pandas has direct support for various web-based data source classes in the pandas.io.data namespace. The primary class of interest is pandas.io.data.DataReader, which is implemented to read data from various supported sources and return it to the application directly as DataFrame.
Currently, support exists for the following sources via the DataReader class:
Daily historical prices' stock from either Yahoo! and Google Finance
Yahoo! Options
The Federal Reserve Economic Data library
Kenneth French's Data Library
The World Bank
The specific source of data is specified via the DataReader object's data_source parameter. The specific items to be retrieved are specified using the name parameter. If the data source supports selecting data between a range of dates, these dates can be specified with the start and end parameters. We will now take a look at reading data from each of these sources.