haver-api

The haver-api library provides a convenient wrapper to Haver Analytics' RESTful API.

This library is not an official product of Haver Analytics and is not covered by its customer support. If you encounter any issue or have suggestions for improvements, please feel free to open an issue or submit a pull request.

Installation

pip install haver-api

❗❗❗ Notice ❗❗❗

The ownership of the old haver domain on PyPi (installable as pip install haver) has been passed on to Haver Analytics.

In order to install this library you now need to use the command above.

Running pip install haver will soon install Haver Analytics' library

How to

Connection and Authentication

from haver import Haver

haver = Haver(api_key='<your-haver-API-token>')

The class Haver also accepts keyword arguments to be passed to requests, which handles the connection to the API under the hood. In this way, by passing e.g. verify and proxy parameters, users can access Haver databases from behind firewalls. For example:

haver = Haver(api_key='<your-haver-API-key>',
              verify=False, # Or local path to certificates 
              proxies={'http': 'http://proxy-username:proxy-password@proxy-server.com:8080',
                       'https': 'http://proxy-username:proxy-password@proxy-server.com:8080'})

Instead of passing the API key explicitely each time, the user can also set an environmental variable HAVER_API_KEY containing the API key. In this case connection will be as simple as haver = Haver().

Obtaining your API Key

You can find your API key at this link.

Exploring available resources

All available databases can easily be listed as

haver.get_databases()

which will return a dictionary with keys the database names and values the corresponding database description:

{'UNPOP': 'U.N. Population Statistics',
 'EPFRECA': 'Fund Country Allocations',
 'EUFIN': 'Financial Data',
 ...
 }

Further information on each dataset can be obtained via the method haver.database_info, and series within each database can be listed e.g. as

haver.get_series(database='UNPOP', full_info=True)

Querying data

In order to retrieve data, the user has the option of querying one series at a time via the dedicated method

haver.read(database='EUDATA', series='N997CE')

which returns data in dictionary format, or querying multiple series as

haver.read_df(haver_codes=['N997CE@EUDATA','N025CE@EUDATA'])

where individual haver_codes are created by joining series and database names as {series}@{database}.

Author

Luca Mingarelli, 2018