Analyzing Historical and Experimental Data

MathWorks data analysis products let you access, visualize and analyze data using a single environment. You can interactively explore and model your data, build customized analyses, and share your discoveries with others through reports, published code, or complete applications.

Access and Integrate Data from Multiple Sources

Without leaving MATLAB, you can access, analyze, and visualize data from:

Manage, Filter, and Preprocess Your Data

Manage and operate on your data using the data types that best suit the data’s characteristics. You can:

  • Work with basic arrays for numeric, string, logical, image, and categorical data types
  • Build heterogeneous data arrays that encapsulate mixed data types into a single data container
  • Create specialized data objects, such as statistical dataset arrays and time series objects, that simplify analysis and enable storage of metadata
  • Build new or customize existing data types for specialized data management tasks

Historical and experimental data often contains noise, bad data, or missing data. With MathWorks data analysis products you preprocess, filter, and transform your data in the same environment that you use to explore, model, and visualize it. Capabilities range from core functions, such as summary statistics, filtering and smoothing operations, interpolation, convolution, and fast Fourier transforms (FFTs), to advanced data analysis tools for statistics, signal processing, image processing, and other specialized applications

Uncover Trends, Test Assumptions, and Build Descriptive Models

Identify patterns, trends, and complex relationships, or build descriptive models using interactive tools that combine curve and surface fitting, multivariate statistics, spectral analysis, image analysis, system identification, and other analysis tasks in a single environment.

Your data-derived model can be combined with a mathematical model of the system. The result is an enhanced model that complements physics-based understanding with complex interactions, uncertainties, and other information from real system performance data.