NumPy Tutorial
NumPy (Numerical Python) is the foundational library for scientific computing in Python, providing a powerful N-dimensional array object and a vast collection of mathematical functions for fast numerical operations. This comprehensive guide covers everything from fundamental NumPy concepts to advanced implementation techniques, helping developers master NumPy for building data science pipelines, machine learning models, and scientific computing applications. Whether you're just getting started with Python's numerical ecosystem or looking to deepen your understanding of array programming, NumPy delivers the performance and flexibility you need through its efficient, vectorized approach to computation.
NumPy Arrays
NumPy Arrays (ndarray) BasicsLearn about the ndarray, NumPy's core data structure for storing and manipulating homogeneous numerical data. Understand how ndarrays differ from Python lists in terms of data type, memory layout, and supported operations. Create one-dimensional and multi-dimensional arrays and access their elements using index notation. Build a solid understanding of ndarray before exploring NumPy's more advanced features.NumPy Data Types and dtypeExplore the wide range of data types NumPy supports including int, float, complex, bool, and string dtypes. Learn how to specify and inspect the dtype of an array using the dtype attribute and type codes. Understand how choosing the right dtype impacts memory usage and numerical precision in your computations. Convert arrays between different data types using the astype function with explicit dtype arguments.NumPy Array Attributes and PropertiesLearn how to inspect NumPy arrays using key attributes like shape, ndim, size, dtype, and itemsize. Understand what each attribute tells you about the structure and memory layout of an array. Use these properties to write flexible code that adapts to arrays of different dimensions and sizes. Access array attributes as the first step in debugging and understanding unfamiliar NumPy data structures.Creating NumPy Arrays - Different MethodsDiscover the many ways to create NumPy arrays including array, zeros, ones, full, arange, linspace, and eye. Understand when to use each creation method based on the type of data you need to generate. Create arrays from Python lists, ranges, and other sequences using the array and asarray functions. Use specialized creation functions to generate identity matrices, random arrays, and evenly spaced sequences.NumPy Array Reshaping and ResizingLearn how to change the shape of a NumPy array without changing its data using reshape and ravel. Understand the difference between reshape which returns a view and resize which modifies the array in place. Use flatten to always get a copy and ravel to get a flattened view when possible for better performance. Apply reshaping to prepare arrays for matrix operations, machine learning models, and data processing pipelines.NumPy Array Indexing and SlicingLearn how to access individual elements and subarrays using basic indexing and slicing in NumPy. Explore advanced indexing techniques including boolean indexing, fancy indexing, and integer array indexing. Use slicing with start, stop, and step values to extract rows, columns, and subarrays from multidimensional arrays. Master NumPy indexing to efficiently select, filter, and manipulate data without writing explicit loops.NumPy Array Copying and Views vs CopiesUnderstand the critical difference between views and copies in NumPy and how it affects your data. Learn that slicing a NumPy array returns a view that shares memory with the original array. Use the copy method to create an independent duplicate that does not affect the original when modified. Avoid subtle bugs caused by unintended shared memory by knowing exactly when NumPy returns a view versus a copy. NumPy Mathematical Functions
NumPy Mathematical Operations and Universal FunctionsLearn how NumPy applies mathematical operations element-wise across arrays without writing explicit loops. Understand universal functions or ufuncs as optimized C-level functions that operate on ndarrays efficiently. Explore arithmetic ufuncs like add, subtract, multiply, divide, power, and mod for fast array math. Use ufuncs as the foundation for fast and readable numerical computation throughout your NumPy workflows.NumPy Array Broadcasting Rules and ExamplesUnderstand NumPy's broadcasting mechanism that allows operations between arrays of different but compatible shapes. Learn the broadcasting rules that NumPy follows to automatically expand smaller arrays to match larger ones. See practical examples of broadcasting for adding scalars to arrays, row-wise and column-wise operations. Use broadcasting to write concise and efficient array operations without manually reshaping or tiling data.NumPy Aggregate Functions (sum, mean, min, max, std)Learn how to compute summary statistics across entire arrays or along specific axes using NumPy aggregate functions. Use sum, mean, min, max, std, var, and median to analyze numerical data quickly and efficiently. Understand how the axis parameter controls whether aggregation happens across rows, columns, or the full array. Apply aggregate functions in data analysis, feature engineering, and scientific computing workflows.NumPy Conditional Operations and where() FunctionLearn how to apply conditions to NumPy arrays using boolean masks and the powerful where function. Use np.where to select values from two arrays based on a condition in a single vectorized operation. Combine conditional logic with array operations to filter, replace, and transform data without Python loops. Apply conditional operations for data cleaning, feature engineering, and applying business logic to large datasets.NumPy Array Comparison OperationsLearn how to compare NumPy arrays element-wise using comparison operators like greater than, less than, and equal to. Understand how comparison operations return boolean arrays that can be used for masking and filtering. Use np.equal, np.greater, np.less, and related ufuncs for explicit and readable array comparisons. Combine comparison results with logical operators like np.logical_and and np.logical_or for complex conditions.NumPy Trigonometric FunctionsExplore NumPy's full set of trigonometric functions including sin, cos, tan, arcsin, arccos, and arctan. Understand how these functions operate element-wise on arrays and expect angles in radians by default. Use np.deg2rad and np.rad2deg to convert between degrees and radians when working with real-world angle data. Apply trigonometric functions in signal processing, geometry calculations, physics simulations, and data visualization.NumPy Logarithmic and Exponential FunctionsLearn how to apply logarithmic and exponential operations to NumPy arrays using built-in ufuncs. Use np.log, np.log2, np.log10, and np.exp for natural log, base-2 log, base-10 log, and exponential calculations. Understand how these functions handle edge cases like zero and negative values and what results to expect. Apply logarithmic scaling and exponential growth calculations in machine learning, finance, and scientific computing.NumPy Polynomial FunctionsLearn how to work with polynomials in NumPy using the numpy.polynomial module and poly1d class. Evaluate, differentiate, integrate, and find roots of polynomials using built-in NumPy functions. Use np.polyfit to fit a polynomial curve to data points using least squares regression. Apply polynomial functions in curve fitting, interpolation, signal processing, and numerical analysis workflows. NumPy Array Manipulation
NumPy Array Concatenation and StackingLearn how to combine multiple NumPy arrays using concatenate, stack, hstack, vstack, and dstack functions. Understand the difference between concatenate which joins along an existing axis and stack which creates a new axis. Use hstack and vstack as convenient shortcuts for horizontal and vertical array joining operations. Apply array joining techniques to merge datasets, combine features, and assemble multi-dimensional data structures.NumPy Array Sorting and SearchingLearn how to sort NumPy arrays using sort and argsort for both in-place and index-based sorting. Understand how the axis parameter controls whether sorting happens along rows, columns, or the full array. Use searchsorted for efficient binary search within sorted arrays and argmin and argmax to find extreme value positions. Apply sorting and searching functions to rank data, find thresholds, and organize numerical datasets efficiently.NumPy Set OperationsLearn how to perform mathematical set operations on NumPy arrays using functions like union1d, intersect1d, and setdiff1d. Use np.unique to find distinct elements and sort them in a single efficient operation. Understand how set operations work on flattened arrays and return sorted one-dimensional results. Apply set operations to compare datasets, find common elements, and remove duplicates in numerical data. NumPy Statistical Functions
NumPy Linear Algebra
NumPy Linear Algebra OperationsLearn how to perform linear algebra computations in NumPy using the numpy.linalg module. Compute matrix multiplication using np.dot and the at symbol operator for clean mathematical notation. Use linalg functions like inv, det, eig, svd, and solve for matrix decomposition and system of equations. Apply NumPy linear algebra in machine learning algorithms, computer graphics, and scientific simulations.NumPy Matrix OperationsLearn how to perform matrix-specific operations in NumPy including transpose, inverse, determinant, and multiplication. Use np.matmul and the at symbol operator for matrix multiplication that follows standard linear algebra rules. Understand the difference between element-wise array multiplication and true matrix multiplication in NumPy. Apply matrix operations for solving linear systems, transformations, and implementing machine learning algorithms from scratch. NumPy Advanced
NumPy Fourier Transform FunctionsLearn how to apply Fast Fourier Transform operations using NumPy's numpy.fft module for signal analysis. Use np.fft.fft, np.fft.ifft, np.fft.fft2, and np.fft.fftfreq for frequency domain transformations. Understand how Fourier transforms convert time-domain signals into their frequency components for analysis. Apply FFT operations in audio processing, image filtering, vibration analysis, and scientific data interpretation.NumPy String OperationsLearn how to perform string operations on arrays of strings using NumPy's numpy.char module. Use functions like add, multiply, upper, lower, strip, split, replace, and startswith on string arrays. Understand how numpy.char functions apply string methods element-wise across entire arrays efficiently. Apply NumPy string operations for text preprocessing, data cleaning, and working with string-type structured arrays.NumPy Date and Time FunctionsLearn how NumPy handles dates and times using the datetime64 and timedelta64 data types. Create date arrays, perform date arithmetic, and calculate time differences using NumPy datetime functions. Understand how NumPy's datetime64 integrates with pandas for time series analysis and data manipulation. Apply NumPy date and time operations for scheduling, financial calculations, and time-based data processing.NumPy Memory Layout and PerformanceUnderstand how NumPy stores array data in memory using C-order row-major and Fortran-order column-major layouts. Learn how memory contiguity affects the speed of array operations and how to check layout using flags. Use np.ascontiguousarray and np.asfortranarray to convert between memory layouts for performance optimization. Write memory-aware NumPy code that minimizes unnecessary copying and maximizes computational throughput.NumPy Integration with Other LibrariesLearn how NumPy serves as the common data layer that connects Python's major scientific computing libraries. Understand how pandas, scikit-learn, TensorFlow, PyTorch, and Matplotlib all accept and return NumPy arrays. Use NumPy arrays as the interchange format for passing data between different tools in your data pipeline. Build end-to-end data science workflows by leveraging NumPy's compatibility across the Python ecosystem.NumPy File Input/Output OperationsLearn how to save and load NumPy arrays from disk using np.save, np.load, np.savetxt, and np.loadtxt. Use the npy and npz file formats for efficient binary storage of single and multiple arrays respectively. Load structured data from CSV and text files directly into NumPy arrays using genfromtxt and loadtxt. Apply file I/O functions to persist computed arrays, share datasets, and build reproducible data workflows.NumPy Structured Arrays and Record ArraysLearn how NumPy structured arrays let you store heterogeneous data with named fields in a single array. Define custom dtypes with multiple field names and types to create table-like data structures in NumPy. Access fields by name and perform operations on specific columns just like a lightweight database table. Use structured arrays for working with binary data formats, C structs, and fixed-schema scientific datasets.NumPy Masked ArraysLearn how NumPy masked arrays let you work with datasets that contain missing, invalid, or ignored values. Use numpy.ma to create masked arrays where specific elements are excluded from computations automatically. Apply masks based on conditions to filter outliers and handle missing data without removing array elements. Use masked arrays in scientific computing, sensor data processing, and geospatial analysis where gaps in data are common.NumPy Memory MappingLearn how NumPy memory mapping lets you work with large array files on disk without loading them entirely into RAM. Use np.memmap to create memory-mapped arrays that read and write data directly from disk as needed. Understand how memory mapping enables processing of datasets larger than available system memory efficiently. Apply memory mapping for large-scale numerical computing, genomics data, image processing, and big data workflows.NumPy Custom Data Types and User-defined FunctionsLearn how to define custom dtypes in NumPy for representing complex or domain-specific data structures. Use np.vectorize and np.frompyfunc to apply custom Python functions element-wise across NumPy arrays. Understand the performance trade-offs of vectorized Python functions compared to native NumPy ufuncs. Extend NumPy's capabilities with user-defined functions and custom types for specialized scientific computing needs.NumPy Error Handling and Debugging ArraysLearn how to manage numerical errors in NumPy using np.seterr to control how division by zero and overflow are handled. Use np.isnan, np.isinf, and np.isfinite to detect and filter invalid values in your arrays. Understand common NumPy warnings and errors and what they tell you about your data and operations. Write robust NumPy code that handles edge cases gracefully and produces reliable numerical results.