plot.glm.normalized.ratio.index: Plot function for (g)lm.nri and cor.test.nri

In hsdar: Manage, Analyse and Simulate Hyperspectral Data

Plot function for (g)lm.nri and cor.test.nri

Description

Plot values in (generalised) linear modes and correlation tests from narrow band indices

Usage

## S4 method for signature 'Nri'
plot(x, coefficient = "p.value", predictor = 2,                      
     xlab = "Wavelength band 1 (nm)",  
     ylab = "Wavelength band 2 (nm)", legend = TRUE,
     colspace = "hcl", col = c(10, 90, 60, 60, 10, 80),                
     digits = 2, range = "auto", constraint  = NULL,
     uppertriang = FALSE, zlog = FALSE, ...)

Arguments

x	Object to be plotted.
coefficient	Name or index of coefficient to plot.
predictor	Name or index of term to plot.
xlab	Label for x-axis.
ylab	Label for y-axis.
legend	Flag if legend is plotted. If legend == "outer" the legend is plotted in the outer margins of the figure. This is useful if both diagonals are used.
colspace	Either "hcl" or "rgb". Colour space to be used for the plots.
col	If colspace == "hcl", the vector is giving the minimum and maximum values of hue (element 1 & 2), chroma (element 3 & 4) and luminance (element 5 & 6). The optional element 7 is used as alpha value. See hcl for further explanation. If colspace == "rgb", a vector of length >=2 giving the colours to be interpolated using colorRamp.
digits	Precision of labels in legend.
range	"auto" or a vector of length = 2 giving the range of values to be plotted.
constraint	A character string giving a constraint which values should be plotted. See examples section.
uppertriang	Flag if upper triangle is used for the plot. Note that if TRUE the current plot is used instead of starting a new plot
zlog	Flag indicating if color should be logarithmically scaled. Useful e.g. for p-values.
...	Further arguments passed to plot.default.

Details

See details in glm.nri and glm.

Value

An invisible vector with minimum and maximum values plotted.

Author(s)

Lukas Lehnert

See Also

nri, glm.nri, glm, cor.test, t.test

Examples

## Not run: 
data(spectral_data)

## Calculate all possible combinations for WorldView-2-8
spec_WV <- spectralResampling(spectral_data, "WorldView2-8",
                              response_function = FALSE)
nri_WV <- nri(spec_WV, recursive = TRUE)

## Fit generalised linear models between NRI-values and chlorophyll
glmnri <- glm.nri(nri_WV ~ chlorophyll, preddata = spec_WV)

## Plot p-values
plot(glmnri, range = c(0, 0.05))
## Plot t-values
plot(glmnri, coefficient = "t.value")
## Plot only t-values where p-values < 0.001
plot(glmnri, coefficient = "t.value", 
     constraint = "p.value < 0.001")

## Fit linear models between NRI-values and chlorophyll
lmnri <- lm.nri(nri_WV ~ chlorophyll, preddata = spec_WV)

## Plot r.squared
plot(lmnri)

## Example for EnMAP (Attention: Calculation time may be long!)
spec_EM <- spectralResampling(spectral_data, "EnMAP", 
                              response_function = FALSE)
mask(spec_EM) <- c(300, 550, 800, 2500)
nri_EM <- nri(spec_EM, recursive = TRUE)
glmnri <- glm.nri(nri_EM ~ chlorophyll, preddata = spec_EM)

## Plot T values in lower and p-values in upper diagonal
## of the plot
## Enlarge margins for legends
par(mar = c(5.1, 4.1, 4.1, 5))
plot(glmnri, coefficient = "t.value", legend = "outer")
plot(glmnri, coefficient = "p.value", uppertriang = TRUE, zlog = TRUE)
lines(c(400,1705),c(400,1705))

## End(Not run)

hsdar documentation built on March 18, 2022, 6:35 p.m.