validate_cie_sky_model: Validate CIE sky models

In GastonMauroDiaz/rcaiman: CAnopy IMage ANalysis

Validate CIE sky models

Description

Validate CIE sky models fitted with fit_cie_sky_model() and ootb_fit_cie_sky_model().

Usage

validate_cie_sky_model(model, rr, k = 10)

Arguments

model	An object of the class list. The output of fit_cie_sky_model().
rr	An object of class list. The output of extract_rel_radiance() or an object with same structure and names.
k	Numeric vector of length ones. Number of folds.

Details

The validation of the CIE sky model is done with a k-fold approach with k equal to 10, following \insertCiteKohavi1995;textualrcaiman, and for the values of relative radiance. The regression of the predicted vs. observed values was done following \insertCitePineiro2008;textualrcaiman. The is_outlier were determined following \insertCiteLeys2013;textualrcaiman and with threshold equal to 3.

Value

A list with the following components:

• An object of class lm (see stats::lm()).

• predicted values.

• observed vales.

• the coefficient of determination (r^2).

• The root mean squared error (RMSE).

• The median absolute error (MAE).

• A logical vector indicating is_outlier within the sky points set attached to the 'model' argument.

References

\insertAllCited

Examples

## Not run: 
caim <- read_caim() %>% normalize_minmax()
z <- zenith_image(ncol(caim), lens())
a <- azimuth_image(z)

# See fit_cie_sky_model() for details on below file
path <- system.file("external/sky_points.csv",
                    package = "rcaiman")
sky_points <- read.csv(path)
sky_points <- sky_points[c("Y", "X")]
colnames(sky_points) <- c("row", "col")
head(sky_points)
plot(caim$Blue)
points(sky_points$col, nrow(caim) - sky_points$row, col = 2, pch = 10)

xy <- c(210, 451) #taken with click() after x11(), then hardcoded here
sun_coord <- zenith_azimuth_from_row_col(z, a, c(nrow(z) - xy[2],xy[1]))
points(sun_coord$row_col[2], nrow(caim) - sun_coord$row_col[1],
       col = 3, pch = 1)

rr <- extract_rel_radiance(caim$Blue, z, a, sky_points)

set.seed(7)
model <- fit_cie_sky_model(rr, sun_coord,
                           general_sky_type = "Clear",
                           twilight = FALSE,
                           method = "CG")
model_validation <- validate_cie_sky_model(model, rr)
model_validation$r_squared
model_validation$rmse

## End(Not run)

GastonMauroDiaz/rcaiman documentation built on April 14, 2025, 9:39 a.m.