plotTest: Detailed plots of Interval-Wise Testing result

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

View source: R/plotTest.R

Description

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).

Usage

1
2
3
4
5
6
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, ...)

Arguments

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.

Value

No value returned. The function produces a graphical output.

Author(s)

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

References

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

See Also

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.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
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))

IWTomics documentation built on Nov. 1, 2018, 2:21 a.m.