Announcement: All noncommercial projects registered to use Earth Engine before April 15, 2025 must verify noncommercial eligibility to maintain Earth Engine access.
Stay organized with collections
Save and categorize content based on your preferences.
Creates a reducer that reduces some number of 1-D arrays of the same length N to a covariance matrix of shape NxN. This reducer uses the one-pass covariance formula from Sandia National Laboratories Technical Report SAND2008-6212, which can lose accuracy if the values span a large range.
[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Missing the information I need","missingTheInformationINeed","thumb-down"],["Too complicated / too many steps","tooComplicatedTooManySteps","thumb-down"],["Out of date","outOfDate","thumb-down"],["Samples / code issue","samplesCodeIssue","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2024-07-13 UTC."],[[["Creates a reducer for calculating the covariance matrix from multiple 1-D arrays."],["Utilizes a one-pass covariance formula which may be less accurate with large value ranges."],["Output is an NxN covariance matrix where N is the length of the input arrays."],["The reducer is created using `ee.Reducer.covariance()`."]]],["This describes a reducer function that computes a covariance matrix from multiple 1-D arrays of identical length. The output is an NxN covariance matrix. It employs a one-pass covariance formula, as detailed in Sandia National Laboratories Technical Report SAND2008-6212. The key action is reducing multiple arrays into a single matrix, and the critical information is the use of a specific formula which can result in decreased accuracy if the input data contains a broad range of values.\n"]]