tsfeatures

library0.4.5pypypi✓ verified 81d ago

Calculates various features from time series data, including autocorrelation, entropy, seasonality, stationarity, and trend-based metrics. Designed for time series classification and preprocessing. Current stable version is 0.4.5. Release cadence is irregular based on fixes and feature additions.

data ai-ml

pip install tsfeatures

github ↗package ↗

INSTALL

IMPORT

SIG · TSFEATURES

tsfeatures

datapythonv0.4.5

Install

21.3s avg

Import

—

Disk

600MB

Pass rate

5/ 10

Env Coverage5 / 10

glibc

3.9–3.13

musl

3.9–3.13

Install & Compatibility

Where this runs

tested against v0.4.5 · pip install

no network on importno background threads

Install × environment matrix

Each cell = how many times install + import succeeded across repeated harness runs. Partial = flaky.
glibc = Debian/Ubuntu slim · musl = Alpine Linux

musl

py 3.10–3.95 runs

build_error

glibc

py 3.10–3.95 runs

installs and imports cleanly · install 21.3s · import 0.000s · 587MB

600MB installed

● package 600MB

Code

Verified usage

Verified import paths — ran on the pinned version, not inferred.

✓ from tsfeatures import AR

✗ from tsfeatures import Tsfeatures

acf

✓ from tsfeatures import acf

✗ from tsfeatures import Tsfeatures

arch_stat

✓ from tsfeatures import arch_stat

✗ from tsfeatures import Tsfeatures

Compute default set of time series features on a single time series column.

import pandas as pd from tsfeatures import Tsfeatures # Sample time series data: one column with datetime index dates = pd.date_range('2020-01-01', periods=50, freq='D') df = pd.DataFrame({'value': [i + (i%10)*0.5 for i in range(50)]}, index=dates) tf = Tsfeatures() features = tf.calculate_features(df) print(features)

Debug

Known issues

breakingChanged from function-based to class-based API in v0.1.0. Old code using `tsfeatures.calculate_features(df)` no longer works. Use `Tsfeatures().calculate_features(df)` instead.

fix

Upgrade to >=0.1.0 and use the Tsfeatures class.

affects: <0.1.0

gotchaInput DataFrame must have a DatetimeIndex; otherwise features relying on frequency detection (like seasonal features) may fail or produce NaNs.

fix

Ensure index is a pandas DatetimeIndex with a recognizable frequency (e.g., 'D', 'H'). If not needed, use `Tsfeatures(freq=1)` to treat as non-seasonal.

affects: >=0.1.0

gotchaSome features require the `antropy` library; if not installed, those features are silently skipped. No error raised.

fix

Install antropy: `pip install antropy` to get entropy-based features.

affects: >=0.3.0

deprecatedThe `freq` parameter in `calculate_features` is deprecated as of v0.4.0. Use `Tsfeatures(freq=...)` constructor instead.

fix

Pass frequency to Tsfeatures constructor: `tf = Tsfeatures(freq=24)`.

affects: 0.4.0 and later

Upgrade

Version history

0.4.5latest on PyPI · released Jun 20, 2023

Audit

Dependencies

pandasrequiredData manipulation required for input/output.

numpyrequiredNumerical operations used internally.

statsmodelsrequiredUsed for ARIMA coefficients, unit root tests, etc.

antropyoptionalProvides entropy features (sample entropy, etc.).

Agent activity

23 hits · last 30 days

node

OpenAI (training)

Resources

githubhttps://github.com/Nixtla/tsfeatures ↗

packagehttps://pypi.org/project/tsfeatures/ ↗