algoseek-connector 1.0.2

A wrapper library for ORM-like SQL builder and executor

algoseek-connector

A wrapper library for ORM-like SQL builder and executor. The library provides a simple pythonic interface to algoseek datasets with custom data filtering/selection.

Installing and Supported Versions

algoseek-connector is available on PyPI:

$ python -m pip install algoseek-connector

or alternatively

$ pip install algoseek-connector

Python versions 3.6+ are supported.

Supported Features

The following query operations on datasets are supported:

• Selecting columns and arbitrary expressions based on columns
• Filtering by column value/column expression
• Grouping by column(s)
• Sorting by column(s)
• All common artithmetic, logical operations on dataset columns and function application
• Fetching query results as a pandas dataframe

Getting Started

Creating a session

A database connection is created with a Session object with the DB host, username and password provided.

import algoseek_connector as aconnect

host = '123.123.123.123'
user = 'demo'
password ='secret-password-2000'

session = aconnect.Session(host, user, password)

Optionally a port number is provided unless it is a default value of 9000.

Configuring a session with environment variables

You can make use of the following environment variables to set up the databse connection:

• AS_DATABASE_HOST
• AS_DATABASE_PORT
• AS_DATABASE_USER
• AS_DATABASE_PASSWORD

In this case an empty session is created with user credentials read from the environment.

session = aconnect.Session()

Executing raw queries

A Session object can be used to execute a SQL query directly

session.execute('''
SELECT * FROM USEquityMarketData.TradeOnly
WHERE Ticker = 'IBM'
LIMIT 10''')

Datagroups and datasets

All datasets available are grouped into data groups Is structured into datagroups, e.g. USEquityMarketData, USFuturesMarketData, etc.

You can browse the list of available data groups with the DataResource

resource = DataResource(session)
for dgr in resource.datagroups.all():
    print(dgr.name)

Similarly, you can access the list of datasets of a specific data group:

datagroup = resource.datagroup('USEquityMarketData')
for dts in datagroup.datasets.all():
    print(dts.name)

Alternatively, getting a specific dataset directly:

dataset = aconnect.Dataset(
    'USEquityMarketData', 'TradeOnlyMinuteBar', session=session
)

Selecting a subset of columns

To get specific columns the Dataset.select method is used:

ds = aconnect.Dataset(
    'USEquityMarketData', 'TradeOnly', session=session
)
ds.select(
    ds.EventDateTime, ds.Ticker, ds.Price
).head()

Dataset filtering

Filtering expressions can be chained using & (AND) and | (OR) operators, a ~ is used for negation (NOT).

ds = aconnect.Dataset(
    'USEquityMarketData', 'TradeOnly', session=session
)
ds.select(
    ds.EventDateTime, ds.Ticker, ds.Price
).filter(
    ds.TradeDate.between('2022-01-01', '2022-01-31') &
    (ds.Ticker = 'TSLA') &
    (ds.Quantity < 100)
).head()

Getting results

You can make use of Dataset.fetch method to execute the generated query and get results as a pandas dataframe:

ds = aconnect.Dataset(
    'USEquityMarketData', 'TradeOnlyMinuteBar', session=session
)

ds.select(
    ds.BarDateTime,
    ds.Ticker,
    ds.Volume
).filter(
    ds.Ticker,isin(['AAPL', 'FB']) &
    ds.TradeDate > '2022-05-01'
).fetch()

TODO

• pandas dataframe parse date/time columns