sensor_wear: Calculate Sensor Wear

In cgmguru: Advanced Continuous Glucose Monitoring Analysis with High-Performance C++ Backend

Calculate Sensor Wear

Description

Calculates the percent of expected CGM readings observed. By default, the calculation uses each subject's original timestamp span from first valid reading to last valid reading. If ndays is supplied, valid readings in [end_date - ndays, end_date] are divided by the expected number of readings over that fixed retrospective window.

For fixed-window calculations, if end_date = NULL, each subject's last valid timestamp defines that subject's retrospective window. If end_date is supplied, the same endpoint is used for all subjects, which is useful for a common study cutoff or report date. Duplicate timestamps within a subject are de-duplicated after sorting, and rows with missing time or gl do not count as observed readings.

Usage

sensor_wear(df, end_date = NULL, ndays = NULL,
 reading_minutes = NULL)

Arguments

df	A dataframe containing CGM data with columns: id: Subject identifier time: POSIXct measurement timestamp gl: Glucose value in mg/dL
end_date	End timestamp for a fixed-window calculation. Requires ndays. If NULL, each subject's last valid timestamp is used. Date values are converted with as.POSIXct(), matching iglu's manual active-percent behavior.
ndays	Number of days in the fixed retrospective window. Defaults to NULL, which uses the original timestamp span.
reading_minutes	Reading interval in minutes. If NULL, it is inferred per id from the median positive difference between valid readings.

Value

A tibble with columns id, sensor_wear_percent, sensor_wear, ndays, start_date, and end_date. sensor_wear is retained as a backward-compatible alias.

See Also

detect_all_events, iglu::active_percent

Examples

library(iglu)
data(example_data_5_subject)
sensor_wear(example_data_5_subject, reading_minutes = 5)
sensor_wear(example_data_5_subject, ndays = 90, reading_minutes = 5)

cgmguru documentation built on July 8, 2026, 9:06 a.m.