Osprey's Coding Challenge

coding_challenge.py

"""
Osprey Code Challenge

We have a report of ~1000 images taken by 10 cameras in the past month.
Each image metadata object contains a 'camera_id' and 'size' property.
For example:

    {
        "camera_id": 1,
        "size": 42048
    }

We would like to be able to query this data to answer some simple
questions:
    - Which cameras have used the most data?
    - Which cameras have the highest number of images?
    - What are the largest images per camera?

Please create an ImageManager class to support these needs. The class
should take in a path to a local JSON file when instantiated. The
ImageManager class should implement the following methods:
    - get_cameras_with_most_images(reverse=False): returns an iterable
        of cameras sorted by the number of images taken per camera
        (descending).
    - get_cameras_with_most_data_usage(reverse=False): returns an
        iterable of cameras sorted by the sum of sizes of images
        taken per camera (descending).
    - get_largest_images(reverse=False): returns an iterable of the
        largest images taken by each camera, sorted by size (descending).
        Note that only one image will be returned per camera.

Each method takes an optional reverse keyword argument. If set to True, the
iterable returned will be in reversed order.

Please do your best to consider performance and legibility in your
implementation. Assume that the solution might be used on datasets up to
10 million images. Additionally, assume that this code will be going in
to production.  You may choose to use either Python 2.7.x or Python
3.x in your solution.

Please provide unit tests to confirm your solution works.
"""

import json
from collections import Counter, defaultdict


# Write your implementation here:
class ImageManager(object):

    def __init__(self, path):
        with open(path) as p:
            self.data = json.load(p)['images']

    def get_cameras_with_most_images(self, reverse=False):
        dic = defaultdict(Counter)
        for item in self.data:
            key = item['camera_id']
            dic[key].update('count')

        iterable = [{'camera_id': k, 'count': v.values()[0]} for k, v in dic.iteritems()]
        return sorted(iterable, key=lambda k: k['count'] if reverse else -k['count'])

    def get_cameras_with_most_data_usage(self, reverse=False):
        dic = defaultdict(Counter)
        for item in self.data:
            key = item['camera_id']
            dic[key].update({key: item['size']})

        iterable = [{'camera_id': k, 'size': v.values()[0]} for k, v in dic.iteritems()]
        return sorted(iterable, key=lambda k: k['size'] if reverse else -k['size'])

    def get_largest_images(self, reverse=False):
        dic = {}
        for item in self.data:
            key = item['camera_id']
            dic[key] = item['size'] if dic.get(key, 0) < item['size'] else dic.get(key, 0)

        iterable = [{'camera_id': k, 'size': v} for k, v in dic.iteritems()]
        return sorted(iterable, key=lambda k: k['size'] if reverse else -k['size'])