Hands-On Data Analysis with Pandas

Predicting the length of a year on a planet

Before we can build our model, we must isolate the columns that are used to predict (semimajoraxis, mass, and eccentricity) from the column that will be predicted (period):

>>> data = planets[...     ['semimajoraxis', 'period', 'mass', 'eccentricity']... ].dropna()>>> X = data[['semimajoraxis', 'mass', 'eccentricity']]>>> y = data.period

This is a supervised task. We want to be able to predict the length of a year on a planet using its semi-major axis, mass, and eccentricity of orbit, and we have the period lengths for most of the planets in the data. Let's create a 75/25 split of training to testing data so that we can assess how well this model predicts year length:

>>> from sklearn.model_selection ...