# This file contains metadata for the Best Fit Interpolator QGIS plugin.

[general]
name=Best Fit Interpolator
qgisMinimumVersion=3.14
qgisMaximumVersion=4.99
description=Decision-support plugin for selecting, validating, and applying spatial interpolation methods in QGIS.
version=2.0
author=Laura Delgado Bejarano and Lucas Rios do Amaral
email=ladelgadobe@unal.edu.co
linkedin_laura=https://www.linkedin.com/in/laura-delgado-bejarano-09b6681a2/
linkedin_lucas=https://www.linkedin.com/in/lucas-rios-do-amaral-bb302449/

about=Best Fit Interpolator supports spatial interpolation workflows for environmental, soil, and precision-agriculture data. The plugin compares deterministic, geostatistical, machine-learning, and hybrid approaches, including IDW, TPS, ordinary kriging, REML-assisted kriging, Random Forest, SVM, and Regression Kriging. It provides data diagnostics, semivariogram visualization, validation metrics, observed-vs-predicted plots, interpolation maps, and a framework-guided decision workflow based on: Delgado Bejarano, L., Loureiro Goncalves Oliveira, A., Fiolo Pozzuto, J. V., Castaneda Sanchez, D., and Rios do Amaral, L. (2026). Performance of interpolation methods in digital soil mapping: the influence of data characteristics. Precision Agriculture, 27(1), 10. https://doi.org/10.1007/s11119-025-10311-8. Contact: ladelgadobe@unal.edu.co. LinkedIn: Laura Delgado Bejarano - https://www.linkedin.com/in/laura-delgado-bejarano-09b6681a2/; Lucas Rios do Amaral - https://www.linkedin.com/in/lucas-rios-do-amaral-bb302449/.
changelog=Version 2.0: Adds configurable Outlier diagnostic with Global Moran, Local Moran/LISA and explicit sample decisions; resets the session on reopening; separates IDW/TPS and keeps RF/residual kriging in one workspace; synchronizes palette, scale and limits with Larger View; restores the MoM experimental and REML theoretical display distinction with three-model validation; enables direct Framework RF/SVM/RK workflows and accurate executed-method reporting; compares aligned maps and adds temporary, navigable, collapsible HTML alongside PDF reports; supports normal, dense and massive processing with bounded tuning, local spatial prediction, block raster output and disclosed hold-out validation; fixes Qt5/Qt6, Matplotlib, native 32-bit NumPy indices and Python/platform ML dependency isolation; updates the English manual and its stable generic URL. Version 1.1 synchronized the Framework semivariogram preview with the active Geostatistics model; blocked interpolation for incompatible CRS or unusable spatial overlap; improved popup notifications; fixed deterministic TPS routing and the REML prediction matrix error; standardized R² labels; and added the metadata-driven About tab.

tracker=https://github.com/ladelgadobe/BestFitInterpolation/issues
repository=https://github.com/ladelgadobe/BestFitInterpolation
manual=https://github.com/ladelgadobe/BestFitInterpolation/blob/main/BestFitInterpolator_User_Manual.pdf
article=https://link.springer.com/article/10.1007/s11119-025-10311-8
article_title=Performance of interpolation methods in digital soil mapping: the influence of data characteristics
article_citation=Laura Delgado Bejarano, Agda Loureiro Gonçalves Oliveira, João Vitor Fiolo Pozzuto, Dario Castañeda Sánchez, and Lucas Rios do Amaral (2026). Performance of interpolation methods in digital soil mapping: the influence of data characteristics. Precision Agriculture, 27, Article 10. https://doi.org/10.1007/s11119-025-10311-8

hasProcessingProvider=False
tags=interpolation, geostatistics, kriging, IDW, TPS, machine learning, digital soil mapping, precision agriculture, QGIS
homepage=https://github.com/ladelgadobe/BestFitInterpolation
category=Raster
icon=icon.png
experimental=False
deprecated=False
server=False
