TFInterpy: A high-performance spatial interpolation Python package
Zhiwen Chen, Baorong Zhong
Abstract
Interpolation algorithms are essential tools for spatial analysis. The Kriging method satisfies the best linear unbiased estimation and is widely used in scenarios where high accuracy is required. But the program running time may be unacceptably long when using the Kriging method for large dataset. To solve the problem, we developed TFInterpy based on the TensorFlow framework. This Python package provides an open-source, cross-platform, easy-to-use API for interpolation algorithms and achieves significant speedups when applied to large-scale tasks.
Topics & Concepts
Python (programming language)Computer scienceKrigingInterpolation (computer graphics)Multivariate interpolationAlgorithmOpen sourceComputational scienceComputer engineeringSoftwareProgramming languageComputer graphics (images)Bilinear interpolationMachine learningComputer visionAnimationGeophysics and Gravity MeasurementsSoil Geostatistics and MappingPrecipitation Measurement and Analysis