The large number of different filter options (and corresponding getter and setter accessors) makes it necessary to maintain them in a special class. This differs from other packages like GOstats, where arguments for controlling the program and the results are stored in the same object.
Objects from the Class
Objects can be created by calls of the form new("concubfilter", ...).
Slots
For a more detailed description of some slots see below.
nfact:
a numeric giving the number of factors
names:
a character vector giving the name of each factor
p.value:
a numeric giving the P-value to be considered. Defaults to 0.1.
test.direction:
a character giving the direction of association. Defaults to “two.sided”.
minimum.l2or:
a numeric giving the minimum absolute log2 odds ratio to be considered. Defaults to 0.
skip.min.group:
a numeric giving the minimum number of tag a group is allowed to have. Defaults to 2.
skip.min.obs:
a numeric giving the minimum number at the position of interest allowed. Defaults to 2.
skip.zeroobs:
a logical. Defaults to TRUE.
drop.insignif.layer:
A vector of logicals. By default, all positions are set to FALSE.
drop.wrongdir.layer:
A vector of logicals. By default, all positions are set to FALSE.
drop.lowl2or.layer:
A vector of logicals. By default, all positions are set to FALSE.
Methods and slot accessors
Several methods are implemented for class concubfilter. They can be roughly grouped into informative, basic, skip-test, and data-reduction methods.
Individual options can be accessed by the corresponding getter and setter methods, for instance
skip.zeroobs
signature(x = "concubfilter"): get current setting to skip test in case of zero cell
skip.zeroobs<-
signature(x = "concubfilter"): set a new value to skip test in case of zero cell
Informative methods
Currently only a single method is implemented.
show
signature(object = "concubfilter"): a short summary about current filter settings
Basic filters/threshoulds
p.value
The maximum P-value that should be taken into account
minimum.l2or
The minumum absolute of log2 odds ratio that should be taken into account
test.direction
The direction of the association. Can be “two.sided”, “greater” (test for over-representation), or “less” (test for under-representation)
Skip test
The following filters cause a skip of a test, i.e. the test is never run if at least one of the conditions is fullfilled.
skip.zeroobs
skip the test, if the position of interest (x_{00} or x_{000}) is zero, i.e. no tag from the population matches the conditions defined at the marginals.
skip.mingroup
skip the test, if the groups considered at the position of interest are too small at all.
skip.minobs
skip the test, if the position of interest contains less than ‘this value’ entries.
Data reduction
The following filters reduce the amout of outcome of the tests. They are applied to both, the (2 or 3 dimensional) table containing the odds ratios and the table containing the corresponding P-values.
drop.insignif.layer
drop all layers in the tables where all P-values are greater than the value defined in ‘p.value’.
drop.wrongdir.layer
drop all layers in the tables where all odds ratios are showing into the opposite direction as defined in ‘test.direction’.
drop.lowl2or.layer
drop all layers in the tables where all absolutes of the log2 odds ratios are smaller than ‘min.l2or’.
Examples
showClass("concub")
Results
R version 3.3.1 (2016-06-21) -- "Bug in Your Hair"
Copyright (C) 2016 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu (64-bit)
R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.
R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.
Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.
> library(geecc)
geecc 1.6.0 loaded
> png(filename="/home/ddbj/snapshot/RGM3/R_BC/result/geecc/concubfilter-class.Rd_%03d_medium.png", width=480, height=480)
> ### Name: concubfilter-class
> ### Title: Class '"concubfilter"'
> ### Aliases: concubfilter-class show,concubfilter-method
> ### initialize,concubfilter-method p.value p.value<-
> ### p.value,concubfilter-method p.value<-,concubfilter-method
> ### minimum.l2or minimum.l2or<- minimum.l2or,concubfilter-method
> ### minimum.l2or<-,concubfilter-method test.direction test.direction<-
> ### test.direction,concubfilter-method
> ### test.direction<-,concubfilter-method skip.min.group skip.min.group<-
> ### skip.min.group,concubfilter-method
> ### skip.min.group<-,concubfilter-method skip.min.obs skip.min.obs<-
> ### skip.min.obs,concubfilter-method skip.min.obs<-,concubfilter-method
> ### skip.zeroobs skip.zeroobs<- skip.zeroobs,concubfilter-method
> ### skip.zeroobs<-,concubfilter-method drop.insignif.layer
> ### drop.insignif.layer<- drop.insignif.layer,concubfilter-method
> ### drop.insignif.layer<-,concubfilter-method drop.lowl2or.layer
> ### drop.lowl2or.layer<- drop.lowl2or.layer,concubfilter-method
> ### drop.lowl2or.layer<-,concubfilter-method drop.wrongdir.layer
> ### drop.wrongdir.layer<- drop.wrongdir.layer,concubfilter-method
> ### drop.wrongdir.layer<-,concubfilter-method
> ### Keywords: classes
>
> ### ** Examples
>
> showClass("concub")
Class "concub" [package "geecc"]
Slots:
Name: fact population
Class: list character
Name: keep.empty.vars options
Class: list list
Name: approx null.model
Class: numeric formula
Name: test.result test.result.filter
Class: list list
Name: test.result.filter.heatmap
Class: list
>
>
>
>
>
> dev.off()
null device
1
>