Generate a graphs showing correlation matrix (lower semi matrix) (Deprecated)

Generate graphs (lower semi matrix) showing lower semi matrix. These graphs are often used to show the structure of a correlation, similarity or dissimilarity matrix.

plotlowertri(input, valuename = "r",
pchlist = c(19, 17, 15, 1, 5, 2, 7), interval = 6,
cex = 1, ncex = 1, int =1.2, add.number = TRUE,
size = FALSE, add.text = FALSE, show.legend = TRUE,
digits = 2)

Arguments

input	The `input`, often a correlation or a distance matrix.
valuename	Name of the value that to show in the legend.
pchlist	A numberic vector specifying the shapes of points, see `pch par()`.
interval	Types of point shapes to show
cex	A number specifying the text size in the legend
ncex	Size of the text shown above each column.
int	Space between lines within the legend
add.number	If the column number should be shown
size	Whether the size of points should change with the value
add.text	Logical, If the number should be shown in the grid.
show.legend	Logical, If the legend should be appear.
digits	Number of digits for the label of each interval.

Details

In the legend, space between lines could be adjusted by specifying int.

Value

lower matrix plot

References

Zhang Qiaoying, Peng Shaolin, Zhang Sumei, Zhang Yunchun, Hou Yuping.(2008). Association of dormintant species in Guia hill Municipal Park of Macao. Ecology and Environment. 17:1541-1547

Author

Jinlong Zhang jinlongzhang01@gmail.com

Examples

data(testdata)
spmatrix <- data2mat(testdata)
result <- sp.pair(spmatrix)

## Check the legend for 0.00 to 0.33 (Unwanted label)
plotlowertri(result$Pearson, int = 0.5, cex=1.5)
#> Warning: Too few rows, please adjust the legend using "cex" and "int".
#>  Some of the points may be missing due to the distribution of data.
#>  You may also have to choose an appropriate number of intervals 
#>  using "interval" and "pchlist".
title("Pearson Correlation Lower Matrix Plot")

## Change the size of points and reset the intervals.
## Warning: The lower matrix plot illustrating Pearson
## Correlation between each pair of species. Note the
## triangle didn't appeared in the plots, but have been
## added to the legend. This is due to the distribution
## of data. Be careful in selection of intervals.


plotlowertri(result$Pearson, int = 0.5, cex=1.5,
interval = 4, pchlist = c(19, 17, 15, 1, 5), size = TRUE)
#> Warning: Too few rows, please adjust the legend using "cex" and "int".
#>  Some of the points may be missing due to the distribution of data.
#>  You may also have to choose an appropriate number of intervals 
#>  using "interval" and "pchlist".

title("Pearson Correlation Lower Matrix Plot")

## "Pure" dots, may have to add legend manually...
plotlowertri(result$Pearson, int = 0.5, cex=2.5,
interval = 4, pchlist = rep(19, 5), size = TRUE,
show.legend = FALSE)
#> Warning: Too few rows, please adjust the legend using "cex" and "int".
#>  Some of the points may be missing due to the distribution of data.
#>  You may also have to choose an appropriate number of intervals 
#>  using "interval" and "pchlist".

title("Pearson Correlation Lower Matrix Plot")


## Using BCI data
library(vegan)
#> Loading required package: permute
#> Loading required package: lattice
#> This is vegan 2.5-7
data(BCI)
## select the top 30 species according to relative frequency.
sub <- sub.sp.matrix(BCI, common = 30)
## Original
plotlowertri(cor(sub))

## Change size
plotlowertri(cor(sub), size = TRUE, cex = 3)

## Set the digits to 1
plotlowertri(cor(sub), size = TRUE, cex = 3, digits = 1,
ncex = 0.7)