bloomberg_py_wrapper

Simplified Python interface to Bloomberg's data.

Prerequisite

Bloomberg API Sources you can also install it via pip:

python -m pip install --index-url=https://bloomberg.bintray.com/pip/simple blpapi

Some of the functions require a subscription to AIM services, they are always preceded by "aim" in their name.

Structure

There are three main classes for data requests:

• Synchronous
• Asynchronous
• Subscription (Not Implemented)

The classes Synchronous and Asynchronous have most of their data request types in common.

=> Synchrnous Functions:

• field_list_request: returns a list of fields matching the field_type (ALL, STATIC, REAL_TIME)
• field_info_request: returns a description for the searched fields
• field_search_request: enables the use of filtered search using keywords
• reference_data: returns reference data or bulk reference data (equivalent of BDP)
• historical_data: returns historical data, the dates can be alligned (equivalent of BDH)
• intraday_tick_data: returns intraday tick data
• aim_position_data: returns the position data of the selected portfolio
• aim_historical_data: returns the historical position of the data of the selected portfolio
• aim_cash_service: returns the cash position of the selected portfolio

=> Asynchronous Functions (all but the field requests):

• reference_data: returns reference data or bulk reference data (equivalent of BDP)
• historical_data: returns historical data, the dates can be alligned
• intraday_tick_data: returns intraday tick data
• aim_position_data: returns the position data of the selected portfolio
• aim_historical_data: returns the historical position of the data of the selected portfolio
• aim_cash_service: returns the cash position of the selected portfolio

=> Subscription (Not Implemented):

• market_data: can be called-back

Notes on Asynchronous

The Asynchronous Class inherits its functions from the Synchronous class with the added benefit that multiple requests can run in the background thus increasing overall speed. On a request, instead of returning the data it returns a code that can be used for getting the state of the request (request_is_ready, and request_is_error) as well as recovering the data if it is available (get_request_data). An alternative to using the code, is by providing a call_back_function wich must take as inputs:

• status: will be either 'READY' or 'ERROR'
• cid: the code that was returned to you when performing the request
• data: your data in a pandas dataframe

Notes on Subscripion

Implementation Planed