agmarknet 0.1.0

Pythonic pandas wrapper for the official AGMARKNET REST API.

agmarknet

Python SDK for the official AGMARKNET REST API.

The SDK supports:

• Fetching filter lookup tables.
• Fetching daily price and arrival reports.
• Returning pandas DataFrames.
• Using requests.Session under the hood.
• Resolving commodity, state, district, market, grade, and variety names from the official filters endpoint.
• Exact, case-insensitive, and fuzzy lookup matching.

from agmarknet import Agmarknet

api = Agmarknet()

filters = api.filters()

report = api.report(
    from_date="2021-01-01",
    to_date="2021-12-31",
    commodity="Tomato",
    state="Karnataka",
    district="Kolar",
    market="Bangarpet APMC",
)

Numeric AGMARKNET IDs still work for users who already have them. If group is omitted, the SDK infers it from the resolved commodity whenever the filters payload includes the commodity group ID.

api.filters() returns a FilterTables object. You can use attributes, friendly keys, or raw AGMARKNET keys:

filters = api.filters()

len(filters.commodities)
len(filters["commodities"])
len(filters["cmdt_data"])

Reports automatically fetch all API pages and return one DataFrame. The default report type is price, matching the live AGMARKNET payload. Use data_type="arrival" or data_type="both" when needed.

Long historical ranges can be requested with start and end. The SDK splits the range into yearly chunks, downloads them sequentially, removes duplicate rows, and preserves chronological order.

history = api.report(
    start="2007-01-01",
    end="2026-07-08",
    commodity="Tomato",
    state="Karnataka",
    district="Kolar",
    market="Bangarpet APMC",
    progress=True,
)

Downloads, caching, async support, and CLI features are planned for later roadmap phases.