qplot_kSp: [!+] Plot spectroscopic curves and spectral components...

In GegznaV/spHelper: Extension of Package `hyperSpec` and Various Convenience Functions

[!+] Plot spectroscopic curves and spectral components (a.k.a. loadings)

Description

Plot spectroscopic curves in different colors.

Functions qplot_kSp, qplot_kSpFacets are designed to plot spectra of spectral components (a.k.a. loadings), extracted in principal component analysis (e.g., princomp), variaus matrix factorization methods (e.g., nmf)) and some other dimension reduction methods.
The function qplot_sp is convenient for all spectroscopic curves as uses no fill.

Usage

qplot_kSp(
  loadings,
  Title = "Components",
  xLabel = labels(loadings, ".wavelength"),
  yLabel = labels(loadings, "spc"),
  by = "cNames",
  palette = hyGet_palette(loadings),
  legendName = FALSE,
  filled = TRUE,
  normalize = FALSE,
  facets = FALSE,
  subTitle = NULL,
  names.in = NULL,
  line_size = 1
)

qplot_kSpFacets(
  ...,
  Title = "Components (stacked)",
  normalize = "auto",
  facets = TRUE
)

qplot_sp(..., Title = NULL, filled = FALSE, line_size = 0.5)

Arguments

loadings	Components' spectra (loadings) as a hyperSpec object.
Title	The main title for the plot.
xLabel	A label for abscisa (x) axis.
yLabel	A label for ordinate (y) axis.
by	A name of variable in loadings that contains variable names. If indicated variable does not exist, row names are used instead. Default is names = 'cNames'
legendName	A name of a legend. Possible entries: logical FALSE - a legend without a name; logical TRUE - a label of a variable name is used as a name of a legend (legendName <- labels(loadings,name)); ...manual input of the name.
filled	Logical. If TRUE, colored fill is used. If FALSE, no fill is used. (Fill is an area between ordinate axis and the curve.)
normalize	A flag that indicates whether components should be normalized before plotting. Possible selections: FALSE, 0 - do not normalize; TRUE, +1 - normalize to max value; -TRUE, -1 - normalize to min value [! this choise can give unexpected results]; "auto" - one of the choices above is selected by determining if spectra have any possitive and any negative peaks: Below0 <- any(loadings$spc < 0); Above0 <- any(loadings$spc > 0); normalize <- (Above0 - Below0)
facets	A logical flag. If TRUE, spectra are plotted on separate graphs/facets (implemented by function facet_grid). If FALSE, all spectra are plotted on one facet.
subTitle	The second line of title, which will be smaller and and in italics.
names.in	DEPRECATED. Use by instead.
...	(The same parameters as above).

Details

qplot_kSp plots spectra on one graph.
qplot_kSpFacets plots spectra on separate graphs (facets).

Value

A ggplot object. More details in package ggplot2.

Note

Before using as an argument loadings, a matrix of components/loadings (loadings) must be treated with function decomposition (or equivalent) which converts it to hyperSpec object.

Author(s)

Vilmantas Gegzna

See Also

Other spHelper plots: check_palette(), layer_spRangeMean(), plot_colors(), plot_hyPalette(), plot_spCompare(), plot_spDiff(), plot_spDistribution(), qplot_confusion(), qplot_crosstab(), qplot_infoDim(), qplot_kAmp(), qplot_prediction(), qplot_spRangeCenter(), qplot_spRangeMedian(), qplot_spStat(), qplot_spc(), rmExpr(), rm_stripes(), stat_chull()

Other component analysis / factorisation related functions in spHelper: getScores(), infoDim(), plot_spDiff(), qplot_infoDim(), qplot_kAmp(), qplot_spc(), reconstructSp(), sortLoadings(), unipeak(), whichOutlier()

Examples

qplot_kSp(Loadings2)

data(flu, package = "hyperSpec")

qplot_kSpFacets(flu, Title = "Flu dataset")
qplot_kSpFacets(flu, Title = "Flu dataset", normalize = 1)
qplot_kSpFacets(flu, Title = "Flu dataset", normalize = FALSE)
qplot_kSpFacets(flu, Title = "Flu dataset", normalize = -1)

## Remove fill -----------------------------------------------------------------


qplot_kSp(flu, filled = FALSE)
qplot_sp(flu)

## Name of a legend ------------------------------------------------------------
flu$c2 <- as.factor(flu$c)

qplot_sp(flu, Title = "Flu dataset", by = 'c2', legendName = FALSE)
qplot_sp(flu, Title = "Flu dataset", by = 'c2', legendName = TRUE)
qplot_sp(flu, Title = "Flu dataset", by = 'c2', legendName = "Concentration")

## Example of line color transitions -------------------------------------------
  qplot_sp(laser)

GegznaV/spHelper documentation built on April 16, 2023, 1:42 p.m.