Assignment 5

1a. Data Utilities (15 pts)

Create a util.py module that has three methods: download_data, get_data, and parse_ingredients.

The download_data method should download the food-data-sample.json datafile and store it locally. The get_data method should load the data in a module variable. Assume that the data file resides in the same directory as util.py. You can then get its absolute path via the __file__ variable of the module via:

import os
fname = os.path.join(os.path.dirname(__file__),'food-data-sample.json')

Use the json module to load the data from the file. The download_data method should download the file just once, otherwise returning the local filename. Refer to the Assignment 3 starter notebook for code that can be used to download data from the file. The get_data method should load and parse the file from disk once, otherwise returning the pre-loaded data. The parse_ingredients method should parse the ingredients string into a list of ingredients by splitting on a comma. Do not worry if this does not do a perfect job (e.g. you may see an ingredient labeled as “LESS THAN 2% OF: SALT”).

1b. Finding Items (15 pts)

Create a find.py module that has two functions that take one parameter each, brand owner and description expression, respectively, and return the list of data items that match. Use the get_data method from the data module to obtain the data. The first method, get_by_brand, should return a list of the items filtered by an exact match to brand owner, ignoring case. The second method, get_by_description, should take a string that searches all item descriptions for those that match the given expression. The expressions may use literal characters but also the wildcard character (*) to match any number of characters. Remember that when you use Python’s regular expression library to deal with wildcards, you need to change * to .*. Your regular expression should ignore case. For example, consider the output when searching for "Quaker*Oat":

>>> [d['description'] for d in food_data.find.get_by_description("Quaker*Oat")]
['QUAKER, OAT BRAN, HOT CEREAL',
 'QUAKER, OLD FASHIONED OATS',
 'QUAKER, INSTANT OATMEAL, BANANA NUT',
 'Quaker Oat Bran Hot Cereal 16 Ounce Box',
 'Quaker Corn Crunch Toasted Corn & Oat Cereal 15 Ounce Paper Box']

1c. Comparison (15 pts)

Create a compare.py module that calculates comparative information between two food items. Given two food items’ fdc_ids as parameters, the diff_nutrition function should return the difference between the nutrition values of the two items, and the diff_ingredients method should return the difference between the ingredients in the items. Also, diff_nutrition should additionally include the difference in serving size; it should return a list of tuples of the form (<name>, <amount>, <unit_name>). Make sure the differences are between the same measures! You should parse the ingredient list via the parse_ingredients method in the util module from Part 1a, and then create three different sets: one for shared ingredients, one for ingredients in the first item and not the second, and one for ingredients in the second item and not the first.

Examples:

>>> food_data.compare.diff_nutrition(345534, 604974)
[('Serving Size', 0.0, 'g'),
 ('Total Fat', -0.41, 'G'),
 ('Fiber', 0.0, 'G'),
 ('Calories', -12.0, 'KCAL'),
 ('Protein', 0.0, 'G'),
 ('Sugar', -2.48, 'G'),
 ('Saturated Fat', 0.0, 'G'),
 ('Sodium', 137.0, 'MG'),
 ('Carbohydrates', -0.8300000000000001, 'G')]

>>> food_data.compare.diff_ingredients(345534, 604974) 
({'Calcium Chloride', 'Citric Acid', 'Tomato Juice', 'Tomatoes'},
 {'Less Than 2% of: Salt'},
 {'Dried Garlic',
  'Dried Onion',
  'Natural Flavor',
  'Olive Oil',
  'Salt',
  'Soybean Oil',
  'Spices'})

1d. Package

Make sure all three analysis modules live in a single food_data package. Add an __init__.py file for completeness. It may contain documentation and the pass keyword.

2a. food_data.find (15 pts)

In this module, we will create a command-line program that accepts two flags, -b for get_by_brand searches, and -d for get_by_description searches. Both will take a single string as the final argument. That string should be passed to the respective functions, but the output should be displayed in a more succinct manner. Each data item should be listed as <fdc_id> <brand_owner> <description>. Make sure to have a usage statement that is shown when a user enters incorrect input. You should test your script via the IPython magic command %run -m food_data.find .... Some sample output:

>>> %run -m food_data.find

Usage: python -m food_data.find -b <brand> | -d <description>

>>> %run -m food_data.find -d "Quaker*Oat"

538773 The Quaker Oats Company QUAKER, OAT BRAN, HOT CEREAL
539625 The Quaker Oats Company QUAKER, OLD FASHIONED OATS
540325 The Quaker Oats Company QUAKER, INSTANT OATMEAL, BANANA NUT
767774 QTG Quaker Oat Bran Hot Cereal 16 Ounce Box
768340 QTG Quaker Corn Crunch Toasted Corn & Oat Cereal 15 Ounce Paper Box

2b. food_data.compare (15 pts)

In this module, we will create a command-line program that accepts two flags, -n for diff_nutrition and -i for diff_ingredients. Both will take two fdc_ids as integers and pass them to their respective functions. The output for diff_nutrition should show the name of the metric, and then a +/- value for the amount (with two decimal places) followed by the unit. The output for diff_ingredients should show the common ingredients prefixed by a space, the ingredients in the first item prefixed by a - and the ingredients in the second item prefixed by a +. You can test your script via the IPython magic command %run -m food_data.compare ...

Some sample output:

>>> %run -m food_data.compare -n 345534 604974

Serving Size: +0.00 g
Total Fat: -0.41 g
Fiber: +0.00 g
Calories: -12.00 kcal
Protein: +0.00 g
Sugar: -2.48 g
Saturated Fat: +0.00 g
Sodium: +137.00 mg
Carbohydrates: -0.83 g

>>> %run -m food_data.compare -i 345534 604974

  Tomatoes
  Tomato Juice
  Citric Acid
  Calcium Chloride
- Less Than 2% of: Salt
+ Natural Flavor
+ Dried Onion
+ Dried Garlic
+ Salt
+ Olive Oil
+ Spices
+ Soybean Oil

Hints

• You’ll need to check particular indices of sys.argv to determine which flag is being used.