Skip to contents

Randomization-to-event scatter

Source: R/pd_RandToDeathScatter.R
pd_RandToDeathScatter.Rd

`r lifecycle::badge("experimental")`

Scatter of every enrolled subject from [pd_Classify()]: x = `x_anchor` (death day for deaths; the window boundary for "alive at window"; follow-up for "alive prior"; discontinuation day for discontinuations), y = `follow_up` (days from randomization to snapshot). Colored by category; each category (including the two death categories) is a separate legend entry (SI-1 design decision: no grouped "Death within `nWindowDays` days" heading). Pass `vXRange`/`vYRange` (computed study-wide by the report) to fix a shared range across the study/country/site views (AXIS-1). Each point's `customdata` packs `[hover, country, invid]` so the report can filter points client-side.

Usage

pd_RandToDeathScatter(
  dfClassified,
  nWindowDays = 90,
  vXRange = NULL,
  vYRange = NULL
)

Arguments

dfClassified

`data.frame` Output of [pd_Classify()]. May already carry the `hover`/`pd_customdata` columns built by [pd_ScatterData()] — in that case the per-point build is skipped (idempotent).

nWindowDays

`numeric` Window in days (color/legend vocabulary). Default 90.

vXRange

`numeric(2)` Optional fixed x-axis range. `NULL` autoranges.

vYRange

`numeric(2)` Optional fixed y-axis range. `NULL` autoranges.

Value

A `plotly` htmlwidget.