Project 1

In this project, you will:

• Work with a new dataset

• Practice working with data in:

  • pandas

  • The Python standard library

Steps¶

1. Read the general Project information.

2. Find a dataset.

   • It must have:

     • At least one numeric column

     • Between one thousand and one million rows

       • If it’s larger than that, you can filter it down.

       • The alumni employment data is smaller than that; that’s ok.

   • Don’t spend too long on this step.

3. If there’s more than one numeric column, pick one.

4. Create a new notebook.

5. Using pandas:

   1. Read in the data.

   2. Compute the following (or -5 points each):

      • The mean

      • The median

      • The mode

6. Repeat the previous step (or -10 points each) using only the Python standard library (or -10 points), a.k.a. the hard way.

7. Create a data visualization, following the instructions below.

8. Submit.

The hard way¶

• You may not use pandas, the statistics module, a spreadsheet program, etc.

  • You should be using the same dataset from the first step, but not accessing the DataFrame/Series.

    • In other words, if put the code for this step in a totally separate notebook, it should still work.

• You should be calculating the mean, median, and mode yourself, not using functions with those names (or equivalent).

  • Hint: Use a dictionary to keep track of value counts.

Data visualization¶

Requirements:

• The data/calculations can come through pandas, but the drawing code should only use the Python standard library.

  • In other words, don’t use plot(), Plotly, or any other external packages.

• The visualization should be visual, using shape, size, symbols, etc. to represent the values.

  • Printing the numbers (as is) isn’t sufficient.

• General requirements

We’ll talk about data visualization in more detail in week 10, but none of that knowledge is expected to complete this.

Example¶

Data that looks like this:

Rat sightings

could be turned into a sparkline that looks like this:

Rat sightings, in thousands

2014: ***
2015: *****
2016: ****

Please don’t print thousands of asterisks (*) 😉

Tips¶

• Start simple.

  • Start with the example above, get that working, then go from there.

  • Use only one or two columns of your dataset.

  • print()ing strings will probably be easiest, but you can get fancy and generate HTML if you want.

• Making your chart vertical (one data point per line) will probably be easier than doing something horizontal.

• Techniques that may be helpful:

  • Repeating strings

  • “Rewinding” a CSV

• Python strings can contain Unicode, including emoji 📈✨