FredAsDataframe

Up to date remote economic data access for ruby, using Polars dataframes.

This package will fetch economic and financial information from the Federal Reserve Economic Database, and return the results as a Polars Dataframe. You will need an API key that can be fetched from the FRED website at https://fredaccount.stlouisfed.org/apikeys .

Requirements

  • Ruby >= 3.3
  • polars-df ~> 0.27.1

Installation

Add this line to your application's Gemfile:

gem 'fred_as_dataframe'

And then execute:

$ bundle install

Or install it yourself as:

$ gem install fred_as_dataframe

Configuration

Some data sources will require the specification of an API key. These keys should be provided as part of a configuration file, e.g., config/fred_as_dataframe.rb

FredAsDataframe::Client.configure do |config|
  config.api_key = '1234567890ABCDEF'
end

Alternatively, you can read the API key from a file:

FredAsDataframe::Client.configure do |config|
  config.api_key = File.read(File.join('', 'home', 'user', '.fred_api_key.txt')).strip
end

You can also pass the API key directly when creating a client instance:

client = FredAsDataframe::Client.new('UNRATE', api_key: '1234567890ABCDEF')

Usage

Consider the following transcript.

3.3.0 :001 > a = FredAsDataframe::Client.new('UNRATE')
 => #<FredAsDataframe::Client:0x0000000105f1dbb8 @api_key="1234567890ABCDEF", @tag="UNRATE"> 
3.3.0 :002 > b = a.fetch
 => 
shape: (919, 2)                                                    
...                                                                
3.3.0 :003 > b
 => 
shape: (919, 2)                                                    
┌────────────┬────────┐                                            
│ Timestamps ┆ UNRATE │                                            
│ ---        ┆ ---    │                                            
│ date       ┆ f64    │                                            
╞════════════╪════════╡                                            
│ 1948-01-01 ┆ 3.4    │                                            
│ 1948-02-01 ┆ 3.8    │                                            
│ 1948-03-01 ┆ 4.0    │                            
│ 1948-04-01 ┆ 3.9    │                            
│ 1948-05-01 ┆ 3.5    │
│ …          ┆ …      │
│ 2024-03-01 ┆ 3.8    │
│ 2024-04-01 ┆ 3.9    │
│ 2024-05-01 ┆ 4.0    │
│ 2024-06-01 ┆ 4.1    │
│ 2024-07-01 ┆ 4.3    │
└────────────┴────────┘ 
3.3.0 :004 > a = FredAsDataframe::Client.new('AAA')
 => #<FredAsDataframe::Client:0x0000000106077d38 @api_key="1234567890ABCDEF", @tag="AAA"> 
3.3.0 :005 > b = a.fetch
 => 
shape: (1_267, 2)        
...                      
3.3.0 :006 > b
 => 
shape: (1_267, 2)        
┌────────────┬──────┐    
│ Timestamps ┆ AAA  │    
│ ---        ┆ ---  │    
│ date       ┆ f64  │    
╞════════════╪══════╡    
│ 1919-01-01 ┆ 5.35 │    
│ 1919-02-01 ┆ 5.35 │
│ 1919-03-01 ┆ 5.39 │
│ 1919-04-01 ┆ 5.44 │
│ 1919-05-01 ┆ 5.39 │
│ …          ┆ …    │
│ 2024-03-01 ┆ 5.01 │
│ 2024-04-01 ┆ 5.28 │
│ 2024-05-01 ┆ 5.25 │
│ 2024-06-01 ┆ 5.13 │
│ 2024-07-01 ┆ 5.12 │
└────────────┴──────┘

Testing

This gem uses RSpec for testing. To run the test suite:

bundle install
bundle exec rake

Or run RSpec directly:

bundle exec rspec

The test suite uses VCR and WebMock to record and replay HTTP interactions with the FRED API, ensuring tests run quickly and don't require a live API connection or API key.

Running Tests with Coverage

Tests are configured with fixtures that mock FRED API responses, so you don't need a FRED API key to run the test suite.

API Documentation

Client#new(series_id, options = {})

Creates a new FRED data client for the specified series.

  • series_id - FRED series identifier (e.g., 'UNRATE', 'AAA', 'SP500')
  • options - Optional hash with :api_key to override configured API key

Client#fetch(start: nil, fin: nil)

Fetches data from FRED and returns a Polars DataFrame.

  • start - Optional start date (String or Date) to filter data from this date forward
  • fin - Optional end date (String or Date) to filter data up to this date

Returns a Polars DataFrame with timestamp and value columns.

Contributing

Others are welcome to contribute to the project.

The following conventions are intended for this project.

  • Different sources are intended to reside in different classes.
  • API keys (if needed) should be able to be set in the single configuration file.
  • Series should be able to be identified via a single unique string, provided in the constructor.
  • When fetched, the dataset may be filtered based on optional (hash) arguments.
  • Output should be provided in a consistent DataFrame format (currently Polars::DataFrame).

Bug reports and pull requests are welcome on GitHub at https://github.com/bmck/fred_as_dataframe.

License

The gem is available as open source under the terms of the MIT License.