NumPy
A biblioteca fundamental para computação científica com Python
NumPy v1.20.0
Suporte a anotações de tipos - Melhorias no desempenho através de SIMD multi-plataformas

Arrays n-dimensionais poderosas
Rápidos e versáteis, os conceitos de vetorização, indexação e broadcasting do NumPy são, na prática, o padrão em computação com arrays.

Ferramentas de computação numérica
O NumPy oferece um conjunto completo de funções matemáticas, geradores de números aleatórios, rotinas de álgebra linear, transformadas de Fourier, e mais.

Interoperabilidade
O NumPy suporta um grande número de plataformas de hardware e computação, e pode ser combinada com bibliotecas de computação com arrays esparsas, distribuidas ou em GPUs.

Alto desempenho
O núcleo do NumPy é feito de código otimizado em C. Experimente a flexibilidade do Python com a velocidade de código compilado.

Fácil de usar
A sintaxe de alto nível do NumPy torna-o acessível e produtivo para programadores de qualquer nível de experiência e formação.

Código aberto
Distribuido com uma licença BSD liberal, o NumPy é desenvolvido e mantido publicamente no GitHub por uma comunidade vibrante, responsiva, e diversa.

Try NumPy
Enable the interactive shell
>

ECOSSISTEMA

  • Nearly every scientist working in Python draws on the power of NumPy.

    NumPy brings the computational power of languages like C and Fortran to Python, a language much easier to learn and use. With this power comes simplicity: a solution in NumPy is often clear and elegant.

    Quantum Computing Statistical Computing Signal Processing Image Processing Graphs and Networks Astronomy Processes Cognitive Psychology
    A computer chip. A line graph with the line moving up. A bar chart with positive and negative values. An photograph of the mountains. A simple graph. A telescope. A human head with gears.
    QuTiP Pandas SciPy Scikit-image NetworkX AstroPy PsychoPy
    PyQuil statsmodels PyWavelets OpenCV graph-tool SunPy
    Qiskit Xarray python-control Mahotas igraph SpacePy
    Seaborn PyGSP
    Bioinformatics Bayesian Inference Mathematical Analysis Chemistry Geoscience Geographic Processing Architecture & Engineering
    A strand of DNA. A graph with a bell-shaped curve. Four mathematical symbols. A test tube. The Earth. A map. A microprocessor development board.
    BioPython PyStan SciPy Cantera Pangeo Shapely COMPAS
    Scikit-Bio PyMC3 SymPy MDAnalysis Simpeg GeoPandas City Energy Analyst
    PyEnsembl ArviZ cvxpy RDKit ObsPy Folium Sverchok
    ETE emcee FEniCS Fatiando a Terra

  • NumPy's API is the starting point when libraries are written to exploit innovative hardware, create specialized array types, or add capabilities beyond what NumPy provides.

    Array Library Capabilities & Application areas
    Dask Dask Distributed arrays and advanced parallelism for analytics, enabling performance at scale.
    CuPy CuPy NumPy-compatible array library for GPU-accelerated computing with Python.
    JAX JAX Composable transformations of NumPy programs differentiate: vectorize, just-in-time compilation to GPU/TPU.
    xarray Xarray Labeled, indexed multi-dimensional arrays for advanced analytics and visualization
    sparse Sparse NumPy-compatible sparse array library that integrates with Dask and SciPy's sparse linear algebra.
    PyTorch PyTorch Deep learning framework that accelerates the path from research prototyping to production deployment.
    TensorFlow TensorFlow An end-to-end platform for machine learning to easily build and deploy ML powered applications.
    MXNet MXNet Deep learning framework suited for flexible research prototyping and production.
    arrow Arrow A cross-language development platform for columnar in-memory data and analytics.
    xtensor xtensor Multi-dimensional arrays with broadcasting and lazy computing for numerical analysis.
    xnd XND Develop libraries for array computing, recreating NumPy's foundational concepts.
    uarray uarray Python backend system that decouples API from implementation; unumpy provides a NumPy API.
    tensorly tensorly Tensor learning, algebra and backends to seamlessly use NumPy, MXNet, PyTorch, TensorFlow or CuPy.

  • NumPy lies at the core of a rich ecosystem of data science libraries. A typical exploratory data science workflow might look like:

    For high data volumes, Dask andRay are designed to scale. Stabledeployments rely on data versioning (DVC),experiment tracking (MLFlow), andworkflow automation (Airflow andPrefect).

    Diagram of three overlapping circle. The circles labeled 'Mathematics', 'Computer Science' and 'Domain Expertise'. In the middle of the diagram, which has the three circles overlapping it, is an area labeled 'Data Science'.
  • An animated gif showing a three-dimensional graph of embeddings made in Tensorflow.

    Source: Google AI Blog

    NumPy forms the basis of powerful machine learning libraries like scikit-learn and SciPy. As machine learning grows, so does the list of libraries built on NumPy. TensorFlow’s deep learning capabilities have broad applications — among them speech and image recognition, text-based applications, time-series analysis, and video detection. PyTorch, another deep learning library, is popular among researchers in computer vision and natural language processing. MXNet is another AI package, providing blueprints and templates for deep learning.

    Statistical techniques called ensemble methods such as binning, bagging, stacking, and boosting are among the ML algorithms implemented by tools such as XGBoost, LightGBM, and CatBoost — one of the fastest inference engines. Yellowbrick and Eli5 offer machine learning visualizations.

  • NumPy is an essential component in the burgeoning Python visualization landscape, which includes Matplotlib, Seaborn, Plotly, Altair, Bokeh, Holoviz, Vispy, Napari, and PyVista, to name a few.

    NumPy’s accelerated processing of large arrays allows researchers to visualize datasets far larger than native Python could handle.

ESTUDOS DE CASO

A Primeira Imagem de um Buraco Negro
Primeira imagem de um buraco negro. É um círculo laranja em um fundo preto.
Como o NumPy, junto com outras bibliotecas como SciPy e Matplotlib que dependem do NumPy, permitiram ao Event Horizon Telescope gerar a primeira imagem de um buraco negro da história.
Descoberta de Ondas Gravitacionais
Duas esferas orbitando a si mesmas. Elas deslocam a gravidade em seu entorno.
Em 1916, Albert Einstein previu ondas gravitacionais; 100 anos depois, sua existência foi confirmada pelos cientistas do LIGO usando NumPy.
Análise Esportiva
Bola de críquete em um campo verde
A análise de críquete está mudando o jogo ao melhorar o desempenho de jogadores e times através de modelagem estatística e análise preditiva. O NumPy possibilita muitas dessas análises.
Estimação de poses usando deep learning
Análise de pose de um guepardo
DeepLabCut usa o NumPy para acelerar estudos científicos que envolvem comportamento animal para entender melhor o controle motor em várias espécies e escalas de tempo.