Time series cross-validation

scikit-learn can perform cross-validation for time series data such as stock market data. We will do so with a time series split, as we would like the model to predict the future, not have an information data leak from the future.

Getting ready

We will create the indices for a time series split. Start by creating a small toy dataset:

from sklearn.model_selection import TimeSeriesSplit
import numpy as np
X = np.array([[1, 2], [3, 4], [1, 2], [3, 4],[1, 2], [3, 4], [1, 2], [3, 4]])
y = np.array([1, 2, 3, 4, 1, 2, 3, 4])

How to do it...

1. Now create a time series split object:

tscv = TimeSeriesSplit(n_splits=7)

2. Iterate through it:

for train_index, test_index in tscv.split(X):

      X_train, X_test = X[train_index], X[test_index]
      y_train, y_test = y[train_index], y[test_index]

      print "Training indices:", train_index, "Testing indices:", test_index

Training indices: [0] Testing indices: [1]
Training indices: [0 1] Testing indices: [2]
Training indices: [0 1 2...