Market Data

data is a BarData object passed to handle_data, scheduled callbacks and before_trading_start. Every read is async and returns a Polars DataFrame with date, sid and the fields you asked for.

current = await data.current(assets=[asset], fields=["close"])
history = await data.history(assets=[asset], fields=["close"], bar_count=20)

Current values

`await data.current(assets, fields, data_source=None)`

The latest available values at the current simulation time.

df = await data.current(
    assets=[context.asset],
    fields=["open", "high", "low", "close", "volume"],
)

close = df["close"][0]
volume = df["volume"][0]

For several assets, map rows back by sid:

df = await data.current(assets=context.assets, fields=["close"])
prices = dict(zip(df["sid"].to_list(), df["close"].to_list()))
price = prices.get(context.assets[0].sid)

Historical windows

`await data.history(assets, bar_count=None, frequency=timedelta(days=1), fields=None, data_source=None, since=None)`

A trailing window ending before the current bar, adjusted for splits and dividends as of the current simulation time. Pass exactly one of bar_count or since:

bar_count — a number of rows. On bar data that is a number of sessions.
since — a span of calendar time, such as since=datetime.timedelta(days=90). Event datasets usually want this: thirty congressional disclosures for one ticker can span four years.

import datetime

hist = await data.history(
    assets=[context.asset],
    fields=["close"],
    bar_count=50,
    frequency=datetime.timedelta(days=1),
)
closes = hist["close"].to_numpy()
sma_50 = closes.mean()

Because the window excludes the current bar, read the current value separately when the signal needs it:

hist = await data.history(assets=[context.asset], fields=["close"], bar_count=19)
current = await data.current(assets=[context.asset], fields=["close"])
window = list(hist["close"]) + [current["close"][0]]

Ask for enough bars to warm every indicator up, and return early when the window is short:

if len(hist) < 50:
    return

Returns across a universe

`await data.returns(assets, bar_count, field="close", data_source=None)`

What each asset returned over the last bar_count bars, as {listing: last / first - 1}. One history read serves the whole universe. Listings that can’t answer — not trading yet, no data, all zeros — are simply absent.

universe = await context.universe_symbols("Q500US")
momentum = await data.returns(assets=universe, bar_count=126)
best = sorted(momentum, key=momentum.get, reverse=True)[:10]
await context.rebalance({listing: 0.1 for listing in best})

Fields

Field	Meaning
`"open"`, `"high"`, `"low"`, `"close"`	Bar prices
`"volume"`	Bar volume
`"price"`	Last known price, when the data source supports it
`"last_traded"`	Last traded datetime, when the data source supports it

Custom sources expose their own fields. Name the source with data_source:

fundamentals = await data.history(
    assets=[context.asset],
    fields=["market_cap", "pe_ratio"],
    since=datetime.timedelta(days=365),
    data_source="limex_us_fundamental_data",
)

data_source also accepts a mounted dataset or an address such as "hf://ZipLime/congress-trading/features" — see Point-in-Time Datasets. Extensions can register their own address schemes; zipfinam adds algopack:// for MOEX AlgoPack.

Working with Polars

import polars as pl

hist = await data.history(assets=context.assets, fields=["close", "volume"], bar_count=20)

summary = hist.group_by("sid").agg(
    pl.col("close").mean().alias("mean_close"),
    pl.col("volume").sum().alias("total_volume"),
)

.iloc, .loc, .at and .iat don’t exist on a Polars frame. Use df["close"][-1], df["close"].to_numpy(), df["close"].mean(), or call .to_pandas() explicitly.

Futures

Method	Returns
`await data.current_contract(continuous_future)`	The contract a continuous future holds right now — this is what you order
`await data.current_chain(continuous_future)`	Active contracts of the chain, front first

A continuous future is a data specifier. Ordering one raises.

Tradability

`data.can_trade(assets)`

Returns a pandas Series of booleans indexed by asset. An asset can trade when it is listed and not yet delisted or auto-closed, is not restricted, and its venue is open. Since 2.11.0 the venue check uses the listing’s own venue calendar: on a daily run it asks whether the venue traded that day; on an intraday run, whether it is open that minute. See Venues & Calendars.

tradeable = data.can_trade(context.assets)
open_now = [asset for asset in context.assets if tradeable[asset]]

can_trade is synchronous and does not check that a price exists. Read the price with data.current before ordering, and skip the asset when it is missing or zero.

Data clock helpers

Property	Returns
`data.current_dt`	Current simulation datetime
`data.current_session`	Current trading session label

Introduction

Writing Strategies

Strategy Language

Migrating to Ziplime

Limitations

Current values

`await data.current(assets, fields, data_source=None)`

Historical windows

`await data.history(assets, bar_count=None, frequency=timedelta(days=1), fields=None, data_source=None, since=None)`

Returns across a universe

`await data.returns(assets, bar_count, field="close", data_source=None)`

Fields

Working with Polars

Futures

Tradability

`data.can_trade(assets)`

Data clock helpers

​Current values

​await data.current(assets, fields, data_source=None)

​Historical windows

​await data.history(assets, bar_count=None, frequency=timedelta(days=1), fields=None, data_source=None, since=None)

​Returns across a universe

​await data.returns(assets, bar_count, field="close", data_source=None)

​Fields

​Working with Polars

​Futures

​Tradability

​data.can_trade(assets)

​Data clock helpers

Current values

`await data.current(assets, fields, data_source=None)`

Historical windows

`await data.history(assets, bar_count=None, frequency=timedelta(days=1), fields=None, data_source=None, since=None)`

Returns across a universe

`await data.returns(assets, bar_count, field="close", data_source=None)`

Fields

Working with Polars

Futures

Tradability

`data.can_trade(assets)`

Data clock helpers