haver 0.1.0

A Python wrapper to Haver Analytics' RESTful API.

haver

Installation

pip install haver

How to

Connection and Authentication

from haver import Haver

haver = Haver(private_token='<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(private_token='<your-haver-API-token>',
              verify=False, # Or local path to certificates 
              proxies={'http': 'http://proxy-username:[e-mail]:8080',
                       'https': 'http://proxy-username:[e-mail]:8080'})

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

Obtaining tokens

To obtain your token, follow these steps:

• Log into haverview.com, then create a graph with any series.
• Click the vertical dots next to the Directory and Series tabs, above the search box.
• Select Export & Sharing in that menu. This will replace the left tab with the Export dialog.
• Copy the contents of the URL field. It is a link which contains your token, specified as a query parameter

E.g. from

https://api.haverview.com/some/path/here?token=3f15493a-9e05-4b61-93ff-8ba56cb3a726

your token would be 3f15493a-9e05-4b61-93ff-8ba56cb3a726.

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)

In addition, a search function is also available to allow the user to search series by their descriptions, for example as:

haver.search(query='employment')

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}.

Finally, a database of available recessions can be obtained as

haver.recessions()

Author

Luca Mingarelli, 2024