Scaling parts of a dataset
In previous videos, you've used the QuantileTransformer
on the full dataset. In this exercise, you will practice scaling only parts of a dataset. The reason for doing this is that the stocks datasets have numerically-encoded categorical features (day_of_week
, day
, month
) that would have been incorrectly scaled if you used QuantileTransformer
on the full dataset.
The transformer has been imported from sklearn
along with the apple
stocks dataset with the extra features.
This exercise is part of the course
Anomaly Detection in Python
Exercise instructions
- Create a list that contains the five numeric column names of
apple
. - Initialize a
QuantileTransformer
that casts features to a normal distribution. - Scale and store the five columns in
to_scale
simultaneously.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Create a list of columns
to_scale = [____]
# Initialize a QuantileTransformer
qt = ____
# Scale and store simultaneously
apple.loc[____] = ____