data-sources

name	data-sources
description	Fetch options/futures market data from supported exchanges (Deribit, Binance, MOEX) with alphavar — snapshot the options book, resolve instrument kinds, and understand each venue's symbol scheme and field→column mapping. Use when a user asks to pull market data from a venue, normalize exchange fields, or parse an instrument symbol.
when_to_use	Use when a user asks to pull market data from a venue, normalize exchange fields, or parse an instrument symbol.
owner	alphavar

Fetch & normalize exchange data in alphavar

USAGE skill (concept → function → how to apply). Verified against src/ — venue APIs drift, so re-confirm both the live docs and the in-repo class before relying on specifics.

Concept

alphavar reads market data from several venues through one abstraction, then normalizes venue fields to its own columns: the venue's raw values are exch_*, the normalized model values are price/iv (R4 / dictionary v2). Each venue has its own host(s), symbol scheme, and field names.

Implementing classes

All extend alphavar.io.exchange._abstract_exchange.AbstractExchange, which exposes the shared facade:

• get_options_assets_books_snapshot(asset_codes=None) → pd.DataFrame — the options book.
• resolve_instrument_kind(native_kind) → (InstrumentKind, ContractKind) | None.
• request_api(endpoint_path, signed=False, **kwargs).

Venue	Class	Notes (verify in the class)
Deribit (crypto, European)	alphavar.io.exchange.deribit.DeribitExchange	API v2 REST, base https://www.deribit.com; symbol {CCY}-{DDMMMYY}-{STRIKE}-{C|P}; mark_price→exch_mark_price, creation_timestamp→exch_timestamp
Binance (crypto, European opts)	alphavar.io.exchange.binance.BinanceExchange	per-product hosts (options host https://eapi.binance.com); symbol {ASSET}-{YYMMDD}-{STRIKE}-{C|P}; primarily a spot/underlying source today
MOEX (FORTS opts & futures)	alphavar.io.exchange.moex.MoexExchange	ISS REST, base https://iss.moex.com/iss; theorprice→exch_mark_price, updatetime→exch_timestamp; groups by base_asset_code

How to apply

from alphavar.io.exchange.deribit import DeribitExchange

ex = DeribitExchange(...)                              # see the class for construction
df = ex.get_options_assets_books_snapshot(["BTC"])     # normalized options book

(Binance / MOEX follow the same facade — swap the class.)

Conventions & failure modes

• Normalized vs raw: use price/iv for model logic; exch_* are the venue's raw values (different conventions/units per venue).
• Symbol parsing differs per venue — a parse failure raises InstrumentParseError (alphavar.io.exchange.exchange_exception); don't assume one scheme across venues.
• Volume / drift: a snapshot can be thousands of instruments (vectorized parsing); endpoint params and rate limits change — verify against live docs, not memory.
• Recording fixtures for tests is a dev concern, not USAGE — see ../../_aitna/skills/refresh-exchange-fixtures.md.