Fit a Govardovskii visual-pigment template to measured spectral sensitivity.

Finds the peak wavelength (\(\lambda_{\max}\)) that minimises the sum of squared residuals between the Govardovskii (2000) template and the supplied sensitivity measurements. Uses optimize for an exact single-parameter search, so the result is a continuous estimate rather than a grid point.

Usage

fit_sensitivity(
  lambda,
  sensitivity,
  chromophore = c("A1", "A2"),
  lambda_max_range = c(300, 700)
)

Arguments

lambda

Wavelengths at which sensitivity was measured (nm).

sensitivity

Measured spectral sensitivity. Normalised to \([0, 1]\) internally; need not be pre-normalised. NA values are silently dropped.

chromophore

"A1" (11-cis retinal, default) or "A2" (11-cis 3,4-didehydroretinal).

lambda_max_range

Length-2 numeric vector giving the search interval for \(\lambda_{\max}\) in nm. Default c(300, 700).

Value

A list with components:

lambda_max

Best-fit peak wavelength (nm).

chromophore

Chromophore used.

SS

Residual sum of squares at the optimum.

fitted

Data frame with columns lambda and S giving the template evaluated on the input wavelength grid.

References

Govardovskii VI, Fyhrquist N, Reuter T, Kuzmin DG, Donner K (2000) In search of the visual pigment template. Visual Neuroscience 17:509–528.

See also

govardovskii_template

Examples

lam   <- seq(400, 700, by = 10)
true  <- govardovskii_template(lam, lambda_max = 530)$S
noisy <- pmax(0, true + rnorm(length(true), 0, 0.02))
fit   <- fit_sensitivity(lam, noisy)
fit$lambda_max  # should be close to 530
#> [1] 529.7793