plotTest: Detailed plots of Interval-Wise Testing result
In marziacremona/IWTomics: Interval-Wise Testing for Omics Data

Description Usage Arguments Value Author(s) References See Also Examples

View source: R/plotTest.R

The function plotTest creates graphical representations of the Interval-Wise Testing results present in the test slot of a "IWTomicsData" object. In particular, it draws a heatmap of the adjusted p-value functions at each possible scale (with the different maximum interval lengths considered), a plot of the adjusted p-value curve at the chosen scale threshold and a plot of the feature measurements in the region datasets tested (aligned curves or pointwise quantile curves).

plotTest(regionsFeatures, alpha=0.05, scale_threshold=NULL, nlevel=100, type="boxplot", 
  N_regions=pmin(lengthRegions(regionsFeatures),10), 
  probs=c(0.25,0.5,0.75), average=TRUE, size=TRUE, 
  id_features_subset=idFeatures(regionsFeatures), 
  col=1+seq_len(nRegions(regionsFeatures)), 
  ask=TRUE, xlab="Windows", ylim=NULL, ...)

`regionsFeatures`	`"IWTomicsData"` object with `test` slot.
`alpha`	level of the hypothesis test used to select and display significant results. Default `alpha=0.05`.
`scale_threshold`	threshold on the test scale (maximum interval length for the p-value adjustment) to be used in the adjusted p-value plot. Can be either a scalar (the same length for all features) or a vector (a length for each feature) or a list of vectors (a vector for each test). If `NULL`, the maximum possible interval length is used.
`nlevel`	number of desired color levels for the adjusted p-value heatmap.
`type`	type of plot to be drawn in addition to the adjusted p-value heatmap and plot. Possible types are: `"boxplot"` for the pointwise quantiles curves (default), `"curves"` for the (aligned) curves.
`N_regions`	number of regions to be randomly chosen (for each region dataset) in `"curves"` type plot. Default plots a maximum of 10 curves.
`probs`	probabilities corresponding to the quantiles to be drawn in `"boxplot"` type plot. Default plots quartile curves.
`average`	if `TRUE`, plot the mean curves.
`size`	if `TRUE`, plot the sample size in each position.
`id_features_subset`	vector with the identifiers of the features to be plotted.
`col`	vector of plotting colors for the different region datasets.
`ask`	if `TRUE` (default) the user is prompted before a new plot is drawn.
`xlab`	a title for the x axis.
`ylim`	the y limits of the feature measurement plot.
`...`	additional plot parameters.

No value returned. The function produces a graphical output.

Marzia A Cremona, Alessia Pini, Francesca Chiaromonte, Simone Vantini

A Pini and S Vantini (2017). Interval-Wise Testing for functional data. Journal of Nonparametric Statistics.

IWTomicsData for "IWTomicsData" class, constructors, accessors and methods; IWTomicsTest function to perform the Interval-wise Testing; plotSummary to draw a summary plot of the test results.

data(regionsFeatures_center)

## One sample test for 'control' regions and feature 'ftr1'
regionsFeatures_test=IWTomicsTest(regionsFeatures_center,
                                  id_region1='control',id_features_subset='ftr1')

## Plotting the results of the one sample test
plotTest(regionsFeatures_test,col=5)

## Set the scale threshold to 25
plotTest(regionsFeatures_test,col=5,scale_threshold=25)

## Plot curves instead of boxplots, do not report sample size in each position
plotTest(regionsFeatures_test,col=5,type='curves',size=FALSE)


## Two sample test for 'elem1', 'elem2' and 'elem3' vs 'control' regions
regionsFeatures_test=IWTomicsTest(regionsFeatures_center,
                                  id_region1=c('elem1','elem2','elem3'),
                                  id_region2=c('control','control','control'))

## Plotting the results of the two sample test
plotTest(regionsFeatures_test)

## Set scale thresholds for the different features
plotTest(regionsFeatures_test,scale_threshold=c(25,30))

## Plot only results regarding feature 'ftr2', setting scale thresholds for each test
plotTest(regionsFeatures_test,id_features_subset='ftr2',
         scale_threshold=list(test1=10,test2=20,test3=30))

## Report also 0.05 and 0.95 quantiles in the boxplots
plotTest(regionsFeatures_test,id_features_subset='ftr2',
         scale_threshold=list(test1=10,test2=20,test3=30),
         probs=c(0.05,0.25,0.5,0.75,0.9))