatted = f"{value:,.1f}"
elif abs(value) >= 10:
formatted = f"{value:.2f}"
else:
formatted = f"{value:.3f}"
return f"{formatted} {unit}".strip()
def _format_yoy(yoy: float, unit: str, lower_better: bool) -> tuple:
"""
Kembalikan (direction_word, change_desc, is_positive_trend).
is_positive_trend: True jika perubahan menguntungkan sesuai direction.
"""
unit = str(unit).strip() if unit else ""
abs_yoy = abs(yoy)
if abs_yoy >= 1000:
yoy_str = f"{abs_yoy:,.1f}"
elif abs_yoy >= 10:
yoy_str = f"{abs_yoy:.2f}"
else:
yoy_str = f"{abs_yoy:.3f}"
change_desc = f"{yoy_str} {unit}".strip()
is_positive = (yoy < 0) if lower_better else (yoy > 0)
direction_word = "decreased by" if yoy < 0 else "increased by"
return direction_word, change_desc, is_positive
# =============================================================================
# PURE HELPER — narrative builder (per indicator, all years, all countries)
# ======================================================================
# NARRATIVE BUILDER — agg_narrative_indicator
# =============================================================================
def _build_narrative_per_indicator(row: pd.Series) -> str:
"""
Bangun 1 paragraf narasi ASEAN-level untuk satu indikator,
merangkum seluruh periode (year_min - year_max) dan seluruh negara.
Kolom yang dibutuhkan dari row:
indicator_name, unit, direction, pillar_name, framework,
year_min, year_max, n_countries,
avg_value_first, avg_value_last,
avg_norm_score_1_100, -- rata-rata seluruh periode
performance, -- Good | Bad | null
n_yoy_total, -- total transisi year-on-year
n_yoy_positive, -- jumlah transisi yang membaik
best_yoy_from, best_yoy_to, -- periode dengan perbaikan terbesar
country_worst, country_best -- negara dengan nilai terburuk / terbaik
Narrative format (no em-dash, bold on key figures):
**{indicator}** ({framework}, {pillar}): ASEAN average {rose/fell} from
**{first}** to **{last}** ({year_min} to {year_max}), **{improving/deteriorating}** trend.
Score: **{score}/100** (*{Good/Bad}*).
Best country: **{best}**; worst: **{worst}**.
Improved in **{n_pos}/{n_total}** YoY transitions.
"""
ind_name = str(row["indicator_name"]).strip()
unit = str(row["unit"]).strip() if row["unit"] else ""
@@ -274,132 +221,81 @@ def _build_narrative_per_indicator(row: pd.Series) -> str:
n_countries = int(row["n_countries"])
avg_score = row["avg_norm_score_1_100"]
performance = row["performance"]
avg_first = row["avg_value_first"]
avg_last = row["avg_value_last"]
n_yoy_total = int(row["n_yoy_total"]) if not pd.isna(row["n_yoy_total"]) else 0
n_yoy_positive = int(row["n_yoy_positive"]) if not pd.isna(row["n_yoy_positive"]) else 0
best_yoy_from = row["best_yoy_from"]
best_yoy_to = row["best_yoy_to"]
country_worst = str(row["country_worst"]).strip() if not pd.isna(row["country_worst"]) else None
country_best = str(row["country_best"]).strip() if not pd.isna(row["country_best"]) else None
lower_better = _is_lower_better(direction)
direction_label = (
"lower values indicate better outcomes"
if lower_better
else "higher values indicate better outcomes"
)
# ---- Kalimat 1: Identifikasi indikator + cakupan -------------------------
member_str = f"{n_countries} member state{'s' if n_countries > 1 else ''}"
sentence1 = (
f"Across ASEAN, {ind_name} under the {framework} framework "
f"({pillar} pillar) was monitored from {year_min} to {year_max} "
f"across {member_str}."
)
# ---- Kalimat 2: Tren keseluruhan (first → last) --------------------------
lower_better = _is_lower_better(direction)
def _fmt(v):
if pd.isna(v):
return "N/A"
abs_v = abs(v)
if abs_v >= 1000:
s = f"{v:,.1f}"
elif abs_v >= 10:
s = f"{v:.2f}"
else:
s = f"{v:.3f}"
return f"{s} {unit}".strip() if unit else s
# Sentence 1: trend first -> last
if not pd.isna(avg_first) and not pd.isna(avg_last):
diff = avg_last - avg_first
abs_diff = abs(diff)
# Format nilai
def fmt(v):
if abs(v) >= 1000:
return f"{v:,.1f}"
elif abs(v) >= 10:
return f"{v:.2f}"
else:
return f"{v:.3f}"
first_str = f"{fmt(avg_first)}{' ' + unit if unit else ''}"
last_str = f"{fmt(avg_last)}{' ' + unit if unit else ''}"
diff_str = f"{fmt(abs_diff)}{' ' + unit if unit else ''}"
# Apakah tren menguntungkan?
diff = avg_last - avg_first
is_improving = (diff < 0) if lower_better else (diff > 0)
trend_word = "improving" if is_improving else "deteriorating"
verb = "declining" if diff < 0 else "rising"
sentence2 = (
f"Since {direction_label}, the region collectively showed "
f"{'an' if trend_word[0] in 'aeiou' else 'a'} {trend_word} trend, "
f"with the ASEAN average {verb} from {first_str} in {year_min} "
f"to {last_str} in {year_max} "
f"(a cumulative {'reduction' if diff < 0 else 'increase'} of {diff_str})."
trend_label = "improving" if is_improving else "deteriorating"
verb = "fell" if diff < 0 else "rose"
sent1 = (
f"**{ind_name}** ({framework}, {pillar}): ASEAN average {verb} from "
f"**{_fmt(avg_first)}** to **{_fmt(avg_last)}** ({year_min} to {year_max}), "
f"**{trend_label}** trend."
)
else:
sentence2 = (
f"Since {direction_label}, trend analysis could not be performed "
f"due to missing data at the start or end of the period."
sent1 = (
f"**{ind_name}** ({framework}, {pillar}): trend data unavailable "
f"({year_min} to {year_max}, {n_countries} members)."
)
# ---- Kalimat 3: Score + performance -------------------------------------
# Sentence 2: score + performance
if not pd.isna(avg_score):
score_str = f"{avg_score:.1f} out of 100"
if performance == "Good":
sentence3 = (
f"The regional normalized score averaged {score_str} "
f"classified as Good performance."
)
elif performance == "Bad":
sentence3 = (
f"The regional normalized score averaged {score_str} "
f"classified as Bad performance, falling below the 60-point threshold."
)
else:
sentence3 = (
f"The regional normalized score averaged {score_str}."
)
perf_label = f"*{performance}*" if performance in ("Good", "Bad") else ""
sent2 = f"Score: **{avg_score:.1f}/100** {perf_label}.".strip()
else:
sentence3 = "The regional normalized performance score could not be assessed."
# ---- Kalimat 4: Negara terbaik & terburuk --------------------------------
if country_worst and country_best and country_worst != country_best:
if lower_better:
worst_label = "highest (most concerning)"
best_label = "consistently performed best (lowest values)"
else:
worst_label = "lowest (most concerning)"
best_label = "consistently performed best (highest values)"
sentence4 = (
f"Among member states, {country_worst} recorded the {worst_label} "
f"levels throughout the period, while {country_best} {best_label}."
)
sent2 = "Score unavailable."
# Sentence 3: best / worst country
if country_best and country_worst and country_best != country_worst:
sent3 = f"Best country: **{country_best}**; worst: **{country_worst}**."
elif country_best:
sentence4 = (
f"Among member states, {country_best} consistently recorded the "
f"best performance throughout the period."
)
sent3 = f"Best country: **{country_best}**."
else:
sentence4 = ""
# ---- Kalimat 5: YoY transitions -----------------------------------------
sent3 = ""
# Sentence 4: YoY transitions
if n_yoy_total > 0:
yoy_sentence = (
f"Year-on-year, the region improved in {n_yoy_positive} out of "
f"{n_yoy_total} transition{'s' if n_yoy_total > 1 else ''}"
)
best_period = ""
if not pd.isna(best_yoy_from) and not pd.isna(best_yoy_to):
yoy_sentence += (
f", with the largest regional gain occurring between "
f"{int(best_yoy_from)} and {int(best_yoy_to)}."
)
else:
yoy_sentence += "."
best_period = f", best gain: **{int(best_yoy_from)} to {int(best_yoy_to)}**"
sent4 = (
f"Improved in **{n_yoy_positive}/{n_yoy_total}** YoY transitions{best_period}."
)
else:
yoy_sentence = "Insufficient data to assess year-on-year transitions."
parts = [sentence1, sentence2, sentence3]
if sentence4:
parts.append(sentence4)
parts.append(yoy_sentence)
sent4 = "Insufficient data for YoY assessment."
parts = [sent1, sent2]
if sent3:
parts.append(sent3)
parts.append(sent4)
return " ".join(parts)
@@ -427,10 +323,10 @@ class IndicatorNormAggregator:
11. Summary log agg_indicator_norm
Alur agg_narrative_indicator (lanjutan, pakai df_final yang sudah ada):
12. Agregasi ke level ASEAN (year x indicator_id)
12. Agregasi ke level ASEAN per indicator_id
13. Hitung YoY avg_value per indikator
14. Assign performance berdasarkan avg_norm_score
15. Build narrative 1 paragraf per baris
15. Build narrative 1 paragraf per indikator
16. Simpan ke BigQuery -> agg_narrative_indicator
17. Summary log agg_narrative_indicator
"""
@@ -769,11 +665,6 @@ class IndicatorNormAggregator:
# =========================================================================
def _assign_performance(self, df: pd.DataFrame) -> pd.DataFrame:
"""
performance = "Good" jika norm_score_1_100 >= 60
= "Bad" jika norm_score_1_100 < 60
= null jika norm_score_1_100 null
"""
self.logger.info("\n" + "=" * 80)
self.logger.info(
f"STEP 9: ASSIGN PERFORMANCE LABEL "
@@ -832,7 +723,6 @@ class IndicatorNormAggregator:
["year", "country_name", "pillar_name", "indicator_name"]
).reset_index(drop=True)
# Cast
out["year"] = out["year"].astype(int)
out["country_id"] = out["country_id"].astype(int)
out["country_name"] = out["country_name"].astype(str)
@@ -949,7 +839,6 @@ class IndicatorNormAggregator:
f"{r['avg_score']:.2f}"
)
# Performance summary per framework
self.logger.info("\n Performance summary per Framework:")
perf_fw = (
df[df["performance"].notna()]
@@ -967,7 +856,6 @@ class IndicatorNormAggregator:
f"({r['count']/total*100:.1f}%)"
)
# Top 5 & Bottom 5 indikator
ind_avg = (
df.groupby(["indicator_id", "indicator_name", "unit", "pillar_name", "direction"])
["norm_score_1_100"].mean()
@@ -995,7 +883,6 @@ class IndicatorNormAggregator:
f"{r['norm_score_1_100']:.2f} {tag} {unit}"
)
# Indikator per pillar
pillar_summary = (
df.drop_duplicates(subset=["indicator_id", "pillar_name"])
.groupby("pillar_name")["indicator_id"]
@@ -1008,29 +895,24 @@ class IndicatorNormAggregator:
self.logger.info(f" {r['pillar_name']:<30}: {r['n_indicators']}")
# =========================================================================
# STEP 12-16: agg_narrative_indicator (lanjutan dari df_final)
# STEP 12-16: agg_narrative_indicator
# =========================================================================
def _build_narrative_table(self, df_final: pd.DataFrame):
"""
Pipeline agg_narrative_indicator — granularity: per indicator_id (1 baris per indikator).
Narasi merangkum seluruh periode + seluruh negara ASEAN.
Dijalankan otomatis setelah agg_indicator_norm selesai.
Pipeline agg_narrative_indicator.
Granularity: per indicator_id (1 baris per indikator, all years, all countries).
"""
self.logger.info("\n" + "=" * 80)
self.logger.info("STEP 12-16: agg_narrative_indicator")
self.logger.info(" Level : per indicator_id (all years + all ASEAN countries)")
self.logger.info("=" * 80)
# -- STEP 12: Hitung statistik agregat per (indicator_id, country_id, year) --
self.logger.info("\n--- STEP 12: COMPUTE INDICATOR-LEVEL STATS ---")
df = df_final.copy()
# Dimensi tetap per indikator
dim_cols = ["indicator_name", "unit", "direction", "pillar_name", "framework"]
# ---- 12a. ASEAN avg per (indicator_id, year) -> untuk first/last & YoY ---
# ---- 12a. ASEAN avg per (indicator_id, year) -------------------------
self.logger.info("\n--- STEP 12: COMPUTE INDICATOR-LEVEL STATS ---")
df_yr = (
df.groupby(["indicator_id", "year"])
.agg(
@@ -1040,41 +922,38 @@ class IndicatorNormAggregator:
)
.reset_index()
)
# ---- 12b. first year / last year avg value per indikator -----------------
# ---- 12b. first / last avg value per indikator -----------------------
df_first = (
df_yr.sort_values("year")
.groupby("indicator_id")
.first()
.reset_index()[["indicator_id", "year", "avg_value"]]
.groupby("indicator_id").first().reset_index()
[["indicator_id", "year", "avg_value"]]
.rename(columns={"year": "year_min", "avg_value": "avg_value_first"})
)
df_last = (
df_yr.sort_values("year")
.groupby("indicator_id")
.last()
.reset_index()[["indicator_id", "year", "avg_value"]]
.groupby("indicator_id").last().reset_index()
[["indicator_id", "year", "avg_value"]]
.rename(columns={"year": "year_max", "avg_value": "avg_value_last"})
)
# ---- 12c. Rata-rata norm_score seluruh periode ----------------------------
# ---- 12c. Rata-rata norm_score seluruh periode -----------------------
df_score_avg = (
df_yr.groupby("indicator_id")
.agg(avg_norm_score_1_100=("avg_norm_score", "mean"))
.reset_index()
)
# ---- 12d. n_countries: maks negara yang pernah hadir ---------------------
# ---- 12d. n_countries ------------------------------------------------
df_nc = (
df.groupby("indicator_id")["country_id"]
.nunique()
.reset_index()
.nunique().reset_index()
.rename(columns={"country_id": "n_countries"})
)
# ---- 12e. YoY per (indicator_id) di level ASEAN avg ----------------------
# ---- 12e. YoY per indicator (ASEAN avg) ------------------------------
self.logger.info("\n--- STEP 13: COMPUTE YoY (ASEAN avg, per indicator) ---")
yoy_parts = []
for ind_id, grp in df_yr.groupby("indicator_id"):
grp = grp.sort_values("year").copy()
@@ -1087,44 +966,36 @@ class IndicatorNormAggregator:
grp = grp.drop(columns=["prev_avg"])
yoy_parts.append(grp)
df_yr = pd.concat(yoy_parts, ignore_index=True)
# Ambil direction per indikator untuk tentukan "improving"
dir_map = (
df[["indicator_id", "direction"]]
.drop_duplicates(subset=["indicator_id"])
.set_index("indicator_id")["direction"]
.to_dict()
)
def _is_positive_yoy(ind_id, yoy_val):
"""True jika perubahan yoy menguntungkan sesuai direction."""
if pd.isna(yoy_val):
return False
lb = _is_lower_better(dir_map.get(ind_id, "positive"))
return (yoy_val < 0) if lb else (yoy_val > 0)
# Hitung n_yoy_total, n_yoy_positive, best_yoy
yoy_stats = []
for ind_id, grp in df_yr.groupby("indicator_id"):
grp_yoy = grp[grp["yoy"].notna()].copy()
lb = _is_lower_better(dir_map.get(ind_id, "positive"))
n_total = len(grp_yoy)
n_positive = int(sum(_is_positive_yoy(ind_id, v) for v in grp_yoy["yoy"]))
# "Best" = perubahan paling menguntungkan
if n_total > 0:
if lb:
idx_best = grp_yoy["yoy"].idxmin() # paling negatif = paling baik
else:
idx_best = grp_yoy["yoy"].idxmax() # paling positif = paling baik
idx_best = grp_yoy["yoy"].idxmin() if lb else grp_yoy["yoy"].idxmax()
best_row = grp_yoy.loc[idx_best]
best_yoy_from = best_row["year"] - 1
best_yoy_to = best_row["year"]
else:
best_yoy_from = np.nan
best_yoy_to = np.nan
yoy_stats.append({
"indicator_id" : ind_id,
"n_yoy_total" : n_total,
@@ -1132,16 +1003,16 @@ class IndicatorNormAggregator:
"best_yoy_from" : best_yoy_from,
"best_yoy_to" : best_yoy_to,
})
df_yoy_stats = pd.DataFrame(yoy_stats)
# ---- 12f. Country terbaik & terburuk (rata-rata value seluruh periode) ---
# ---- 12f. Country terbaik & terburuk ---------------------------------
df_country_avg = (
df.groupby(["indicator_id", "country_id", "country_name"])
.agg(country_avg_value=("value", "mean"))
.reset_index()
)
country_stats = []
for ind_id, grp in df_country_avg.groupby("indicator_id"):
lb = _is_lower_better(dir_map.get(ind_id, "positive"))
@@ -1152,33 +1023,33 @@ class IndicatorNormAggregator:
worst_row = grp.loc[grp["country_avg_value"].idxmin()]
best_row = grp.loc[grp["country_avg_value"].idxmax()]
country_stats.append({
"indicator_id": ind_id,
"indicator_id" : ind_id,
"country_worst": worst_row["country_name"],
"country_best" : best_row["country_name"],
})
df_country_stats = pd.DataFrame(country_stats)
# ---- 12g. Dimensi tetap per indikator ------------------------------------
# ---- 12g. Dimensi tetap per indikator --------------------------------
df_dim = (
df[["indicator_id"] + dim_cols]
.drop_duplicates(subset=["indicator_id"])
)
# ---- 12h. Merge semua -------------------------------------------------------
# ---- 12h. Merge semua ------------------------------------------------
df_agg = (
df_dim
.merge(df_first, on="indicator_id", how="left")
.merge(df_last, on="indicator_id", how="left")
.merge(df_score_avg, on="indicator_id", how="left")
.merge(df_nc, on="indicator_id", how="left")
.merge(df_yoy_stats, on="indicator_id", how="left")
.merge(df_country_stats,on="indicator_id", how="left")
.merge(df_first, on="indicator_id", how="left")
.merge(df_last, on="indicator_id", how="left")
.merge(df_score_avg, on="indicator_id", how="left")
.merge(df_nc, on="indicator_id", how="left")
.merge(df_yoy_stats, on="indicator_id", how="left")
.merge(df_country_stats, on="indicator_id", how="left")
)
self.logger.info(f" Rows (1 per indicator) : {len(df_agg):,}")
self.logger.info(f" Indicators : {df_agg['indicator_id'].nunique()}")
# -- STEP 14: Assign performance --------------------------------------------
# -- STEP 14: Assign performance ---------------------------------------
self.logger.info("\n--- STEP 14: ASSIGN PERFORMANCE ---")
df_agg["performance"] = pd.NA
has_score = df_agg["avg_norm_score_1_100"].notna()
@@ -1187,8 +1058,8 @@ class IndicatorNormAggregator:
n_good = (df_agg["performance"] == "Good").sum()
n_bad = (df_agg["performance"] == "Bad").sum()
self.logger.info(f" Good: {n_good:,} | Bad: {n_bad:,}")
# -- STEP 15: Build narrative -----------------------------------------------
# -- STEP 15: Build narrative ------------------------------------------
self.logger.info("\n--- STEP 15: BUILD NARRATIVE (per indicator, all years) ---")
df_agg["narrative"] = df_agg.apply(_build_narrative_per_indicator, axis=1)
self.logger.info(f" Narratives generated: {len(df_agg):,}")
@@ -1198,8 +1069,8 @@ class IndicatorNormAggregator:
f"\n [{int(row['indicator_id'])}] {row['indicator_name'][:60]}"
f"\n -> {row['narrative'][:300]}..."
)
# -- STEP 16: Save ----------------------------------------------------------
# -- STEP 16: Save -----------------------------------------------------
self.logger.info("\n--- STEP 16: SAVE -> [Gold] agg_narrative_indicator ---")
out = df_agg[[
"indicator_id", "indicator_name", "unit", "direction",
@@ -1212,10 +1083,9 @@ class IndicatorNormAggregator:
"country_worst", "country_best",
"narrative",
]].copy()
out = out.sort_values(["pillar_name", "indicator_name"]).reset_index(drop=True)
# Cast
out["indicator_id"] = out["indicator_id"].astype(int)
out["indicator_name"] = out["indicator_name"].astype(str)
out["unit"] = out["unit"].fillna("").astype(str)
@@ -1236,7 +1106,7 @@ class IndicatorNormAggregator:
out["country_worst"] = out["country_worst"].astype(str).replace("nan", pd.NA).astype("string")
out["country_best"] = out["country_best"].astype(str).replace("nan", pd.NA).astype("string")
out["narrative"] = out["narrative"].astype(str)
schema = [
bigquery.SchemaField("indicator_id", "INTEGER", mode="REQUIRED"),
bigquery.SchemaField("indicator_name", "STRING", mode="REQUIRED"),
@@ -1259,17 +1129,17 @@ class IndicatorNormAggregator:
bigquery.SchemaField("country_best", "STRING", mode="NULLABLE"),
bigquery.SchemaField("narrative", "STRING", mode="NULLABLE"),
]
rows_loaded = load_to_bigquery(
self.client, out, "agg_narrative_indicator",
layer="gold", write_disposition="WRITE_TRUNCATE", schema=schema,
)
log_update(self.client, "DW", "agg_narrative_indicator", "full_load", rows_loaded)
self.logger.info(
f" [OK] agg_narrative_indicator: {rows_loaded:,} rows -> [Gold] fs_asean_gold"
)
metadata = {
"source_class" : self.__class__.__name__,
"table_name" : "agg_narrative_indicator",
@@ -1293,9 +1163,8 @@ class IndicatorNormAggregator:
}
save_etl_metadata(self.client, metadata)
self.logger.info(" Metadata -> [AUDIT] etl_metadata")
self.pipeline_metadata["rows_loaded_narrative"] = rows_loaded
# =========================================================================
# RUN
@@ -1326,7 +1195,6 @@ class IndicatorNormAggregator:
self.pipeline_metadata["rows_loaded"] = rows_loaded
self._log_summary(df_final)
# Lanjut build agg_narrative_indicator dari df_final (tanpa re-load BQ)
self._build_narrative_table(df_final)
self.pipeline_metadata["end_time"] = datetime.now()
@@ -1345,7 +1213,7 @@ class IndicatorNormAggregator:
# =============================================================================
# AIRFLOW TASK <-- tidak berubah
# AIRFLOW TASK
# =============================================================================
def run_indicator_norm_aggregation():

264
scripts/bigquery_aggregate_layer.py
View File

@@ -219,113 +219,58 @@ def _build_overview_narrative(
most_declined_country,
most_declined_delta,
) -> str:
# Sentence 1: indicator breakdown
parts_ind = []
"""
Narrative format (no em-dash):
In {year}, ASEAN scored {score} ({performance}) across {n_total} indicators
({n_mdg} MDGs, {n_sdg} SDGs). Score {increased/decreased} by {delta} pts from
{prev_year} ({prev_score}). {top_country} led the region; {bottom_country} ranked
last. Biggest gain: {country}; biggest drop: {country}.
"""
# Sentence 1: score + performance + indicators
ind_parts = []
if n_mdg > 0:
parts_ind.append(f"{n_mdg} MDG indicator{'s' if n_mdg > 1 else ''}")
ind_parts.append(f"**{n_mdg} MDGs**")
if n_sdg > 0:
parts_ind.append(f"{n_sdg} SDG indicator{'s' if n_sdg > 1 else ''}")
ind_parts.append(f"**{n_sdg} SDGs**")
ind_detail = f" ({', '.join(ind_parts)})" if ind_parts else ""
if parts_ind:
ind_detail = " and ".join(parts_ind)
sent1 = (
f"In {year}, the ASEAN food security assessment incorporated a total of "
f"{n_total_ind} indicator{'s' if n_total_ind != 1 else ''}, "
f"consisting of {ind_detail}."
)
else:
sent1 = (
f"In {year}, the ASEAN food security assessment incorporated "
f"{n_total_ind} indicator{'s' if n_total_ind != 1 else ''}."
)
# Sentence 2: score + performance status + YoY
status_phrase = (
f"classified as \"{performance_status}\" performance "
f"(threshold: {PERFORMANCE_THRESHOLD:.0f})"
sent1 = (
f"In **{year}**, ASEAN scored **{_fmt_score(score)}** (*{performance_status}*) "
f"across **{n_total_ind} indicators**{ind_detail}."
)
# Sentence 2: YoY
if yoy_val is not None and prev_score is not None:
direction_word = "increasing" if yoy_val >= 0 else "decreasing"
pct_clause = ""
if yoy_pct is not None:
abs_pct = abs(yoy_pct)
trend_word = "improvement" if yoy_val >= 0 else "decline"
pct_clause = f", representing a {abs_pct:.2f}% {trend_word} year-over-year"
status_change = ""
if prev_performance_status not in ("N/A", None) and prev_performance_status != performance_status:
status_change = (
f" This marks a shift from \"{prev_performance_status}\" in {prev_year} "
f"to \"{performance_status}\" in {year}."
)
direction_word = "increased" if yoy_val >= 0 else "decreased"
sent2 = (
f"The ASEAN overall score (Total framework) reached {_fmt_score(score)}, "
f"{status_phrase}, {direction_word} by {abs(yoy_val):.2f} points compared to "
f"{prev_year} ({_fmt_score(prev_score)}, \"{prev_performance_status}\"){pct_clause}.{status_change}"
f"Score {direction_word} by **{abs(yoy_val):.2f} pts** "
f"from {prev_year} ({_fmt_score(prev_score)}, *{prev_performance_status}*)."
)
else:
sent2 = (
f"The ASEAN overall score (Total framework) reached {_fmt_score(score)} in {year}, "
f"{status_phrase}. No prior-year data is available for year-over-year comparison."
)
sent2 = "No prior-year data available for comparison."
# Sentence 3: country ranking
sent3 = ""
if ranking_list:
first = ranking_list[0]
last = ranking_list[-1]
middle = ranking_list[1:-1]
first = ranking_list[0]
last = ranking_list[-1]
if len(ranking_list) == 1:
sent3 = (
f"In terms of country performance, {first['country_name']} was the only "
f"country assessed, scoring {_fmt_score(first['score'])} in {year}."
)
elif len(ranking_list) == 2:
sent3 = (
f"In terms of country performance, {first['country_name']} led the region "
f"with a score of {_fmt_score(first['score'])}, while "
f"{last['country_name']} recorded the lowest score of "
f"{_fmt_score(last['score'])} in {year}."
)
sent3 = f"**{first['country_name']}** was the only country assessed ({_fmt_score(first['score'])})."
else:
middle_parts = [
f"{c['country_name']} ({_fmt_score(c['score'])})" for c in middle
]
if len(middle_parts) == 1:
middle_str = middle_parts[0]
else:
middle_str = ", ".join(middle_parts[:-1]) + f", and {middle_parts[-1]}"
sent3 = (
f"In terms of country performance, {first['country_name']} led the region "
f"with a score of {_fmt_score(first['score'])}, followed by {middle_str}. "
f"At the other end, {last['country_name']} recorded the lowest score "
f"of {_fmt_score(last['score'])} in {year}."
f"**{first['country_name']}** led the region ({_fmt_score(first['score'])}); "
f"**{last['country_name']}** ranked last ({_fmt_score(last['score'])})."
)
# Sentence 4: most improved / declined country
# Sentence 4: most improved / declined
sent4_parts = []
if most_improved_country and most_improved_delta is not None:
sent4_parts.append(
f"the most notable improvement was seen in {most_improved_country}, "
f"which gained {_fmt_delta(most_improved_delta)} points from the previous year"
)
sent4_parts.append(f"Biggest gain: **{most_improved_country}** ({_fmt_delta(most_improved_delta)} pts)")
if most_declined_country and most_declined_delta is not None:
if most_declined_delta < 0:
sent4_parts.append(
f"while {most_declined_country} experienced the largest decline "
f"of {_fmt_delta(most_declined_delta)} points"
)
else:
sent4_parts.append(
f"while {most_declined_country} recorded the smallest gain "
f"of {_fmt_delta(most_declined_delta)} points"
)
sent4 = ""
if sent4_parts:
sent4 = ", ".join(sent4_parts) + "."
sent4_parts.append(f"biggest drop: **{most_declined_country}** ({_fmt_delta(most_declined_delta)} pts)")
sent4 = ("; ".join(sent4_parts) + ".") if sent4_parts else ""
if sent4:
sent4 = sent4[0].upper() + sent4[1:]
return " ".join(s for s in [sent1, sent2, sent3, sent4] if s)
@@ -351,70 +296,55 @@ def _build_pillar_narrative(
most_declined_pillar,
most_declined_delta,
) -> str:
"""
Narrative format (no em-dash):
In {year}, {pillar} ranked {rank}/{n} with score {score}, {up/down} {delta} pts YoY.
Top country: {top_country}; bottom: {bot_country}.
Strongest pillar: {pillar}; weakest: {pillar}.
"""
rank_suffix = {1: "st", 2: "nd", 3: "rd"}.get(rank_in_year, "th")
# Sentence 1: rank + score + YoY
if yoy_val is not None:
direction_word = "up" if yoy_val >= 0 else "down"
yoy_clause = f", {direction_word} **{abs(yoy_val):.2f} pts** YoY"
else:
yoy_clause = ", no prior-year data"
sent1 = (
f"In {year}, the {pillar_name} pillar scored {_fmt_score(pillar_score)}, "
f"ranking {rank_in_year}{rank_suffix} out of {n_pillars} pillars assessed across ASEAN."
f"In **{year}**, **{pillar_name}** ranked **{rank_in_year}{rank_suffix}/{n_pillars}** "
f"with score **{_fmt_score(pillar_score)}**{yoy_clause}."
)
# Sentence 2: top / bottom country
sent2 = ""
if strongest_pillar and weakest_pillar:
if strongest_pillar == pillar_name:
sent2 = (
f"This made {pillar_name} the strongest performing pillar in {year}, "
f"compared to the weakest pillar, {weakest_pillar}, "
f"which scored {_fmt_score(weakest_score)}."
)
elif weakest_pillar == pillar_name:
sent2 = (
f"This made {pillar_name} the weakest performing pillar in {year}, "
f"compared to the strongest pillar, {strongest_pillar}, "
f"which scored {_fmt_score(strongest_score)}."
)
else:
sent2 = (
f"Across all pillars in {year}, {strongest_pillar} was the strongest "
f"(score: {_fmt_score(strongest_score)}), while {weakest_pillar} "
f"was the weakest (score: {_fmt_score(weakest_score)})."
)
sent3 = ""
if top_country and bot_country:
if top_country != bot_country:
sent3 = (
f"Within the {pillar_name} pillar, {top_country} led with a score of "
f"{_fmt_score(top_country_score)}, while {bot_country} recorded the lowest "
f"score of {_fmt_score(bot_country_score)}."
sent2 = (
f"Top country: **{top_country}** ({_fmt_score(top_country_score)}); "
f"bottom: **{bot_country}** ({_fmt_score(bot_country_score)})."
)
else:
sent3 = (
f"Within the {pillar_name} pillar, {top_country} was the only country "
f"with available data, scoring {_fmt_score(top_country_score)}."
)
sent2 = f"**{top_country}** was the only country with data ({_fmt_score(top_country_score)})."
if yoy_val is not None:
direction_word = "improved" if yoy_val >= 0 else "declined"
sent4 = (
f"Compared to the previous year, the {pillar_name} pillar "
f"{direction_word} by {abs(yoy_val):.2f} points"
)
else:
sent4 = (
f"No prior-year data is available to calculate year-over-year change "
f"for the {pillar_name} pillar in {year}"
# Sentence 3: strongest / weakest pillar
sent3 = ""
if strongest_pillar and weakest_pillar:
sent3 = (
f"Strongest pillar: **{strongest_pillar}** ({_fmt_score(strongest_score)}); "
f"weakest: **{weakest_pillar}** ({_fmt_score(weakest_score)})."
)
if (most_improved_pillar and most_improved_delta is not None
and most_declined_pillar and most_declined_delta is not None
and most_improved_pillar != most_declined_pillar):
sent4 += (
f". Across all pillars, {most_improved_pillar} showed the greatest improvement "
f"({_fmt_delta(most_improved_delta)} pts), while {most_declined_pillar} "
f"recorded the largest decline ({_fmt_delta(most_declined_delta)} pts)"
)
sent4 += "."
sent4 = sent4[0].upper() + sent4[1:]
# Sentence 4: most improved / declined pillar
sent4_parts = []
if most_improved_pillar and most_improved_delta is not None:
sent4_parts.append(f"Best gain: **{most_improved_pillar}** ({_fmt_delta(most_improved_delta)} pts)")
if most_declined_pillar and most_declined_delta is not None:
sent4_parts.append(f"largest drop: **{most_declined_pillar}** ({_fmt_delta(most_declined_delta)} pts)")
sent4 = ("; ".join(sent4_parts) + ".") if sent4_parts else ""
if sent4:
sent4 = sent4[0].upper() + sent4[1:]
return " ".join(s for s in [sent1, sent2, sent3, sent4] if s)
@@ -610,10 +540,6 @@ class FoodSecurityAggregator:
# =========================================================================
# METADATA BUILDER
# Menyesuaikan dengan signature: save_etl_metadata(client, metadata: dict)
# dan skema etl_metadata: source_class, table_name, execution_timestamp,
# duration_seconds, rows_fetched, rows_transformed, rows_loaded,
# completeness_pct, config_snapshot, validation_metrics
# =========================================================================
def _build_etl_metadata(
@@ -1419,50 +1345,7 @@ class FoodSecurityAggregator:
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 _build_etl_metadata(
self,
table_name: str,
rows_loaded: int,
start_time: datetime,
end_time: datetime,
status: str,
error_msg: str = None,
) -> dict:
"""
Susun dict metadata sesuai signature save_etl_metadata(client, metadata: dict)
dan kolom skema etl_metadata di bigquery_helpers.py:
source_class, table_name, execution_timestamp, duration_seconds,
rows_fetched, rows_transformed, rows_loaded, completeness_pct,
config_snapshot, validation_metrics
"""
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 "",
}),
}
def _finalize(self, table_name: str, rows_loaded: int):
"""
Tandai tabel sukses. Catat ke etl_logs dan etl_metadata.
Pemanggilan: save_etl_metadata(client, metadata_dict)
"""
end_time = datetime.now()
start_time = self.load_metadata[table_name].get("start_time")
@@ -1486,7 +1369,6 @@ class FoodSecurityAggregator:
)
)
except Exception as meta_err:
# Error metadata tidak boleh menghentikan pipeline
self.logger.warning(
f" [METADATA WARNING] Gagal simpan etl_metadata untuk {table_name}: {meta_err}"
)
@@ -1494,10 +1376,6 @@ class FoodSecurityAggregator:
self.logger.info(f" [OK] {table_name}: {rows_loaded:,} rows -> [Gold] fs_asean_gold")
def _fail(self, table_name: str, error: Exception):
"""
Tandai tabel gagal. Catat ke etl_logs dan etl_metadata.
Pemanggilan: save_etl_metadata(client, metadata_dict)
"""
end_time = datetime.now()
start_time = self.load_metadata[table_name].get("start_time")
error_msg = str(error)
@@ -1579,10 +1457,6 @@ class FoodSecurityAggregator:
# =============================================================================
def run_aggregation():
"""
Airflow task: Hitung semua agregasi dari fact_asean_food_security_selected.
Dipanggil setelah analytical_layer_to_gold selesai.
"""
from scripts.bigquery_config import get_bigquery_client
client = get_bigquery_client()
agg = FoodSecurityAggregator(client)