Quick exploratory summaries and plots to assess whether VR reporting appears to change over time, particularly around a conflict start.
This is designed to be run before model fitting.
Usage
vrc_diagnose_reporting(
data,
t0 = NULL,
id_cols = c("region", "time", "age", "sex", "cause"),
y_col = "y",
cause_labels = NULL,
aggregate_duplicates = TRUE
)Arguments
- data
A VR long-format data.frame.
- t0
Conflict start time (optional). May be an index or a time value.
- id_cols
Identifier columns.
- y_col
Outcome column.
- cause_labels
Optional named character vector mapping cause levels to human-readable labels.
- aggregate_duplicates
If TRUE (default), aggregates duplicate cells by summing
y.