Time series cross validation tool, supplementing or replacing relevant existing sklearn libraries.
Project description
#Overview:
The package provides quick train/test split indexing for cross validation, specifically optimized for time series data. There are two primary output options:
- Expanding Window - in which the training window becomes larger with each fold and is always overlapping with part of the previous.
- Rolling Window - in which the training window is of a fixed pre specified dimension, and it may or may not overlap depending on the rolling step specified.
#Example:
''' from time_cross_validation import TimeCV import pandas as pd
#sample X and Y variables: X = pd.DataFrame([10,20,10,4,5,1,7,20]) Y = pd.DataFrame([5,1,7,20,10,20,10,4])
CV = TimeCV(X, train_sample_size = 3, test_sample_size = 3, step = 1) for train_index, test_index in CV.expanding_train_test_split(): x_train = X.iloc[train_index] x_test = X.iloc[test_index] '''
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Hashes for time-cross-validation-0.0.2.tar.gz
Algorithm | Hash digest | |
---|---|---|
SHA256 | f678707f3d9978f92a00eccba5a2348a81f058a43c8d3a2dae542e5d1f694095 |
|
MD5 | 41d9dd05496ea60badcfd98306ff60d2 |
|
BLAKE2b-256 | 3218179af891db7c6a0370804ce796baf134994f0a985c5b956d9df65fb77bb8 |