Visualizing Regression Data

In regression, you need training and testing data, just like in classification. This has already been set up in the starter code. Change the value of test_color from "b" to “r” (for “red”), and rerun . Note: For those students converting Python 2 code to Python 3, see below for some important remarks regarding compatibility.

You will be fitting your regression using the blue (training) points only. (You may have noticed that instead of the standard 10%, we’ve put 50% of our data into the test set--that’s because in Part 5, we will switch the training and testing datasets around and splitting the data evenly makes this more straightforward.)

INSTRUCTOR NOTE:

From Python 3.3 forward, a change to the order in which dictionary keys are processed was made such that the orders are randomized each time the code is run. This will cause some compatibility problems with the graders and project code, which were run under Python 2.7. To correct for this, add the following argument to the featureFormat call on line 26 of finance_regression.py :

sort_keys = '../tools/python2_lesson06_keys.pkl'

This will open up a file in the tools folder with the Python 2 key order.

Next Concept