LightLogR (version 0.9.2)

spectral_reconstruction: Reconstruct spectral irradiance from sensor counts

Description

This function takes sensor data in the form of (normalized) counts and reconstructs a spectral power distribution (SPD) through a calibration matrix. The matrix takes the form of sensor channel x wavelength, and the spectrum results form a linear combination of counts x calibration-value for any wavelength in the matrix. Handles multiple sensor readings by returning a list of spectra

Usage

spectral_reconstruction(
  sensor_channels,
  calibration_matrix,
  format = c("long", "wide")
)

Value

• "long": List of tibbles (wavelength, irradiance)

• "wide": Dataframe with wavelength columns and one row per spectrum

Arguments

sensor_channels

Named numeric vector or dataframe with sensor readings. Names must match calibration matrix columns.

calibration_matrix

Matrix or dataframe with sensor-named columns and wavelength-indexed rows

format

Output format: "long" (list of tibbles) or "wide" (dataframe)

Details

Please note that calibration matrices are not provided by LightLogR, but can be provided by a wearable device manufacturer. Counts can be normalized with the normalize_counts() function, provided that the output also contains a gain column.

See Also

Other Spectrum: normalize_counts(), spectral_integration()

Examples

# Calibration matrix example
calib <- matrix(1:12, ncol=3, dimnames = list(400:403, c("R", "G", "B")))

# Named vector input
spectral_reconstruction(c(R=1, G=2, B=3), calib)

# Dataframe input
df <- data.frame(R=1, G=2, B=3, other_col=10)
spectral_reconstruction(dplyr::select(df, R:B), calib)

# Multiple spectra: as list columns
df <- data.frame(Measurement = c(1,2), R=c(1,2), G=c(2,4), B=c(3,6))
df <- 
df |> 
  dplyr::mutate(
      Spectrum = spectral_reconstruction(dplyr::pick(R:B), calib)
      )
df |> tidyr::unnest(Spectrum)

# Multiple spectra: as extended dataframes
df |> 
  dplyr::mutate(
      Spectrum = spectral_reconstruction(dplyr::pick(R:B), calib, "wide"))