coreforecast
Fast implementations of common forecasting routines
Decision gist · record as of 2026-08-14
Yes, if you are working with time-series forecasting pipelines and need fast grouped transformations. The library is actively maintained, has no known vulnerabilities, and carries a permissive Apache-2.0 license. However, most users should access it indirectly through mlforecast or similar higher-level libraries rather than using it directly; install it directly only if you need fine-grained control over grouped-array operations or are building a forecasting tool.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires Python 3.10 or later; wheels are pre-compiled for common platforms but may require a C++ build environment on unsupported architectures.
- Medium install friction due to compiled wheels across multiple Python versions and platforms.
- Active maintenance with recent release (36 days ago) and no known vulnerabilities.
License · maintenance · safety
Apache-2.0 (permissive) — Apache-2.0 permissive license allows commercial and private use with minimal restrictions; you must include a copy of the license and note any modifications to the library itself.
last release 2026-07-09 (36 days) · last repo commit 2026-07-27 · 42 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 1,900,154 downloads/mo, #3,444 on PyPI
Alternatives
Verify before relying
pip install coreforecast
import numpy as np
from coreforecast.grouped_array import GroupedArray
from coreforecast.scalers import LocalStandardScaler
data = np.arange(10)
indptr = np.array([0, 3, 10])
ga = GroupedArray(data, indptr)
scaler = LocalStandardScaler().fit(ga)
standardized = scaler.transform(ga)- Whether the library's multithreading performance gains justify the medium install friction for typical use cases.
- How cold-start performance compares to the numba-based approach it was designed to replace.
- Whether higher-level libraries like mlforecast are the recommended entry point rather than direct use.
What it is and what it does
Coreforecast is a C++ acceleration library for time-series data transformations, built by Nixtla to replace slower Python and numba-based implementations. It operates on grouped arrays—a compact representation where multiple independent time series are stored in a single flattened numpy array with index pointers marking series boundaries—and applies operations like scaling, lag transforms, rolling statistics, and seasonal decomposition with multithreading support.
The library is designed as a backend for higher-level forecasting tools rather than for direct use in most cases. It depends only on numpy and avoids the cold-start overhead and LLVM dependency of numba-based approaches. You interact with it through a grouped-array abstraction and call transformations like LocalStandardScaler or ExpandingMean; the library handles the C++ computation and parallelization internally.
Use it for
- Preprocessing large collections of time series for machine-learning forecasting pipelines where scaling and lag features are needed.
- Computing expanding-window statistics (mean, standard deviation) across many independent time series efficiently.
- Applying rolling-window operations like moving averages to grouped time-series data with minimal Python overhead.
- Standardizing or normalizing multiple time series in batch using LocalStandardScaler or similar transformers.
- Serving as the computational backbone for higher-level forecasting libraries that need fast grouped transformations.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you are working with time-series forecasting pipelines and need fast grouped transformations.
The library is actively maintained, has no known vulnerabilities, and carries a permissive Apache-2.0 license. However, most users should access it indirectly through mlforecast or similar higher-level libraries rather than using it directly; install it directly only if you need fine-grained control over grouped-array operations or are building a forecasting tool.
Install
coreforecast on PyPI
Before you install
Medium install friction due to compiled wheels across multiple Python versions and platforms. Active maintenance with recent release (36 days ago) and no known vulnerabilities. Single runtime dependency on numpy keeps the footprint minimal.
Requires Python 3.10 or later; wheels are pre-compiled for common platforms but may require a C++ build environment on unsupported architectures.
License in practice
Apache-2.0 permissive license allows commercial and private use with minimal restrictions; you must include a copy of the license and note any modifications to the library itself.
Quickstart
pip install coreforecast
import numpy as np
from coreforecast.grouped_array import GroupedArray
from coreforecast.scalers import LocalStandardScaler
data = np.arange(10)
indptr = np.array([0, 3, 10])
ga = GroupedArray(data, indptr)
scaler = LocalStandardScaler().fit(ga)
standardized = scaler.transform(ga)Verify before relying
- Whether the library's multithreading performance gains justify the medium install friction for typical use cases.
- How cold-start performance compares to the numba-based approach it was designed to replace.
- Whether higher-level libraries like mlforecast are the recommended entry point rather than direct use.
Package facts
| License | Apache-2.0 permissive |
| Python support | Supports the current Python release >=3.10 |
| Install friction | Medium. Platform-specific wheel |
| Runtime dependencies | 1 packagenumpy |
| Maintenance | Actively maintained 36 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 1,900,154 / month, #3,444 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Development Status :: 3 - AlphaIntended Audience :: Science/ResearchLicense :: OSI Approved :: Apache Software LicenseNatural Language :: EnglishProgramming Language :: Python :: 3.10Programming Language :: Python :: 3.11Programming Language :: Python :: 3.12Programming Language :: Python :: 3.13Programming Language :: Python :: 3.14 |
Evidence: coreforecast-0.0.18-cp310-cp310-macosx_10_9_x86_64.whl; coreforecast-0.0.18-cp310-cp310-macosx_11_0_arm64.whl; coreforecast-0.0.18-cp310-cp310-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl; coreforecast-0.0.18-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl; coreforecast-0.0.18-cp310-cp310-win_amd64.whl; coreforecast-0.0.18-cp311-cp311-macosx_10_9_x86_64.whl; coreforecast-0.0.18-cp311-cp311-macosx_11_0_arm64.whl; coreforecast-0.0.18-cp311-cp311-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl; coreforecast-0.0.18-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl; coreforecast-0.0.18-cp311-cp311-win_amd64.whl; coreforecast-0.0.18-cp312-cp312-macosx_10_13_x86_64.whl; coreforecast-0.0.18-cp312-cp312-macosx_11_0_arm64.whl; coreforecast-0.0.18-cp312-cp312-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl; coreforecast-0.0.18-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl; coreforecast-0.0.18-cp312-cp312-win_amd64.whl; coreforecast-0.0.18-cp313-cp313-macosx_10_13_x86_64.whl; coreforecast-0.0.18-cp313-cp313-macosx_11_0_arm64.whl; coreforecast-0.0.18-cp313-cp313-manylinux_2_27_aarch64.manylinux_2_28_aarch64.whl; coreforecast-0.0.18-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl; coreforecast-0.0.18-cp313-cp313-win_amd64.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “time series transformations”
- coreforecastProvides fast C++ implementations of time-series transformations and…
- sktimesktime provides a unified interface for time series machine learning…
- tslearntslearn provides machine learning algorithms optimized for time…
Give your agent the search over MCP, or paste the wish link into any chat.
More Scientific/Engineering packages
NumPy provides an N-dimensional array object and a comprehensive suite of mathematical, linear algebra, Fourier transform, and random number functions for scientific computing in Python.
pandas provides fast, flexible data structures (Series and DataFrame) for loading, cleaning, transforming, and analyzing labeled or relational data in Python.
scipy provides numerical algorithms for mathematics, science, and engineering—including optimization, integration, linear algebra, Fourier transforms, signal and image processing, and ODE solvers—built on numpy arrays.
scikit-learn provides a comprehensive Python library for supervised and unsupervised machine learning, including classification, regression, clustering, dimensionality reduction, and model evaluation tools built on NumPy and SciPy.
Install it if you need to train, evaluate, or deploy supervised or unsupervised learning models.
dill extends Python's pickle module to serialize and deserialize a much wider range of Python objects, including functions, lambdas, classes, and interpreter sessions, to byte streams for storage or network transmission.
Multiprocess is an enhanced fork of Python's standard multiprocessing library that uses dill for better serialization, allowing you to spawn processes with a threading-like API and share complex objects between them.
Install it if you use multiprocessing and encounter pickle serialization limits with lambdas or complex objects.
See also mlforecast · neuralforecast · statsforecast · window-ops · nixtla · utilsforecast · tbats · sktime · timesfm · dbl-tempo