Train/Test split
To avoid overfitting, it's common practice in Machine Learning to split data into train and test datasets. This is done to ensure that the model is able to correctly predict new, unseen data.
Since we're working with time-series data, we cannot use random split methods, as that would allow the model to know the future.
A function to print the start and end of a DataFrame is available as show_start_end()
, which takes a DataFrame as the only argument, and returns a string.
The data is available as environment
.
This exercise is part of the course
Analyzing IoT Data in Python
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Define the split day
limit_day = ____
# Split the data
train_env = ____[____]
test_env = ____[____]