code/090-reports.R

In pat-s/2019-feature-selection: Research Compendium package for the publication "Monitoring forest health using hyperspectral imagery: Does feature selection improve the performance of machine-learning techniques?"

reports_plan_paper <- drake_plan(
  eda_wfr = target({
    wflow_publish(knitr_in("analysis/eda.Rmd"), view = FALSE, verbose = TRUE)
    copy_figures()
  }),
  spectral_signatures_wfr = target({
    wflow_publish(knitr_in("analysis/spectral-signatures.Rmd"), view = FALSE)
    copy_figures()
  }),
  filter_correlations_wfr = target({
    wflow_publish(knitr_in("analysis/filter-correlation.Rmd"),
      view = FALSE,
      verbose = TRUE
    )
    copy_figures()
  }),
  eval_performance_wfr = target({
    wflow_publish(knitr_in("analysis/eval-performance.Rmd"),
      view = FALSE,
      verbose = TRUE
    )
    copy_figures()
  }),
  ### not used/published anymore since the whole idea is questionable
  # response_normality_wfr = target({
  #   wflow_publish(knitr_in("analysis/response-normality.Rmd"),
  #     view = FALSE,
  #     verbose = TRUE
  #   )
  #   copy_figures()
  # }),
  feature_importance_wfr = target({
    wflow_publish(knitr_in("analysis/feature-importance.Rmd"),
      view = FALSE,
      verbose = TRUE
    )
    copy_figures()
  }),

  inspect_glmnet_wfr = target({
    wflow_publish(knitr_in("analysis/inspect-glmnet-models.Rmd"),
                  view = FALSE,
                  verbose = TRUE
    )
    copy_figures()
  }),
)

reports_plan_project <- drake_plan(
  defoliation_maps_wfr = target(
    wflow_publish(knitr_in("analysis/report-defoliation.Rmd"), view = FALSE)
  )
)