tsflex 0.1.1.1

Toolkit for flexible operations on time-series data

tsflex stands for: flexible time-series operations

It is a time-series first toolkit for processing & feature extraction, making few assumptions about input data.

• example notebooks

Table of contents

• Installation
• Usage
  • Series processing
  • Feature extraction
• Documentation

Installation

If you are using pip, just execute the following command:

pip install tsflex

Usage

tsflex is built to be intuitive, so we encourage you to copy-paste this code and toy with some parameters!

Series processing

:WIP:

Feature extraction

import pandas as pd; import scipy.stats as ss; import numpy as np
from tsflex.features import FeatureDescriptor, FeatureCollection, NumpyFuncWrapper

# 1. -------- Get your time-indexed data --------
series_size = 10_000
series_name="lux"

data = pd.Series(
    data=np.random.random(series_size), 
    index=pd.date_range("2021-07-01", freq="1h", periods=series_size)
).rename(series_name)
# -- 1.1 drop some data, as we don't make frequency assumptions
data = data.drop(np.random.choice(data.index, 200, replace=False))


# 2 -------- Construct your feature collection --------
fc = FeatureCollection(
    feature_descriptors=[
        FeatureDescriptor(
            function=NumpyFuncWrapper(func=ss.skew, output_names="skew"),
            series_name=series_name, 
            window="1day", stride="6hours"
        )
    ]
)
# -- 2.1. Add multiple features to your feature collection
fc.add(FeatureDescriptor(np.min, series_name, '2days', '1day'))


# 3 -------- Calculate features --------
fc.calculate(data=data)

Documentation

To see the documentation locally, install pdoc and execute the succeeding command from this folder location.

pdoc3 --template-dir docs/pdoc_template/ --http :8181 tsflex

---

👤 Jonas Van Der Donckt, Jeroen Van Der Donckt, Emiel Deprost