Convolutional Neural Networks for Digital Soil Mapping

[Up] [Top]

Documentation for package ‘soilcnn’ version 0.1.0

Help Pages

ale_effect ALE: how the prediction changes along each variable's range.
aoa_di The DI of new rows of RAW predictor values, QC'd and scaled as the reference was.
aoa_reference The dissimilarity reference of a fitted model.
aoa_threshold The DI threshold that separates the area of applicability.
block_bootstrap A metric, or two models' difference in it, with an interval from whole blocks.
block_bootstrap_by_size The interval as the blocks grow.
calc_metrics Compute all regression metrics for one obs/pred pair.
caret_available Every caret method that could be borrowed here.
caret_spec Turn a caret method into a model_spec.
ccc Lin's Concordance Correlation Coefficient.
check_fold_plan Check a plan's folds are well formed, and describe them.
compare_importance Several importances, side by side.
conformal_calibrate Calibrate a conformal interval from held-out residuals.
conformal_cv Calibrate on the folds of a resampling run, then report coverage.
conformal_interval Turn predictions into intervals.
conformal_scaled_calibrate Calibrate an interval whose width is a fitted scale.
conformal_scaled_interval Turn predictions into intervals with a fitted scale.
conformal_scale_fit Fit the scale of an interval: |residual| on covariates that say how wrong.
context_importance Context importance: how far from the point, and at which scale, the model reads.
cv_plus_calibrate Calibrate a CV+ interval from the folds of a cross-validation.
cv_plus_interval Turn the fold models' predictions into CV+ intervals.
cv_residuals Cross-validated residuals from a tuning run, for calibration.
cv_residuals_for_config Cross-validated residuals of a configuration, from any tuning run.
dissimilarity_index The dissimilarity index of new data.
di_reference Summarise a training set for dissimilarity-index computation.
dsm_final Refit the configuration a tuning run selects, under N seeds.
dsm_importance Variable importance of a fitted model, by the method you choose.
dsm_load Load everything a run needs, and check that it fits together.
dsm_predict Predict a map from a final model.
dsm_prepare Build a patch store from a point table and a folder of aligned rasters.
dsm_report_final The declaration of a final model that already exists.
dsm_train Fit and tune a model over a resampling plan.
equal_area_blocks Square blocks of equal area, in kilometres.
example_landscape A small synthetic landscape, for the examples and the tests.
example_run A small fitted run on the example landscape, made once a session.
fold_leakage_report How much does each fold still leak?
freeze_selection Record which config was chosen, and when.
get_model Fetch a registered model.
holdout A single train/validation/test split of a point table.
holdout_cv A single train/validation/test split.
importance_groups Which channels the importance treats as one variable.
importance_map Maps of SHAP values, from an importance computed at points of the map.
importance_points Points of the map to explain: a regular grid of the model's raster cells.
importance_weights One weight per channel, for the area of applicability, from an importance.
inside_aoa Is each row inside the area of applicability?
knndm_cv Folds matched to where the map will be predicted (kNNDM).
knndm_folds Fold plan from k-fold Nearest Neighbour Distance Matching.
latest_run_dir The most recent run under 'base', by time, and only if it finished.
list_models Every registered model, as a table.
make_manual_tune_grid Build a manual grid from explicit lists of values per parameter.
make_tune_grid Generate a random hyperparameter grid (like caret's tuneLength).
model_spec Describe a model the framework can fit.
occlusion_report What each part of the patch is worth to one trained unit of a tuning run.
one_se Pick the simplest config within one standard error of the best.
paired_family_test Paired comparison between two model families.
permutation_importance Permutation importance: how much a fitted model relies on each variable.
picp_report Coverage, and whether it holds where it is needed.
plot.block_bootstrap_by_size Plot the interval as the blocks grow.
plot.dsm_importance Plot a variable importance.
plot.importance_comparison Plot importances side by side.
plot.importance_map Plot maps of SHAP values.
plot.spatial_correlogram Plot a spatial correlogram.
prediction_sample A regular sample of the prediction area, as kNNDM's 'predpoints'.
print_aoa Report what an AOA covers.
print_noise_floor Print a noise-floor report in the terms it should be read in.
print_one_se Say what one_se() did, including when it did nothing.
random_cv Random k-fold. Ignores geography by construction.
random_folds Random k-fold over the train+validation pool.
refit_importance Refit importance: what the model loses when trained without a variable.
region_cv Leave-region-out, on a grouping that already exists (biome, catchment, ...).
region_folds Fold by a grouping column: no group is split across folds.
register_model Add a model to the registry.
resolve_resampling Turn a spec into a fold plan against real points.
sage_importance SAGE: how much of the model's skill each variable carries, shared fairly.
score_test_grid Score every trained unit of a tuning run on the held-out test set.
seed_noise_floor How much does a metric move when ONLY the seed changes?
selected_config_id The configuration a final run deployed.
setup_torch_device Configure torch threads and select compute device.
set_torch_threads Set torch's thread pools.
shap_importance SHAP importance: how each variable pushes each prediction, up or down.
smear The conditional MEAN surface, from transform-space predictions.
smearing_check The smearing factors on held-out points: which one unbiases the mean surface.
smearing_factor The smearing factor, from held-out residuals in transform space.
smearing_from_run The smearing factor from a tuning run's out-of-fold predictions.
smear_map The mean surface of a median map already made.
spatial_correlogram Moran's I by distance class: how far apart two points stop resembling each other.
spatial_cv Spatially blocked k-fold: whole blocks of ground go to one fold.
spatial_folds Spatially blocked folds of a point table.
summarise_resamples Aggregate a unit-level comparison table into one row per config.
target_transform_spec The forward and inverse functions for a named target transform.