comexpy

Access the Brazilian Foreign Trade Statistics API (ComexStat) from Python.

comexpy is a pandas-friendly interface to the ComexStat API of the Brazilian Ministry of Development, Industry, Trade and Services (MDIC). It provides programmatic access to detailed Brazilian export and import data — every query returns a pandas.DataFrame.

This is the Python port of the R package comexr; the public function names mirror the R API so knowledge transfers directly between the two.

Features

• General trade data (1997–present), city-level data, and historical records (1989–1996)
• Auxiliary tables: countries, economic blocs, NCM/NBM/HS product codes, CGCE/SITC/ISIC classifications, states, cities, transport modes, customs units
• Friendly aliases — pass "hs4", "transport_mode", "cgce_n1" and the package translates them to the API's internal names
• Multilingual responses: Portuguese, English, Spanish
• SSL auto-fallback — handles ICP-Brasil certificate issues transparently
• Configurable retry/timeout for the API's aggressive rate limiting

Installation

pip install comexpy

Requires Python 3.9+, requests and pandas.

Quick start

import comexpy

# Exports by country in January 2024
exports = comexpy.comex_export(
    start_period="2024-01",
    end_period="2024-01",
    details="country",
)

# Imports with CIF value, by country, for all of 2024
imports = comexpy.comex_import(
    start_period="2024-01",
    end_period="2024-12",
    details="country",
    metric_cif=True,
)

# Exports to China (160), grouped by HS4
# (the package translates "hs4" to the API's `heading`)
soy = comexpy.comex_export(
    start_period="2024-01",
    end_period="2024-12",
    details=["country", "hs4"],
    filters={"country": 160},
)

Discover available options

comexpy.comex_details("general")     # grouping fields available
comexpy.comex_filters("general")     # filters available
comexpy.comex_metrics("general")     # metrics available

# Look up country codes
countries = comexpy.comex_countries()
countries[countries["text"].str.contains("China", case=False)]

# Economic blocs in Portuguese
comexpy.comex_blocs(language="pt")

City-level and historical data

# Exports declared in Pernambuco (state 26) in 2023
comexpy.comex_query_city(
    flow="export",
    start_period="2023-01",
    end_period="2023-12",
    details=["country", "state"],
    filters={"state": 26},
)

# Historical exports 1995–1996 by country (NBM classification)
comexpy.comex_historical(
    flow="export",
    start_period="1995-01",
    end_period="1996-12",
    details="country",
)

Rate limiting and timeouts

The ComexStat API frequently returns rate-limit errors (HTTP 429, "Você excedeu o limite de solicitações...") or times out. Adjust the behaviour with set_options:

comexpy.set_options(retry_time=30)   # wait 30s between retries
comexpy.set_options(max_tries=5, timeout_post=180)

Silence progress/success messages with comexpy.set_verbose(False).

Public API

License

MIT — see LICENSE. Developed by the comexpy authors.