Validate Philippine Addresses in Python with PSGC

How to validate Philippine addresses against the official PSGC masterlist in Python using the barangay package. Fuzzy matching handles misspellings and unstandardized formats with validate() and validate_many().

Address validation is one of the most common reasons teams adopt the Philippine Standard Geographic Code (PSGC). Misspelled barangays, missing provinces, and inconsistent casing make user-entered Philippine addresses notoriously hard to clean. This post shows how to validate them in Python — fully offline — with the barangay package.

Why address validation is hard in the Philippines

• 42,010 barangays with similar or duplicated names across provinces.
• Users often omit the municipality or use abbreviations (“San”, “Sta.”).
• Free-text fields mix barangay, city, and province in arbitrary order.
• Spellings drift (“Tongmageng” vs “Tongmagen”).

Install

pip install barangay

Validate a single address

from barangay import validate

result = validate("Tongmageng, Sitangkai, Tawi-Tawi")
print(result.valid, result.matched_name, result.score)
# True Tongmageng 100.0

validate() returns a ValidationResult with valid, matched_name, matched_psgc_id, and a fuzzy score (0–100). The default threshold is 95.0.

Validate in bulk

from barangay import validate_many

addresses = [
    "Tongmageng, Tawi-Tawi",
    "Bagumbayan, Quezon City",
    "Nonexistent Place, Nowhere",
]
for r in validate_many(addresses):
    print(f"{r.input!r} -> {'valid' if r.valid else 'invalid'} ({r.score})")

Tuning the threshold

Lower the threshold to catch more approximate matches, or raise it to be stricter:

validate("Tongmagen, Tawi-Tawi", threshold=80.0)

Why do it offline?

barangay bundles the complete PSGC dataset, so validation runs in-process with no API calls, rate limits, or network dependency — perfect for batch cleaning of large datasets, ETL pipelines, and CI checks.

---

Read the full address validation tutorial or browse the validation API reference.