Adventures in Ruby - Arrays of Hashes

Coming from a data background, I am very used to viewing and handling data in highly structured tables - and languages such as Python and R have dataframes which simulate this structure by way of dataframes which allow for very quick navigation and data manipulation. When looking at other popular languages things get a bit tricky though as tabular- style data structures aren’t as common. Even less so in Ruby - so what have we got?

Well there’s your classic hash which gives you the ability to store a value with a key, like survey results:

favourite_star_wars = {'A New Hope': 16, 'The Empire Strikes Back': 18,
                     'Return of the Jedi':12 }

And we can return the result of a particular film like so:

ep_5_result = favourite_star_wars[:"The Empire Strikes Back"]
=> 18

However, what if we wanted to calculate something more aggregate? Like what is the average score across all films? Or we wanted to categorise the data by another dimension (like year of release?). It becomes unwieldy with just a hash.

By organising the data into an array of hashes (each hash representing a film) we have consistent keys across all our films ready for comparison.

favourite_star_wars =[ {'Title': 'A New Hope', 'Release Date': 1977, 'Score': 16 },
                       {'Title': 'The Empire Strikes Back', 'Release Date': 1980, 'Score': 18 },
                       {'Title': 'Return of the Jedi', 'Release Date': 1983, 'Score': 12 }
                     ]

As our film titles are now values rather than keys it means we have to be a bit more verbose if we want our score back for a film - using the array method select to pick out the entire film listing and then finding the score:

ep_5 = favourite_star_wars.select{|film| film[:Title] == 'The Empire Strikes Back'}[0]
=> {'Title': 'The Empire Strikes Back', 'Release Date': 1980, 'Score': 18 }
ep_5_result = ep_5[:Score]
=> 18

However it now means it becomes more manageable to do data manipulation as we can explicitly refer to what data we are selecting for each film and what we are doing with it:

scores = favourite_star_wars.map { |film| film[:Score] }
=> [16, 18, 12]
average_result = scores.reduce(:+) / scores.size
=> 15

Hopefully by bundling up your data this way, you’ll have a much easier time playing with the data.