plot.PCoA.cumulative

Source: R/plot.PCoA.cumulative.R

Function to plot the cumulative variance of all the principal coordinates of a PCoA.

# S3 method for class 'PCoA.cumulative'
plot(
  DEprot.PCoA.object,
  bar.color = "steelblue",
  line.color = "navyblue",
  title = NULL,
  broken.stick = TRUE
)

Arguments

DEprot.PCoA.object

An object of class DEprot.PCoA, as generated by perform.PCoA.

bar.color

String indicating the color to use for the bar fill. Default: "steelblue".

line.color

String indicating the color to use for the line and the dots of the cumulative curve. Default: "navyblue".

title

String indicating the title of the plot (markdown annotation supported). Default: NULL.

broken.stick

Logical value indicating whether the broken-stick expectation should be displayed as a dashed gray line. Default: TRUE.

Value

A barplot in ggplot format.

Details

Only the principal coordinates corresponding to a positive eigenvalue are displayed, and the proportions of variance are computed using the positive eigenvalues as denominator (see perform.PCoA).
The broken-stick expectation provides a criterion for the axis retention: the axes whose proportion of variance exceeds the corresponding broken-stick value are usually considered non-trivial.

Author

Sebastian Gregoricchio

Examples

# Compute the sample correlations and the PCoA
correlation <- plot.correlation.heatmap(DEprot.object = DEprot::test.toolbox$dpo.imp,
                                        correlation.method = "spearman")

pcoa <- perform.PCoA(DEprot.correlation.object = correlation)
#> Only 7 of the 11 principal coordinates requested correspond to a positive eigenvalue and could be computed.
#> The dissimilarities are not euclidean-embeddable: 4 negative eigenvalue(s) detected (largest relative magnitude = 0.083).
#>        The negative eigenvalues are non-negligible. Consider using 'distance.transformation = "sqrt"' or 'correction = "cailliez"'/'correction = "lingoes"'.
#>        The proportions of variance are computed using only the positive eigenvalues as denominator.

# Plot cumulative variance barplot
plot.PCoA.cumulative(pcoa)