1. Learn
    2. /
  2. Courses
    3. /
  3. Introduction to Regression with statsmodels in Python
Connected

Exercise

The limits of prediction

In the last exercise, you made predictions on some sensible, could-happen-in-real-life, situations. That is, the cases when the number of nearby convenience stores were between zero and ten. To test the limits of the model's ability to predict, try some impossible situations.

Use the console to try predicting house prices from mdl_price_vs_conv when there are -1 convenience stores. Do the same for 2.5 convenience stores. What happens in each case?

mdl_price_vs_conv is available.

Instructions 1/2

undefined XP
    1
    2

Create some impossible explanatory data. Define a DataFrame impossible with one column, n_convenience, set to -1 in the first row, and 2.5 in the second row.