Method to compute the "parallel" mean of values across members of a collection of spectra or of a spectral object containing multiple spectra in long form.
s_mean(x, trim, na.rm, ...)# S3 method for default
s_mean(x, trim = 0, na.rm = FALSE, ...)
# S3 method for generic_spct
s_mean(x, trim = 0, na.rm = FALSE, ...)
# S3 method for source_mspct
s_mean(x, trim = 0, na.rm = FALSE, ...)
# S3 method for response_mspct
s_mean(x, trim = 0, na.rm = FALSE, ...)
# S3 method for filter_mspct
s_mean(x, trim = 0, na.rm = FALSE, ...)
# S3 method for reflector_mspct
s_mean(x, trim = 0, na.rm = FALSE, ...)
# S3 method for calibration_mspct
s_mean(x, trim = 0, na.rm = FALSE, ...)
# S3 method for cps_mspct
s_mean(x, trim = 0, na.rm = FALSE, ...)
# S3 method for raw_mspct
s_mean(x, trim = 0, na.rm = FALSE, ...)
If x is a collection spectral of objects, such as a
"filter_mspct" object, the returned object belongs to the same class
as the members of the collection, such as "filter_spct", containing
the summary spectrum, with variables with names tagged for summaries other
than mean or median.
An R object.
numeric The fraction (0 to 0.5) of observations to be trimmed from each end of x before the mean is computed. Values of trim outside that range are taken as the nearest endpoint.
logical A value indicating whether NA values should be stripped before the computation proceeds.
Further arguments passed to or from other methods.
Parallel summaries differ fundamentally from the "deepest curves" obtained through functional data analysis (FDA) in that in functional data analysis one of the input curves is returned as the deepest one based on a decision criterion. In contrast the parallel summaries from package 'photobioloy' return one or more "fictional" curves different to any of those passed as inputs. This curve is constructed from independent summaries at each wavelength value.
Method specializations compute the mean at each wavelength across a
group of spectra stored in an object of one of the classes defined in
package 'photobiolgy'. Trimming of extreme values is
possible (trimmed mean) and omission of NAs is done separately at each
wavelength. Interpolation is not applied, so all spectra in x must
share the same set of wavelengths. An error is triggered if this condition
is nor fulfilled.