Feature Engineering First, let's convert our features to numerical values. A straightforward approach is to map some of these non-numerical features into numerical ones. For example, we can treat the day as a value from 1-7. However, one of the disadvantages in directly translating to a numeric value is that we unintentionally assign certain features disproportionate weight. Consider assigning Sunday to the numeric value of 7, and Monday to the numeric value of 1. In our linear model, Sunday will have 7 times the influence of Monday, which can lower the accuracy of our model Instead, let's use one-hot encoding to better represent these features! As discussed in lecture, one-hot encoding will produce a binary vector indicating the non-numeric feature, Sunday would be encoded as a [0000001). This assigns a more even weight across each category in non-numeric features. Complete the code below to one-hot encode our dataset, allowing us to see the transformed dataset named one hot x. This dataframe holds our "featurized" data, which is also often denoted by ϕ.
def one_hot_encode(data): Return the one-hot encoded dataframe of our input data. Parameters datai a dataframe that may include non-numerical features Returns A one-hot encoded dataframe that only contains numeric features one_hot_x - one hot encode(x) one_hot X. head()