numpyarray

NumPy array, or ndarray, is the core data structure of the NumPy library for Python. It represents a multi-dimensional, homogeneous collection of items, all stored as elements of a single data type in a contiguous memory block. This design enables efficient numerical computations and seamless interaction with other scientific libraries.

An array has attributes such as shape, dtype, size, and ndim. Shape specifies the extents of each dimension, dtype indicates the element type (for example int32 or float64), and size is the total number of elements. Arrays may be stored in C order (row-major) or Fortran order (column-major), affecting memory layout and indexing.

Indexing and slicing typically produce views rather than copies, so modifying a view affects the original data. NumPy also supports broadcasting, which allows operations between arrays of different shapes by virtually expanding smaller arrays. Elementwise operations, reductions, and linear algebra are performed efficiently with universal functions (ufuncs).

Arrays are created using constructors such as array, zeros, ones, empty, arange, and linspace. They can be converted to Python lists with tolist and interoperate with other libraries that accept array-like objects. High-performance routines are implemented in C and Fortran and accessed from Python.

Common use cases include numerical simulations, data analysis, and machine learning preprocessing. NumPy arrays underpin many data pipelines, and they serve as the foundation for higher-level libraries such as SciPy, pandas, and scikit-learn.

Limitations include the requirement for homogeneous data types and potential memory overhead for very large datasets. For irregular data, object arrays or specialized structures may be more appropriate.