Implements automated variable-type detection, preprocessing,
encoding, scaling, model diagnostics, variable selection, and yield gap
computation for agricultural comparative performance analysis (CPA).
The comparative performance analysis approach is described in de Bie
(2004)
VeraCrop provides an end-to-end workflow for yield gap analysis using Comparative Performance Analysis (CPA): automatic variable-type detection, preprocessing (missing-value handling, encoding, scaling, filtering), regression diagnostics, variable selection with cross-validation, and yield gap decomposition with reporting.
# Install from source tarball
install.packages("VeraCrop_0.1.0.tar.gz", repos = NULL, type = "source")
library(VeraCrop)
# 1. Preprocess: type detection, imputation, encoding, scaling, filtering
prep <- prep_yield_gap(my_data, response_var = "Yield")
# 2. Check regression assumptions
model <- lm(Yield ~ ., data = prep$data)
diag <- check_assumptions(model, data = prep$data, plot = FALSE)
# 3. Run the full yield gap analysis
result <- yield_gap_analysis(prep$data, response = "Yield")
result$yield_gap
# 4. Visualize and export
plot_all_graphs(result)
save_results_excel(result, output_path = "yield_gap_results.xlsx")
See vignette("getting-started", package = "VeraCrop") for a complete
worked example.
| Step | Function |
|---|---|
| Variable type detection | detect_variable_types() |
| Missing value handling | handle_missing_values() |
| Encoding | encode_variables() |
| Scaling | scale_variables() |
| Filtering | remove_constant_vars(), remove_near_zero_variance(), remove_highly_correlated() |
| Full preprocessing pipeline | prep_yield_gap() |
| Regression diagnostics | check_assumptions() |
| Variable selection | select_vars_ftest() |
| Cross-validation | validate_model_with_cv() |
| Yield gap analysis | yield_gap_analysis() |
| Reporting | save_results_excel(), save_field_level_excel(), plot_all_graphs() |
MIT © Abolfazl Derakhshan