Skip to content

Latest commit

 

History

History
15 lines (10 loc) · 879 Bytes

README.md

File metadata and controls

15 lines (10 loc) · 879 Bytes

statmodels

Functions I developed as part of my MSc. thesis 'Modelling Pesticide Biodegradation Using Regulatory Data'. Most are simply wrappers to make model fitting more efficient. Dataset is from the EAWAG-Soil package of envipath, see https://envipath.org/wiki/index.php/EAWAG-SOIL_package

crossvalidation

Functions for internal cross-validation (10-fold and Leave-One-Out) of lm model objects. Output R2, adjR2 (where possible), RMSE, and MAE.

myplsr

Running partial least squares regression using {pls} by Bjorn Mevik. Wrapper functions to group together some modelling steps for quicker pls regression model fitting.

marginals

Based on {ggplot}, plot and facets the marginal distributions of the enumerated variables of a dataset.

unadjR2

Calculate the unadjusted R2 value. Gam objects derived using {mgcv} give adjusted R2 by default in model outputs.