Accounting for non-stationarity
In this exercise, you will again visualize the variations in model scores, but now for data that changes its statistics over time.
An instance of the Linear regression model object is stored in model, a cross-validation object in cv, and the data in X and y.
This exercise is part of the course
Machine Learning for Time Series Data in Python
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Pre-initialize window sizes
window_sizes = [25, 50, 75, 100]
# Create an empty DataFrame to collect the stores
all_scores = ____(index=times_scores)
# Generate scores for each split to see how the model performs over time
for window in window_sizes:
# Create cross-validation object using a limited lookback window
cv = ____(n_splits=100, max_train_size=window)
# Calculate scores across all CV splits and collect them in a DataFrame
this_scores = ____(____, ____, ____, cv=cv, scoring=my_pearsonr)
all_scores['Length {}'.format(window)] = this_scores