Contents

1. Presidential ratings
2. Employee database
3. On your own

Presidential ratings

Skills used: CSV reading, basic math, dealing with data

Your task is to write a program that summarizes the data contained in presidents.csv. This contains data from the presidents dataset in R, quarterly approval ratings of US presidents. Your program avg_ratings.py should take two filenames as command-line arguments: the CSV file to read (presidents.csv) and the output filename.

The input file contains three columns: id is a unique identifier for each row, year gives the year and quarter of the rating, and rating gives the approval rating.

You should write your output to a new CSV with three columns: year will contain whole years (i.e., 1975 but no information about specific quarters), mean will have the mean approval rating for that year, and median will have the median approval rating for that year.

As some quarters are missing ratings (NA) you will have to exclude those from the mean/median calculations. You should write functions to calculate the mean and median of a list of numbers and remove any NA from the list before passing it to those functions.

Your output should look like the following for the first five years:

year,mean,median
1945,81.33333333333333,82
1946,47.0,46.5
1947,51.0,54.5
1948,37.5,37.5
1949,58.5,57.0

An example solution is available in avg_ratings.py.

Employee database

Skills used: Listing directory contents, classes, string parsing

Your task is to write a program that aggregates data spread across a number of files. Start by downloading the file employee_info.zip and unzipping it to create a folder named employee_info.

In this folder, there’s a set of files of the format <employee id>.txt with information about each employee. For example, 372.txt contains the following:

Name: John Smith
Title: Researcher

You should write a program that reads in every file in the employee_info folder and stores the data in a dictionary. Each key for the dictionary should be the employee ID (which is numeric, but you can treat it as a string without any problems), and each value should be a object of the class Employee.

You should define the class Employee yourself such that:

1. It has an __init__ function that takes the employee name, id, and title as arguments and stores them in the class.
2. It has a __str__ function that allows for nice, human-readable printing of the object. For example, in might return: John Smith, Researcher (372).

Your program should:

1. Take the target directory (employee_info) as a command-line argument.
2. List the contents of the directory.
3. Open each file in the directory and store the information in that file in a new Employee object in the dictionary.
4. At the end, print out all of the key-value pairs of the dictionary, one per line, separated by a colon. For example, a line would be: 372: John Smith, Researcher (372).

Hints:

1. You will need to put together the name of the directory of files and the individual filenames in order to open them. Use os.path.join.
2. You can assume the first line of each file contains the name field and that the second line contains the title field. (If you want an extra challenge, write a solution that does not make this assumption.)
3. You can use split to parse each string in the file. If you want an additional challenge, use a regular expression, using the module re.
4. You should not assume that a name only consists of a first and last name; your solution should still work if the name is “John Q Public.” If you notice that you are only extracting the first name, you are probably using split incorrectly.

An example solution is available in employee_info.py.

On your own

Now’s a great time to come up with your own exercise and try it out!