Package index • AgriFusionR

The project

The object everything else operates on: management units observed over seasons, with a provenance ledger.

agri_project(): Create an agricultural analysis project
units_of() seasons_of() provenance(): Accessors for project components
demo_agri_data(): A demonstration agricultural data set

Covariates

Attaching climate, soil and remote-sensing layers through the source registry, and reducing them to one row per unit and season.

add_layer() add_climate() add_soil() add_satellite(): Attach a covariate layer to a project
build_features(): Build the model design matrix
growing_degree_days(): Growing degree days
crop_parameters(): Indicative crop thermal parameters
phenology_windows(): Derive phenological windows from accumulated thermal time

Checking

The faults that invalidate an analysis, including covariate windows that reach past the harvest they predict.

check_project(): Check a project for the faults that invalidate an analysis

Modelling

Spatial resampling by default, with the optimism of random folds reported alongside.

resample_scheme(): Build a resampling scheme
train_model(): Train and honestly validate a model
predict(<agri_model>): Predict from a fitted model
stack_weights(): Stack weights from a fitted ensemble

Interpretation

Importance, marginal effects, prediction intervals and model cards.

explain(): Explain a fitted model
uncertainty(): Prediction intervals by split conformal inference
report(): Write a model card

Plots

plot(<agri_project>): Plot a project's management units
plot(<agri_resample>): Plot a resampling scheme
plot(<agri_model>): Plot a fitted model
plot_map(): Map predictions, residuals or uncertainty
plot_effect(): Plot a marginal effect
plot_uncertainty(): Plot prediction intervals and their coverage

Extending

Adding a data provider or a learning algorithm without modifying the package.

register_source(): Register a covariate source
register_learner(): Register a learning algorithm
list_sources() list_learners(): List registered sources and learners