Subcategories
Packages
CuPy provides a NumPy- and SciPy-compatible array library that runs computations on NVIDIA GPUs using CUDA 12.x, letting you write GPU-accelerated numerical code with a familiar NumPy-like API.
kornia-rs provides low-level computer vision operations—image I/O, resizing, color conversion, and video capture—implemented in Rust with Python bindings for efficient, thread-safe processing.
OpenCLIP provides open-source implementations of CLIP (Contrastive Language-Image Pre-training) models that encode images and text into a shared embedding space for zero-shot classification, retrieval, and multimodal tasks.
PyMatting solves the alpha matting problem—extracting foreground objects from images given a hand-drawn trimap—using multiple algorithmic approaches and provides foreground estimation with optional GPU acceleration.
Install it if you need to solve alpha matting problems or extract foreground objects from images with trimap guidance.
Provides IERS Earth Rotation and Leap Second tables as data files for the astropy package, intended for internal use by astropy rather than direct user installation.
MuJoCo is a physics simulation engine with Python bindings that provides direct access to the C API for rigid-body dynamics, collision detection, and rendering.
Encodes latitude/longitude coordinates to geohash strings and decodes them back, with utilities for distance calculations, adjacent geohash lookup, and optional visualization.
VTK is a 3D graphics, image processing, and visualization toolkit that provides algorithms for surface reconstruction, volume rendering, and advanced rendering techniques.
SeleniumBase is a browser automation framework that combines Selenium WebDriver with pytest integration, CDP mode for bot-detection bypass, and tools for web testing, scraping, and crawling.
However, the 60 runtime dependencies are substantial—evaluate whether you need the full feature set or if a lighter alternative suits your use case.
Converts dates between the Hijri (Islamic) and Gregorian calendars using the Umm al-Qura standard, with support for dates from 1343 AH to 1500 AH.
Install it if you need reliable Hijri-Gregorian conversion within the supported date range; the main constraint is the 1343–1500 AH window and the explicit…
spython is a Python API for programmatically managing Singularity containers, letting you build, run, and interact with container images from Python code.
Provides OpenCV's computer vision algorithms and image processing functions in a headless (no GUI) package with contrib modules, designed for server and containerized environments.
HDBSCAN performs hierarchical density-based clustering that automatically finds clusters of varying densities with minimal parameter tuning, and includes outlier detection and branch-detection capabilities.
embreex provides Python bindings to Intel's Embree ray tracing engine, enabling fast ray-geometry intersection queries with only numpy as a runtime dependency.
Hyperopt performs serial and parallel hyperparameter optimization over mixed search spaces using algorithms like Tree of Parzen Estimators (TPE), Random Search, and Adaptive TPE.
Param provides declarative class attributes with runtime validation and a reactive programming API for automatic updates when attribute values change.
Install it if you need runtime parameter validation with class-level configuration, or if you're building interactive systems where automatic synchronization matters.
PyDOE constructs experimental designs for scientists and engineers, providing factorial, mixture, response-surface, space-filling, and optimal design methods to plan efficient experiments.
Install it if you need to design experiments systematically—it covers classical and modern DOE methods and integrates cleanly with numpy and scipy workflows.
Julius provides differentiable, GPU-accelerated digital signal processing for audio and 1D signals using PyTorch, including resampling, FFT convolutions, and frequency-domain filtering.
Install it if you need signal processing as part of a neural network or GPU pipeline; skip it if you only do offline audio analysis on CPU.
GraphFrames provides distributed graph processing and analytics on Apache Spark using DataFrame abstractions, with built-in algorithms and flexible APIs like Pregel and AggregateMessages for custom graph computation.
However, verify that your Spark version and Python environment are compatible, since the latest release date and classifier information suggest the package may not…
Open3D provides data structures and algorithms for working with 3D point clouds, meshes, and RGB-D data, with GPU-accelerated processing and visualization capabilities.
PsmPy implements propensity score matching for observational studies, using logistic regression and KNN algorithms to match treated and control subjects on covariates and estimate causal effects.
However, maintenance is aging (last release 275 days ago), so consider it stable for established workflows but not actively developed; for active support or…
Causallib estimates causal effects from observational data using a modular, scikit-learn-inspired API that supports pluggable machine learning models for flexible causal inference.
Install it if you need to estimate treatment effects from non-experimental data.
Implements Coarsened Exact Matching (CEM), a statistical matching technique that reduces covariate imbalance in observational data to produce more stable causal treatment effect estimates.
However, the dormant maintenance status and unclear license terms warrant verification before production use.
MediaPipe provides pre-built machine learning models and cross-platform APIs for vision, text, and audio tasks that run on-device without sending input data to external servers.
Performs arithmetic and mathematical operations on values with uncertainties, automatically propagating errors through calculations and tracking correlations between expressions.
Install it if you work with measurements or experimental data where uncertainty quantification matters.
Provides base classes and design patterns for building scikit-learn-like and sktime-like parametric objects, making it easier to create packages that follow these established conventions.
Provides NVIDIA CUBLAS native runtime libraries for CUDA 11, enabling GPU-accelerated linear algebra operations in Python applications on x86_64 and ARM64 Linux, Windows platforms.
Install only if you are maintaining legacy code explicitly pinned to CUDA 11 and cannot upgrade; otherwise, use current nvidia-cublas packages or let your framework…
Computes approximate convex decomposition of triangle meshes by wrapping the V-HACD algorithm, taking vertex coordinates and face indices as NumPy arrays and returning a list of convex hull pairs.
However, verify the upstream license terms before use in proprietary work, and be prepared for minimal maintenance support if issues arise.
lifelines implements survival analysis methods—Kaplan-Meier, Nelson-Aalen, and regression models—to estimate time-to-event distributions and identify factors influencing when events occur.
Install it if you need to fit survival models or estimate event-time distributions.
Provides Python bindings to Microsoft's Speech Service SDK for speech-to-text, text-to-speech, and speech translation via Azure Cognitive Services.
Converts and matches country names across different classification standards (ISO 2/3 letter codes, numeric codes, UN/EU/OECD membership schemes) using regex-based matching and can build concordance matrices between schemes.
Install it if you regularly work with messy country data and can accept the copyleft requirement.
Provides the NVIDIA CUDA nvcc compiler for building CUDA applications from Python, bundling the compiler with its runtime dependencies.
PaddleX is a low-code framework for training, inference, and deployment of computer vision and document processing models, integrating pre-trained models across OCR, object detection, image classification, segmentation, and time-series tasks.
Provides cuDNN runtime libraries for GPU-accelerated deep neural network operations on NVIDIA CUDA 11 hardware.
einx provides a universal notation for expressing tensor operations across multiple frameworks (NumPy, PyTorch, JAX, TensorFlow, MLX) using a single string-based syntax that compiles to optimized code for each backend.
ORAS Python is an SDK that lets you push OCI Artifacts to OCI Conformant registries, treating them as storage from Python applications.
fastText is a library for learning word embeddings and training text classifiers using efficient algorithms based on subword information and bag-of-words methods.
Suitable for production use of existing models or training on stable datasets, but not for active development or cutting-edge NLP research.
fasttext-wheel provides Python bindings for fastText, enabling efficient learning of word representations and text classification through unsupervised and supervised training models.
However, the package is abandoned—last released in 2020 and its repository archived in 2024.
Provides bidirectional communication between Python and JavaScript in Jupyter notebooks and JupyterLab, enabling HoloViz tools to send and receive messages across the kernel-frontend boundary.
Provides the NVVM compiler IR library for building and optimizing CUDA applications on NVIDIA GPUs.