GSM YAML Workflow: HTML Report

This vignette walks through the end-to-end GSM YAML workflow to generate an HTML report for the timeline KRI using the funnel-based z-score method. It follows the standard four-phase GSM pipeline: mapping raw data into a standardised layer, running the metric analysis, assembling the reporting data, and rendering the HTML output. The result is the Sample Report linked in the navbar.

Prepare Source Data

lSource <- list(
  Source_STUDY = clindata::ctms_study,
  Source_SITE = clindata::ctms_site,
  Source_SUBJ = clindata::rawplus_dm,
  Source_AE = clindata::rawplus_ae %>% dplyr::select(-aetoxgr),
  # gsm.mapping expects a studyid column, but clindata::edc_data_pages uses
  # protocolname instead. DATAENT itself is not used by gsm.timez but is
  # required by gsm.mapping's AE workflow.
  Source_DATAENT = clindata::edc_data_pages %>%
    dplyr::mutate(studyid = protocolname),
  Source_VISIT = clindata::rawplus_visdt
)

Step 1 — Map Raw Data

Use Ingest() to standardise column names based on the workflow specs, then run the mapping workflows to produce the Mapped_* data layer.

mapping_wf <- gsm.core::MakeWorkflowList(
  strPath = system.file("workflow/1_mappings", package = "gsm.mapping"),
  strNames = c("SUBJ", "AE", "VISIT", "SITE", "STUDY", "COUNTRY")
)
#> Warning: `MakeWorkflowList()` was deprecated in gsm.core 1.3.0.
#> ℹ Please use `workr::MakeWorkflowList()` instead.
#> This warning is displayed once per session.
#> Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
#> generated.
mapping_spec <- gsm.mapping::CombineSpecs(mapping_wf)
lRaw <- gsm.mapping::Ingest(lSource, mapping_spec)
lMapped <- gsm.core::RunWorkflows(mapping_wf, lRaw)
#> Warning: `RunWorkflows()` was deprecated in gsm.core 1.3.0.
#> ℹ Please use `workr::RunWorkflows()` instead.
#> This warning is displayed once per session.
#> Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
#> generated.

names(lMapped)
#> [1] "Mapped_AE"      "Mapped_SUBJ"    "Mapped_VISIT"   "Mapped_DATAENT"
#> [5] "Mapped_COUNTRY" "Mapped_SITE"    "Mapped_STUDY"

Step 2 — Run Analysis

Run the kri_TimeZFunnel metric workflow to produce the analysis results.

metrics_wf <- gsm.core::MakeWorkflowList(
  strPath = system.file("workflow/2_metrics", package = "gsm.timez"),
  strNames = "kri_TimeZFunnel",
  bExact = TRUE
)
lAnalyzed <- gsm.core::RunWorkflows(metrics_wf, lMapped)

names(lAnalyzed)
#> [1] "Analysis_kri_TimeZFunnel"

Step 3 — Prepare Reporting Data

The 3_reporting workflows stack and reshape the analysis results into the flat data frames expected by the report template.

reporting_wf <- gsm.core::MakeWorkflowList(
  strPath = system.file("workflow/3_reporting", package = "gsm.timez")
)
lReporting <- gsm.core::RunWorkflows(
  reporting_wf,
  c(lMapped, list(lAnalyzed = lAnalyzed, lWorkflows = metrics_wf))
)
#> Warning: `RunQuery()` was deprecated in gsm.core 1.3.0.
#> ℹ Please use `workr::RunQuery()` instead.
#> ℹ The deprecated feature was likely used in the workr package.
#>   Please report the issue to the authors.
#> This warning is displayed once per session.
#> Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
#> generated.

names(lReporting)
#> [1] "Reporting_Bounds"            "Reporting_Results_allmonths"
#> [3] "Reporting_Groups"            "Reporting_Metrics"          
#> [5] "Reporting_Results"

Step 4 — Generate HTML Report

The 4_modules workflow creates all charts and renders the HTML report using gsm.timez’s custom template (inst/report/Report_TimeZFunnel.Rmd).

module_wf <- gsm.core::MakeWorkflowList(
  strPath = system.file("workflow/4_modules", package = "gsm.timez"),
  strNames = "Report_TimeZFunnel",
  bExact = TRUE
)
lReport <- gsm.core::RunWorkflows(module_wf, lReporting)

lReport$Module_Report_TimeZFunnel
#> [1] "/home/runner/work/gsm.timez/gsm.timez/vignettes/kri_report_AAAA0000000_Site_20261004.html"

The workflow generates Report_TimeZFunnel.html. See the Sample Report for an example output.