Best Fit Interpolator

A QGIS plugin for selecting, validating, and applying spatial interpolation methods for environmental, soil, and precision-agriculture data.

Best Fit Interpolator compares deterministic, geostatistical, machine-learning, and hybrid interpolation approaches, including IDW, TPS, Ordinary Kriging, REML-assisted kriging, Random Forest, SVM, and Regression Kriging.

Version 2.0

• Outlier diagnostic combines IQR, MAD and optional Z flags with Global Moran and Local Moran/LISA; sample exclusions remain explicit user decisions.
• Reopening starts a clean session. IDW and TPS controls are separated; Regression Kriging keeps RF and residual kriging side by side with a shared adjustment dialog.
• MoM shows experimental semivariances and its model; REML shows the fitted theoretical model only. Advanced dialogs compare all three models and explain the numerical selection.
• Framework can prepare, validate and run RF, SVM and RK directly, and reports the method that successfully produced the final interpolation.
• Comparison shows aligned maps and their signed difference. Report includes an interactive summary, temporary browser HTML with collapsible sections, and PDF export.
• Map settings show palette gradients in a popup and preserve colors, scale and value limits in Larger View. Validation plots have no palette controls.
• Normal, dense and massive profiles bound tuning, prediction blocks and spatial neighborhoods; the actual validation strategy and sample count are disclosed.
• Automatic Global Moran uses all valid samples with spatial indexing and adaptive permutation counts. Outlier diagnostic retains its explicit neighborhood and permutation settings.
• Qt5/Qt6 and Matplotlib compatibility fixes restore embedded previews; native NumPy indices support 32-bit QGIS and machine-learning dependencies are isolated by Python/platform ABI.
• The updated English manual uses current Paulínia screenshots and documents outlier parameters, MoM/REML, map comparison, reports and dense/massive workflows.

Version 1.1

• Synchronized Framework and Geostatistics semivariogram previews.
• Spatial compatibility checks before interpolation.
• Popup alerts for warnings and errors.
• TPS and REML interpolation fixes.
• Standardized R² validation labels.
• About tab with documentation, support, article, and author links.

Main features

• Data diagnostics and spatial-pattern support.
• Semivariogram preview and kriging tools.
• Framework-guided method selection.
• Cross-validation metrics and observed-vs-predicted plots.
• Interpolation maps and PDF report support.

Authors

• Laura Delgado Bejarano
• Lucas Rios do Amaral

Contact: ladelgadobe@unal.edu.co

Reference article

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

User Manual

Open the current user manual

Repository

https://github.com/ladelgadobe/BestFitInterpolation