Train-test split for time series in ML Python - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to import the function used for splitting time series data.

ML Python

from sklearn.model_selection import [1]

Drag options to blanks, or click blank then click option'

Atrain_test_split

BKFold

CTimeSeriesSplit

Dcross_val_score

Attempts:

3 left

💡 Hint

Common Mistakes

Using train_test_split which shuffles data and breaks time order.

The TimeSeriesSplit function is designed specifically for splitting time series data while preserving order.

2fill in blank

medium

Complete the code to create a time series splitter with 3 splits.

ML Python

tscv = [1](n_splits=3)

Drag options to blanks, or click blank then click option'

ATimeSeriesSplit

BKFold

Ctrain_test_split

DStratifiedKFold

Attempts:

3 left

💡 Hint

Common Mistakes

Using KFold which shuffles data and breaks time order.

TimeSeriesSplit is used to create splits that respect time order. Here, we specify 3 splits.

3fill in blank

hard

Fix the error in the code to correctly split features and target for time series cross-validation.

ML Python

for train_index, test_index in tscv.[1](X):
    X_train, X_test = X.iloc[train_index], X.iloc[test_index]
    y_train, y_test = y.iloc[train_index], y.iloc[test_index]

Drag options to blanks, or click blank then click option'

Atrain_test_split

Bsplit

Csplit_data

Dcross_val_score

Attempts:

3 left

💡 Hint

Common Mistakes

Using train_test_split as a method of TimeSeriesSplit causes an error.

The split method of TimeSeriesSplit returns indices for train and test sets preserving time order.

4fill in blank

hard

Fill both blanks to create a train-test split manually for time series data without shuffling.

ML Python

train_size = int(len(data) * [1])
train, test = data[:train_size], data[[2]:]

Drag options to blanks, or click blank then click option'

A0.8

B0.2

Ctrain_size

Dlen(data) - train_size

Attempts:

3 left

💡 Hint

Common Mistakes

Using 0.2 for train size or slicing test incorrectly causing overlap.

We use 80% of data for training (0.8) and slice the test set starting from train_size index to keep time order.

5fill in blank

hard

Fill both blanks to create a dictionary of train and test indices using TimeSeriesSplit.

ML Python

splits = {}
for i, ([1], [2]) in enumerate(tscv.split(data)):
    splits[i] = {'train': [1], 'test': [2]

Drag options to blanks, or click blank then click option'

Atrain_index

Btest_index

Ctrain_idx

Dtest_idx

Attempts:

3 left

💡 Hint

Common Mistakes

Mixing variable names or assigning test indices to 'train' key.

We unpack the split into train_index and test_index, then store them in the dictionary with keys 'train' and 'test'.