An overview of the Space Time Pattern Mining toolbox

The Space Time Pattern Mining toolbox contains statistical tools for analyzing data distributions and patterns in the context of both space and time. The toolbox contains toolsets for clustering analysis, forecasting, and a tool that creates a space-time cube layer that can be used to visualize the data stored in the space-time netCDF cube in both 2D and 3D. The toolbox also includes options for estimating and filling missing values in the data before cube creation.

Note:

See Visualize the space-time cube for strategies on how to visualize the contents of a space-time cube.

ToolsetDescription

Space Time Cube Creation toolset

The Space Time Cube Creation toolset contains tools that summarize data into a netCDF data structure that can be used as input to tools in the Space Time Pattern Analysis and Time Series Forecasting toolsets. The data aggregated and summarized into the space-time cube must have time stamps but can come from many different formats such as a set of points, panel data, related tables, or multidimensional raster layers. When the space-time cube is created, initial summary statistics and trend are calculated.

Space Time Cube Visualization toolset

The Space Time Cube Visualization toolset contains tools to create a space-time cube layer and visualize the variables stored in the space-time cube in 2D and 3D. Space-time cube layers can be used to visualize the cube and its structure, understand how the cube aggregation process works, and visualize patterns over time at specific locations of interest.

Space Time Pattern Analysis toolset

Using the analytical and statistical tools in the Space Time Pattern Analysis toolset, you can identify patterns and interrogate the data in a space-time cube. After the space-time cube is created, use these analysis tools to gain a better understanding of the data aggregated in the cube.

Time Series Forecasting toolset

The tools in the Time Series Forecasting toolset allow you to forecast and estimate future values at locations in a space-time cube as well as evaluate and compare forecast models for each location. Various time series forecasting models are available, including simple curve fitting, exponential smoothing, and a forest-based method.

Utilities toolset

The Utilities toolset contains tools that perform data conversion tasks on space-time cubes and on datasets prior to the creation of a space-time cube. Use these tools to fill missing values in data, spatially or temporally subset cubes, or describe cube properties.

Additional resources

The Spatial Statistics Resources page contains a variety of resources to help you use the Spatial Statistics and Space Time Pattern Mining tools, including the following:

  • Hands-on tutorials
  • Workshop videos and presentations
  • Training and web seminars
  • Links to books, articles, and technical papers
  • Sample scripts and case studies

Related topics


In this topic
  1. Additional resources