Skip to content

README.md

Latest commit

 

History

History
104 lines (74 loc) · 3.02 KB

README.md

File metadata and controls

104 lines (74 loc) · 3.02 KB

Frequenz Resampling

Build Status PyPI Package Docs

Introduction

frequenz-resampling provides fast, typed resampling for timestamped numeric data in Python.

The library is backed by the Rust resampler frequenz-floss/frequenz-resampling-rs and exposed through a small Python API. It supports incremental stream resampling through a Resampler object as well as one-shot resampling with the resample() convenience function.

Main capabilities:

  • fixed-interval resampling with timezone-aware datetime timestamps
  • configurable interval semantics via Closed and Label
  • multiple aggregation functions via ResamplingFunction: AVERAGE, SUM, MAX, MIN, FIRST, LAST, COUNT, COALESCE
  • typed Python API (py.typed) and generated API docs

Supported Platforms

The following platforms are officially supported (tested):

  • Python: 3.11, 3.12, 3.13, 3.14
  • Operating System: Ubuntu Linux 24.04, Windows, macOS
  • Architectures: amd64, arm64

Quick Start

Install

python -m pip install frequenz-resampling

Stream resampling with Resampler

import datetime as dt

from frequenz.resampling import Closed, Label, Resampler, ResamplingFunction

start = dt.datetime(1970, 1, 1, tzinfo=dt.timezone.utc)
step = dt.timedelta(seconds=1)

resampler = Resampler(
    dt.timedelta(seconds=5),
    ResamplingFunction.AVERAGE,
    max_age_in_intervals=1,
    start=start,
    closed=Closed.LEFT,
    label=Label.RIGHT,
)

for i in range(10):
    resampler.push_sample(timestamp=start + i * step, value=float(i + 1))

result = resampler.resample(start + 10 * step)
print(result)
# [(1970-01-01 00:00:05+00:00, 3.0), (1970-01-01 00:00:10+00:00, 8.0)]

One-shot resampling with resample()

import datetime as dt

from frequenz.resampling import Closed, Label, ResamplingFunction, resample

start = dt.datetime(1970, 1, 1, tzinfo=dt.timezone.utc)
step = dt.timedelta(seconds=1)
data = [(start + i * step, float(i + 1)) for i in range(10)]

result = resample(
    data,
    dt.timedelta(seconds=5),
    ResamplingFunction.SUM,
    closed=Closed.LEFT,
    label=Label.RIGHT,
)
print(result)

Documentation

Contributing

Contributions are welcome. For development setup, testing, docs, and release workflow, see the Contributing Guide.

License

This project is licensed under the MIT License.