CS 111: Introduction to Computer Science

Processing weather data

Hand in via Moodle as weather.py.

You may work alone or with a partner of your choosing on this assignment. Put your name and the name of you partner (if any) in the comment at the top of your source code. If you work with a partner, submit your code only once (i.e. to one partner's Moodle account, not to both).

Data and starter code

• weather.csv, the raw data file you'll need
• weather.py, some code to get you started

Goals

• Build a more complex program than you have built before.
• Practice designing and implementing functions suitable for completing subtasks of the program as a whole.
• Think about lists and lists of lists.
• Experience one of the things computers are important for and good at: processing large amounts of data.

Summary

The website of the Minnesota Department of Natural Resources makes available daily weather data for the Minneapolis/St. Paul area (and other locations) collected since 1871. For this assignment, you're going to write functions to compute various statistics from this raw data, and a program to call the functions and generate a report.

The program: basic version

Your program, called weather.py, should produce a report looking something like the following. (The numbers I've included in this example are fictional—the correct answers will be something different.)

The program: the fancy version

Each of the statistics in the basic version of this assignment can be computed by performing a simple loop and storing only a small amount of intermediate data. For example, to compute the average high temperature over all the days in the dataset, you only need to keep track of the sum of the high temperatures and the number of days that have high temperatures listed.

Some statistics, however, require a little more sophisticated code. For example, to compute each year's total snowfall, you need to somehow keep a separate running total of snowfall for each year.

With that in mind, the basic version of the program described above will be worth 16 of 20 points, while the fancy version will add another 4 possible points to your total. That is, you have the potential to get a B (which is a good grade) by completing just the features described above. To get those extra 4 points, however, your program should produce a report like this:

(Again, the numbers in this example are fictional.)

Notes and suggestions

• Think about the "incremental development" approach I have described more than once in class. Before you start coding, try to write down a list of stages your work will proceed in.
  Stage 0: run Jeff's starter code see what it does. Stage 1: fix Jeff's code to report the correct year Stage 2: add the "lowest temperature" feature Stage 3: ...
  Planning the stages of your work in this way can make things go much more smoothly.
• Most of your code should be in functions. The exact design of the functions is up to you, but I recommend keeping it simple and thinking in terms of one function per statistic you're trying to compute (at least for the basic version of the program). I have written most of one such function in weather.py so you can see roughly what I'm thinking of.
• Some entries in weather.csv contain just "M", which means "missing". There are many M's at the beginning of the file, because much of the data for years like 1871 is not available, either because it was never collected or because it was not saved. For rows containing M for a relevant statistic, you should ignore that entry. For example, if you're computing the average high temperature over the entire dataset, you should just skip over any row whose high temperature is listed as M—neither count that day as a day, nor add any value to your running total of high temperatures.
• Similarly, the snowfall and snow depth columns sometimes have "T" listed, which stands for "trace". When you encounter a T when doing a snowfall computation, count it as 0, since "trace" means, roughly, "a little bit of snow, but close enough to zero that you might as well call it zero".

Start early, ask questions, and have fun!