PyDSWS

Python wrapper for the Datastream Web Services API (DSWS)

Connect to the Thomson Reuters Datastream database via Datastream Web Services. You need to have a Datastream subscription and a username/password to use this package. Please note that this is an unofficial/unsupported package and it is still under development. The package has basic functionality and most of the error handling still needs to be done.

Installation

PyDSWS is written in Python 3. Dependencies:

pandas
requests

Installation:

pip install PyDSWS

Usage

Quick Start:

An example how to retrieve closing prices for Vodafone.

import the 'PyDSWS' package
authenticate with your username and password
use the 'get_data' function

import PyDSWS

ds = PyDSWS.Datastream(username='XXXXXXX', password='XXXXXXX')

df = ds.get_data('VOD')

print(df)

Please find below examples on how to set parameters for the requests.

For static data:

ds.get_data(tickers='VOD', fields='P', date='2017-01-01')

For time series:

ds.get_data(tickers='VOD', fields='P,MV,VO', start='-10D', end='-0D', freq='D')

*for multiple tickers/fields, just use a comma. In this example: fields='P,MV,VO'

The output is a Pandas DataFrame:

Instrument     VOD
Field            P        MV       VO
Date
2017-11-21  229.75  61283.06  55100.4
2017-11-22  228.75  61016.34  79602.5
2017-11-23  225.40  60122.75  35724.1
2017-11-24  225.50  60149.44  42918.0
2017-11-27  224.60  59909.38  50355.3
2017-11-28  226.45  60402.83  49027.0
2017-11-29  225.25  60082.74  61618.1
2017-11-30  224.30  59829.99  95423.4
2017-12-01  224.00  59749.96  54855.4
2017-12-04  225.55  60163.41  47524.1
2017-12-05  225.25  60083.38  43846.7

Please check http://datastream.thomsonreuters.com/DswsClient/Docs/Default.aspx for further documentation.

Resources

You can use the Datastream Navigator to look up codes and data types: http://product.datastream.com/navigator/

Development

If you discover any issues with regards to this project, please feel free to create an Issue. If you have coding suggestions that you would like to provide for review, please create a Pull Request.

Acknowledgements