File size: 19,738 Bytes
64f689c da250a5 64f689c da250a5 64f689c | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 | #!/usr/bin/env python3
"""
Phase 4 — Variant Assignment / Variant Modeling
Compliments Reference DB Pipeline
Input: Phase 3 outputs + Phase 2 output
Output: Variant tables and mappings
This phase transforms the Phase 3 product-group representation into a
structured product variant model, where products belonging to the same
reference product group can have different purchasable variants such as
size, package quantity, formulation, flavour, fat level, etc.
CRITICAL: This phase does NOT use nutrition data.
CRITICAL: This phase does NOT use any external taxonomy.
CRITICAL: This phase is 100% deterministic (no LLM).
"""
import json
import hashlib
import sys
from datetime import datetime, timezone
from pathlib import Path
import pandas as pd
import numpy as np
# ---------------------------------------------------------------------------
# Configuration
# ---------------------------------------------------------------------------
BASE_DIR = Path(__file__).resolve().parent.parent
PHASE3_DIR = BASE_DIR.parent / "phase3"
PHASE2_DIR = BASE_DIR.parent / "phase2"
OUTPUT_DIR = BASE_DIR / "outputs"
VALIDATION_DIR = BASE_DIR / "validation"
STATISTICS_DIR = BASE_DIR / "statistics"
AUDIT_DIR = BASE_DIR / "audit"
VERSION = "1.0.0"
TIMESTAMP = datetime.now(timezone.utc).isoformat()
def log(msg: str) -> None:
print(f"[Phase4] {msg}")
# ---------------------------------------------------------------------------
# 1. VARIANT DEFINITION
# ---------------------------------------------------------------------------
# A variant is a purchasable representation of the same reference product
# that differs in attributes such as:
# - Size (500 g, 1 kg, 2 kg)
# - Package quantity (1 × 500 g, 2 × 500 g)
# - Pack configuration (single, multipack, case)
#
# Variants within the same group share the same identity attributes:
# - is_organic, is_gluten_free, is_naturally_simple, etc.
# - product_line
# - flavour
# - formulation
# - fat_level
#
# Variants differ only in size/package attributes.
# ---------------------------------------------------------------------------
# 2. VARIANT KEY GENERATION
# ---------------------------------------------------------------------------
def build_variant_key(row: dict) -> str:
"""
Build a deterministic variant key from normalized variant attributes.
The variant key captures size/package information that distinguishes
variants within the same reference product group.
Structure: amount|unit|qty|count
"""
va_str = row.get("variant_attributes", "{}")
if isinstance(va_str, str):
try:
va = json.loads(va_str)
except (json.JSONDecodeError, TypeError):
va = {}
elif isinstance(va_str, dict):
va = va_str
else:
va = {}
# Extract variant-distinguishing attributes
amount = va.get("amount")
unit = va.get("unit", "")
qty = va.get("qty", 1)
count = va.get("count")
multiplier = va.get("multiplier", 1)
raw = va.get("raw", "")
# Build key parts
parts = []
if count is not None:
# Count-based variant (e.g., "20 per pack")
parts.append(f"count{int(count)}")
elif amount is not None:
# Size-based variant
parts.append(f"amt{amount}")
if unit:
parts.append(f"unit{unit}")
if qty and qty > 1:
parts.append(f"qty{int(qty)}")
if multiplier and multiplier > 1:
parts.append(f"mult{int(multiplier)}")
elif raw:
# Fallback to raw string
parts.append(f"raw{raw.strip().lower()}")
else:
# No size information
parts.append("nosize")
return "|".join(parts)
def generate_variant_id(group_id: str, variant_key: str) -> str:
"""
Generate deterministic variant_id from group_id + variant_key.
Uses UUID5 (namespace-based) for reproducibility.
"""
import uuid
namespace = uuid.NAMESPACE_DNS
name = f"{group_id}||{variant_key}"
return str(uuid.uuid5(namespace, name))
# ---------------------------------------------------------------------------
# 3. VARIANT ASSIGNMENT
# ---------------------------------------------------------------------------
def assign_variants(df: pd.DataFrame) -> pd.DataFrame:
"""
Assign variant_ids to products within each group.
Products in the same group with the same variant_key get the same variant_id.
Products in the same group with different variant_keys get different variant_ids.
"""
log("Building variant keys ...")
df["variant_key"] = df.apply(lambda row: build_variant_key(row.to_dict()), axis=1)
log("Generating variant IDs ...")
df["variant_id"] = df.apply(
lambda row: generate_variant_id(row["group_id"], row["variant_key"]),
axis=1
)
n_variants = df["variant_id"].nunique()
log(f"Total unique variants: {n_variants}")
return df
# ---------------------------------------------------------------------------
# 4. OUTPUT GENERATION
# ---------------------------------------------------------------------------
def build_variant_table(df: pd.DataFrame) -> pd.DataFrame:
"""
Build the main variant table: one row per unique variant.
"""
log("Building variant table ...")
# Group by variant_id and aggregate
variant_rows = []
for vid, grp in df.groupby("variant_id"):
row = grp.iloc[0]
# Get all sizes in this variant
sizes = grp["size"].dropna().unique().tolist()
variant_rows.append({
"variant_id": vid,
"group_id": row["group_id"],
"group_name": row["group_name"],
"brand": row["brand"],
"product_line": row["product_line"],
"product_domain": row["product_domain"],
"reference_db_taxonomy": row["reference_db_taxonomy"],
"core_title": row["core_title"],
"variant_key": row["variant_key"],
"size": sizes[0] if len(sizes) == 1 else "; ".join(str(s) for s in sizes),
"size_amount": row.get("size_amount"),
"size_unit": row.get("size_unit"),
"size_qty": row.get("size_qty"),
"flavour": row.get("flavour"),
"formulation": row.get("formulation"),
"fat_level": row.get("fat_level"),
"fat_percentage": row.get("fat_percentage"),
"product_count": len(grp),
"unique_upcs": grp["upc"].nunique(),
"source": row["source"],
"source_url": row["source_url"],
})
variant_df = pd.DataFrame(variant_rows)
log(f" Variant table: {len(variant_df)} variants")
return variant_df
def build_variant_mapping(df: pd.DataFrame) -> pd.DataFrame:
"""
Build the product → variant mapping table.
"""
log("Building variant mapping ...")
mapping = df[[
"external_id", "upc", "group_id", "variant_id",
"core_title", "original_title", "brand", "size",
"variant_key", "source", "source_url"
]].copy()
log(f" Variant mapping: {len(mapping)} products")
return mapping
def build_variant_summary(df: pd.DataFrame) -> pd.DataFrame:
"""
Build the group-level variant summary.
"""
log("Building variant summary ...")
summary_rows = []
for gid, grp in df.groupby("group_id"):
variant_ids = grp["variant_id"].unique().tolist()
row = grp.iloc[0]
summary_rows.append({
"group_id": gid,
"group_name": row["group_name"],
"brand": row["brand"],
"product_domain": row["product_domain"],
"reference_db_taxonomy": row["reference_db_taxonomy"],
"variant_count": len(variant_ids),
"variant_ids": ", ".join(variant_ids),
"product_count": len(grp),
})
summary_df = pd.DataFrame(summary_rows)
log(f" Variant summary: {len(summary_df)} groups")
return summary_df
# ---------------------------------------------------------------------------
# 5. VALIDATION
# ---------------------------------------------------------------------------
def validate_phase4(df: pd.DataFrame, variant_table: pd.DataFrame,
variant_mapping: pd.DataFrame, variant_summary: pd.DataFrame) -> dict:
"""Comprehensive Phase 4 validation."""
checks = {}
# Rule 1: Every product has exactly one variant
checks["rule_1_product_has_one_variant"] = {
"description": "Every product has exactly one variant_id",
"products": len(df),
"unique_variant_ids_per_product": int(df.groupby("external_id")["variant_id"].nunique().max()),
"pass": int(df.groupby("external_id")["variant_id"].nunique().max()) == 1,
}
# Rule 2: Every variant belongs to exactly one group
checks["rule_2_variant_belongs_to_one_group"] = {
"description": "Every variant belongs to exactly one group_id",
"variants": len(variant_table),
"unique_groups_per_variant": int(variant_table.groupby("variant_id")["group_id"].nunique().max()),
"pass": int(variant_table.groupby("variant_id")["group_id"].nunique().max()) == 1,
}
# Rule 3: No product maps to multiple variants
checks["rule_3_no_product_multiple_variants"] = {
"description": "No product maps to multiple variant_ids",
"products_with_multiple_variants": int((df.groupby("external_id")["variant_id"].nunique() > 1).sum()),
"pass": int((df.groupby("external_id")["variant_id"].nunique() > 1).sum()) == 0,
}
# Rule 4: Variant count matches
checks["rule_4_variant_count"] = {
"description": "Variant count in mapping equals variant table",
"mapping_variants": int(variant_mapping["variant_id"].nunique()),
"table_variants": len(variant_table),
"pass": int(variant_mapping["variant_id"].nunique()) == len(variant_table),
}
# Rule 5: All products have group_id
checks["rule_5_all_products_have_group"] = {
"description": "All products have group_id",
"null_group_ids": int(df["group_id"].isna().sum()),
"pass": int(df["group_id"].isna().sum()) == 0,
}
# Rule 6: All products have variant_id
checks["rule_6_all_products_have_variant"] = {
"description": "All products have variant_id",
"null_variant_ids": int(df["variant_id"].isna().sum()),
"pass": int(df["variant_id"].isna().sum()) == 0,
}
# Overall
all_pass = all(c.get("pass", True) for c in checks.values())
checks["overall"] = {"result": "PASS" if all_pass else "FAIL"}
return checks
# ---------------------------------------------------------------------------
# 6. STATISTICS
# ---------------------------------------------------------------------------
def build_statistics(df: pd.DataFrame, variant_table: pd.DataFrame,
variant_mapping: pd.DataFrame, variant_summary: pd.DataFrame) -> dict:
"""Build Phase 4 statistics."""
# Variant count distribution
variant_counts = variant_summary["variant_count"].value_counts().sort_index()
return {
"version": VERSION,
"timestamp": TIMESTAMP,
"input": {
"source": "phase3_product_group_mapping.csv + phase2_output.parquet",
"products": len(df),
"groups": df["group_id"].nunique(),
},
"output": {
"variants": len(variant_table),
"products_mapped": len(variant_mapping),
"groups": len(variant_summary),
},
"variant_distribution": {
"groups_with_1_variant": int((variant_summary["variant_count"] == 1).sum()),
"groups_with_2_variants": int((variant_summary["variant_count"] == 2).sum()),
"groups_with_3_variants": int((variant_summary["variant_count"] == 3).sum()),
"groups_with_4plus_variants": int((variant_summary["variant_count"] >= 4).sum()),
"max_variants_per_group": int(variant_summary["variant_count"].max()),
"avg_variants_per_group": float(variant_summary["variant_count"].mean()),
},
"variant_count_distribution": {str(k): int(v) for k, v in variant_counts.items()},
"missing_size": {
"products_with_missing_size": int(df["size"].isna().sum()),
"products_with_no_size_variant": int((df["variant_key"] == "nosize").sum()),
},
}
# ---------------------------------------------------------------------------
# 7. PEANUT BUTTER EXAMPLE
# ---------------------------------------------------------------------------
def generate_peanut_butter_example(df: pd.DataFrame) -> pd.DataFrame:
"""Generate Peanut Butter variant example for audit."""
log("Generating Peanut Butter example ...")
# Find peanut butter products
pb = df[df["original_title"].str.contains("peanut butter", case=False, na=False)].copy()
# Select relevant columns
example = pb[[
"external_id", "upc", "group_id", "variant_id",
"original_title", "core_title", "brand", "size",
"variant_key", "variant_attributes", "product_line",
"identity_hash", "fat_level", "product_domain", "reference_db_taxonomy"
]].copy()
# Sort by group_id and size
example = example.sort_values(["group_id", "size"]).reset_index(drop=True)
log(f" Peanut Butter example: {len(example)} products in {example['group_id'].nunique()} groups")
return example
# ---------------------------------------------------------------------------
# MAIN
# ---------------------------------------------------------------------------
def main():
log("Starting Phase 4 (v1.0.0 — Variant Assignment)")
# Ensure output dirs exist
OUTPUT_DIR.mkdir(parents=True, exist_ok=True)
VALIDATION_DIR.mkdir(parents=True, exist_ok=True)
STATISTICS_DIR.mkdir(parents=True, exist_ok=True)
AUDIT_DIR.mkdir(parents=True, exist_ok=True)
# ------------------------------------------------------------------
# 1. LOAD PHASE 3 OUTPUTS
# ------------------------------------------------------------------
log("Loading Phase 3 outputs ...")
mapping_path = PHASE3_DIR / "outputs" / "product_group_mapping.csv"
mapping = pd.read_csv(mapping_path, dtype={"upc": "string"})
log(f" product_group_mapping: {len(mapping)} rows")
# UPC integrity validation
if "upc" in mapping.columns:
upc_col = mapping["upc"]
dot_zero_count = upc_col.dropna().astype(str).str.endswith(".0").sum()
if dot_zero_count > 0:
raise ValueError(f"UPC integrity check failed: {dot_zero_count} UPCs end with '.0'")
log(f" UPC integrity: PASS (no .0 suffixes, dtype={upc_col.dtype})")
# ------------------------------------------------------------------
# 2. LOAD PHASE 2 OUTPUT (for variant-relevant columns)
# ------------------------------------------------------------------
log("Loading Phase 2 output ...")
p2_path = PHASE2_DIR / "outputs" / "phase2_output.parquet"
p2 = pd.read_parquet(p2_path)
log(f" phase2_output: {len(p2)} rows, {len(p2.columns)} columns")
# ------------------------------------------------------------------
# 3. MERGE PHASE 3 MAPPING WITH PHASE 2 VARIANT DATA
# ------------------------------------------------------------------
log("Merging Phase 3 mapping with Phase 2 variant data ...")
variant_cols = ["external_id", "variant_attributes", "flavour", "formulation",
"fat_level", "fat_percentage", "size_amount", "size_unit", "size_qty"]
# Drop duplicate columns from Phase 3 mapping before merge to avoid _x/_y suffixes
overlap_cols = [c for c in variant_cols if c != "external_id" and c in mapping.columns]
mapping_clean = mapping.drop(columns=overlap_cols)
df = mapping_clean.merge(p2[variant_cols], on="external_id", how="left")
log(f" Merged: {len(df)} rows, {len(df.columns)} columns")
# ------------------------------------------------------------------
# 4. ASSIGN VARIANTS
# ------------------------------------------------------------------
df = assign_variants(df)
# ------------------------------------------------------------------
# 5. GENERATE OUTPUTS
# ------------------------------------------------------------------
variant_table = build_variant_table(df)
variant_mapping = build_variant_mapping(df)
variant_summary = build_variant_summary(df)
# ------------------------------------------------------------------
# 6. VALIDATION
# ------------------------------------------------------------------
log("Running validation ...")
validation = validate_phase4(df, variant_table, variant_mapping, variant_summary)
# ------------------------------------------------------------------
# 7. STATISTICS
# ------------------------------------------------------------------
log("Building statistics ...")
statistics = build_statistics(df, variant_table, variant_mapping, variant_summary)
# ------------------------------------------------------------------
# 8. PEANUT BUTTER EXAMPLE
# ------------------------------------------------------------------
pb_example = generate_peanut_butter_example(df)
# ------------------------------------------------------------------
# 9. SAVE OUTPUTS
# ------------------------------------------------------------------
log("Saving outputs ...")
# Main outputs (Parquet)
variant_table.to_parquet(OUTPUT_DIR / "reference_product_variants.parquet", index=False)
log(f" Saved reference_product_variants.parquet ({len(variant_table)} rows)")
variant_mapping.to_parquet(OUTPUT_DIR / "product_variant_mapping.parquet", index=False)
log(f" Saved product_variant_mapping.parquet ({len(variant_mapping)} rows)")
variant_summary.to_parquet(OUTPUT_DIR / "reference_product_variant_summary.parquet", index=False)
log(f" Saved reference_product_variant_summary.parquet ({len(variant_summary)} rows)")
# CSV versions for inspection
variant_table.to_csv(OUTPUT_DIR / "reference_product_variants.csv", index=False)
variant_mapping.to_csv(OUTPUT_DIR / "product_variant_mapping.csv", index=False)
variant_summary.to_csv(OUTPUT_DIR / "reference_product_variant_summary.csv", index=False)
# Validation
with open(VALIDATION_DIR / "phase4_validation.json", "w") as f:
json.dump(validation, f, indent=2, default=str)
log(" Saved phase4_validation.json")
# Statistics
with open(STATISTICS_DIR / "phase4_statistics.json", "w") as f:
json.dump(statistics, f, indent=2, default=str)
log(" Saved phase4_statistics.json")
# Peanut Butter example
pb_example.to_parquet(AUDIT_DIR / "phase4_peanut_butter_example.parquet", index=False)
log(f" Saved phase4_peanut_butter_example.parquet ({len(pb_example)} rows)")
# ------------------------------------------------------------------
# SUMMARY
# ------------------------------------------------------------------
log("")
log("=== PHASE 4 COMPLETE (v1.0.0) ===")
log(f"Input: {len(df)} products, {df['group_id'].nunique()} groups")
log(f"Output: {len(variant_table)} variants, {len(variant_mapping)} products mapped")
log(f"Validation: {validation['overall']['result']}")
log("========================")
return variant_table, variant_mapping, variant_summary, validation, statistics
if __name__ == "__main__":
main()
|