nvRC_metrics {LightLogR}

Performance metrics for circadian response

Description

These functions compare the non-visual circadian response (see nvRC) for measured personal light exposure to the nvRC for a reference light exposure pattern, such as daylight.

Usage

nvRC_circadianDisturbance(nvRC, nvRC.ref, as.df = FALSE)

nvRC_circadianBias(nvRC, nvRC.ref, as.df = FALSE)

nvRC_relativeAmplitudeError(nvRC, nvRC.ref, as.df = FALSE)

Arguments

nvRC

Time series of non-visual circadian response (see nvRC.

nvRC.ref

Time series of non-visual circadian response circadian response (see nvRC for a reference light exposure pattern (e.g., daylight). Must be the same length as nvRC.

as.df

Logical. Should the output be returned as a data frame? Defaults to TRUE.

Details

nvRC_circadianDisturbance() calculates the circadian disturbance (CD). It is expressed as

CD(i,T)=\frac{1}{T}\int_{t_{i}}^{t_{i}+T} {\lvert r_{C}(t)-r_{C}^{ref}(t)\rvert dt},

and quantifies the total difference between the measured circadian response and the circadian response to a reference profile.

nvRC_circadianBias() calculates the circadian bias (CB). It is expressed as

CB(i,T)=\frac{1}{T}\int_{t_{i}}^{t_{i}+T} {(r_{C}(t)-r_{C}^{ref}(t))dt},

and provides a measure of the overall trend for the difference in circadian response, i.e. positive values for overestimating and negative for underestimating between the measured circadian response and the circadian response to a reference profile.

nvRC_relativeAmplitudeError() calculates the relative amplitude error (RAE). It is expressed as

RAE(i,T)=r_{C,max}-r_{C,max}^{ref},

and quantifies the difference between the maximum response achieved in a period to the reference signal.

Value

A numeric value or single column data frame.

References

Amundadottir, M.L. (2016). Light-driven model for identifying indicators of non-visual health potential in the built environment [Doctoral dissertation, EPFL]. EPFL infoscience. doi:10.5075/epfl-thesis-7146

Examples

dataset1 <- 
  tibble::tibble(
    Id = rep("B", 60 * 24),
    Datetime = lubridate::as_datetime(0) + lubridate::minutes(0:(60*24-1)),
    Illuminance = c(rep(0, 60*8), rep(sample(1:1000, 16, replace = TRUE), each = 60)),
    MEDI = Illuminance * rep(sample(0.5:1.5, 24, replace = TRUE), each = 60),
  ) %>%
  dplyr::mutate(
    nvRC = nvRC(MEDI, Illuminance, Datetime, sleep.onset = hms::as_hms("22:00:00"))
  )

dataset.reference <-
  tibble::tibble(
    Id = rep("Daylight", 60 * 24),
    Datetime = lubridate::as_datetime(0) + lubridate::minutes(0:(60*24-1)),
    Illuminance = c(rep(0, 60*6), rep(10000, 12*60), rep(0, 60*6)),
    MEDI = Illuminance
  ) %>%
  dplyr::mutate(
    nvRC = nvRC(MEDI, Illuminance, Datetime, sleep.onset = hms::as_hms("22:00:00"))
  )

# Circadian disturbance
nvRC_circadianDisturbance(dataset1$nvRC, dataset.reference$nvRC)

# Circadian bias
nvRC_circadianBias(dataset1$nvRC, dataset.reference$nvRC)

# Relative amplitude error
nvRC_relativeAmplitudeError(dataset1$nvRC, dataset.reference$nvRC)