Query market data in Python

TL;DR

Use a lake-scoped API key and the optional Python lake dependencies to run bounded read-only SQL and retain its durable Parquet results. The worked query retrieves five AAPL sessions from January 2024, preserving adjusted prices and source timestamps. It was verified through the canonical lake reader on October 10, 2026; that receipt is separate from the SDK's query-operation transport. Save the operation, manifest and checked row count before analysis.

Follow one worked path

Step 1 of 3

Install the shown SDK version

Use a registered API key with lake scope. Install nexustrade[lake] for the DuckDB and pandas analysis helpers. The base SDK is standard-library-only; these analysis dependencies are optional.

What you should see

The matching package is available and the request uses the appropriate read, write or lake scope.

Shell

Copy command

python3 -m pip install 'nexustrade[lake]==1.42.0'

Step 2 of 3

Run the bounded query

Run the exact bounded SQL example below. Preserve the adjusted-price basis and returned session timestamps.

What you should see

The durable query job returns schema and parts; read the actual returned values before treating them as market observations.

Python

Copy Python example

import nexustrade as nt

result = nt.lake.sql(
    "SELECT ticker, date, closingPrice FROM lake.sec_daily_ohlc "
    "WHERE ticker = ? AND date::DATE BETWEEN CAST(? AS DATE) AND CAST(? AS DATE) "
    "ORDER BY date ASC LIMIT 5",
    ["AAPL", "2024-01-02", "2024-01-08"],
    max_rows=5,
)
frame = result.to_pandas()
print(frame)
for batch in result.iter_batches():
    print(batch)

Step 3 of 3

Run the query and read its results

This query asks for at most five AAPL sessions from January 2 through January 8, 2024, sorted oldest first. Both the SQL and result budget are bounded. The catalog recommends sec_daily_ohlc for canonical equity and ETF history; daily_ohlc is the legacy vendor table. Results are durable Parquet parts.

AAPL

Source timestamp (UTC): 2024-01-02T21:00:00.000Z · closingPrice (USD): 185.64

AAPL

Source timestamp (UTC): 2024-01-03T21:00:00.000Z · closingPrice (USD): 184.25

AAPL

Source timestamp (UTC): 2024-01-04T21:00:00.000Z · closingPrice (USD): 181.91

AAPL

Source timestamp (UTC): 2024-01-05T21:00:00.000Z · closingPrice (USD): 181.18

AAPL

Source timestamp (UTC): 2024-01-08T21:00:00.000Z · closingPrice (USD): 185.56

What you should see

The shown product receipt demonstrates its originating workflow, not proof that your local SDK request has executed.

Actual output from the canonical lake reader

We executed the exact bounded SQL below on the deployed NexusTrade lake worker on October 10, 2026. It returned these five AAPL sessions in 438 milliseconds, including setup. This verifies the canonical source query and returned values; the request used the synchronous lake reader rather than the SDK's authenticated query-operation transport.

The date values are the source's session timestamps. closingPrice is the catalog's adjusted series, so preserve that basis when comparing it with an unadjusted quote or deriving market capitalization.

Install the lake extra

python3 -m pip install 'nexustrade[lake]==1.42.0'

Run the query and read its results

This query asks for at most five AAPL sessions from January 2 through January 8, 2024, sorted oldest first. Both the SQL and result budget are bounded. The catalog recommends sec_daily_ohlc for canonical equity and ETF history; daily_ohlc is the legacy vendor table. Results are durable Parquet parts.

import nexustrade as nt

result = nt.lake.sql(
    "SELECT ticker, date, closingPrice FROM lake.sec_daily_ohlc "
    "WHERE ticker = ? AND date::DATE BETWEEN CAST(? AS DATE) AND CAST(? AS DATE) "
    "ORDER BY date ASC LIMIT 5",
    ["AAPL", "2024-01-02", "2024-01-08"],
    max_rows=5,
)
frame = result.to_pandas()
print(frame)
for batch in result.iter_batches():
    print(batch)

Save the query and result manifest

Continue exploring