Fix: use dose-relative times for manual interval parameter renaming

[Gero1999]

Issue

Closes #1168

Description

Two bugs in the descriptive statistics module:

1. Parameter columns not displayed in the table. The parameter selector was refactored from updatePickerInput (raw column names like CMAX[ng/mL]) to selector_label (cleaned names like CMAX). But summary_stats_filtered still used select(any_of(input$select_display_parameters)) — the clean names didn't match the actual column names with units, so any_of() silently dropped all parameter columns.

2. Interval parameter renaming mismatch. Commit 742ca2434 changed signif(start_dose) / signif(end_dose) to start / end in descriptive_statistics.R. This produced absolute interval times instead of dose-relative times, mismatching pivot_wider_pknca_results which uses signif(start_dose) / signif(end_dose). The same incorrect convention was propagated to parameter_plots.R and flexible_violinboxplot.R.

Changes

• descriptive_statistics.R — map clean parameter names back to actual column names (with units) before selecting; revert interval renaming to signif(start_dose) / signif(end_dose)
• parameter_plots.R — align interval renaming with pivot_wider_pknca_results
• flexible_violinboxplot.R — align interval renaming with pivot_wider_pknca_results

How to test

1. Upload data and run NCA, if possible with multiple profiles
2. Go to Results > Descriptive Statistics — verify parameter columns (CMAX, AUC, etc.) show numeric values
3. If manual AUC intervals are configured, verify interval parameters appear correctly (e.g., AUCINT_0-12)
4. Go to Results > Parameter Plots — verify interval parameters appear in the selector

Contributor checklist

• Code passes lintr checks
• Code passes all unit tests
• New logic covered by unit tests
• New logic is documented
• App or package changes are reflected in NEWS
• Package version is incremented
• R script works with the new implementation (if applicable)
• Settings upload works with the new implementation (if applicable)

[Shaakon35]

add_label_attribute() in R/pivot_wider_pknca_results.R (line 218) still uses start / end for manual interval naming:

type_interval == "manual" ~ paste0(
  PPTESTCD, "_", start, "-", end,
  ifelse(PPSTRESU != "", paste0("[", PPSTRESU, "]"), "")
)

This is the same inconsistency this PR fixes in the other three files. The column labels generated here would use absolute times while the column names (from pivot_wider_pknca_results) use dose-relative times via signif(start_dose) / signif(end_dose).

Suggested fix:

type_interval == "manual" ~ paste0(
  PPTESTCD, "_", signif(start_dose), "-", signif(end_dose),
  ifelse(PPSTRESU != "", paste0("[", PPSTRESU, "]"), "")
)

[Shaakon35]

looks good