| Title: | Empirical Distribution Ordering Inference Framework (EDOIF) |
|---|---|
| Description: | A non-parametric framework based on estimation statistics principle. Its main purpose is to infer orders of empirical distributions from different categories based on a probability of finding a value in one distribution that is greater than an expectation of another distribution. Given a set of ordered-pair of real-category values the framework is capable of 1) inferring orders of domination of categories and representing orders in the form of a graph; 2) estimating magnitude of difference between a pair of categories in forms of mean-difference confidence intervals; and 3) visualizing domination orders and magnitudes of difference of categories. The publication of this package is at Chainarong Amornbunchornvej, Navaporn Surasvadi, Anon Plangprasopchok, and Suttipong Thajchayapong (2020) <doi:10.1016/j.heliyon.2020.e05435>. |
| Authors: | Chainarong Amornbunchornvej [aut, cre]
|
| Maintainer: | Chainarong Amornbunchornvej <[email protected]> |
| License: | BSD_3_clause + file LICENSE |
| Version: | 0.1.3 |
| Built: | 2024-12-11 06:59:21 UTC |
| Source: | CRAN |
bootDiffmeanFunc is a support function for bootstrapping method. Its main task is to infer mean-difference confidence intervals of distributions for all categories except the first category in idx (idx[2],idx[3],...) minus a target category (idx[1]).
bootDiffmeanFunc(Group, Values, idx, reps, ci, methodType)bootDiffmeanFunc(Group, Values, idx, reps, ci, methodType)
Group |
is a vector of categories of each real number in Values |
Values |
is a vector of real-number values |
idx |
is an order list of categories; idx[1] is a target category while others (idx[2],idx[3],...) are compared against idx[1] in order to compute mean-difference confidence intervals. |
reps |
is a number of time of sampling with replacement in a bootstrapping method. |
ci |
is a level of confidence interval inferred. |
methodType |
is a type of method for inferring confidence intervals. It is a parameter of two.boot function of simpleboot package. |
This function returns a list of mean-difference confidence intervals of categories idx[2],idx[3],... minus category idx[1].
result a list of objects that contains mean-difference confidence intervals of pairs of distributions.
It contains mean-difference confidence intervals of categories idx[2],idx[3],... minus category idx[1].
checkSim3Res is a support function for checking whether an adjacency matrix of inferred
a dominant-distribution network adjMat is corrected w.r.t. generator SimNonNormalDist().
checkSim3Res(adjMat, flag = 0)checkSim3Res(adjMat, flag = 0)
adjMat |
is an adjacency matrix of inferred a dominant-distribution network. |
flag |
is a flag of matrix. It should be set only to shift the low of matrix for comparison. |
This function returns precision, recall, and F1-score of inferred adjacency matrix.
# Generate simulation data with 100 samples per categories simData<-SimNonNormalDist(nInv=100) # Performing ordering infernce from simData resultObj<-EDOIF(simData$Values,simData$Group) # Compare the inferred adjacency matrix with the ground truth checkSim3Res(adjMat=resultObj$adjMat)# Generate simulation data with 100 samples per categories simData<-SimNonNormalDist(nInv=100) # Performing ordering infernce from simData resultObj<-EDOIF(simData$Values,simData$Group) # Compare the inferred adjacency matrix with the ground truth checkSim3Res(adjMat=resultObj$adjMat)
EDOIF is a non-parametric framework based on Estimation Statistics principle. Its main purpose is to infer orders of empirical distributions from different categories base on a probability of finding a value in one distribution that greater than the expectation of another distribution.
Given a set of ordered-pair of real-category values the framework is capable of 1) inferring orders of domination of categories and representing orders in the form of a graph; 2) estimating magnitude of difference between a pair of categories in forms of confidence intervals; and 3) visualizing domination orders and magnitudes of difference of categories.
EDOIF(Values, Group, bootT, alpha, methodType)EDOIF(Values, Group, bootT, alpha, methodType)
Values |
is a vector of real-number values |
Group |
is a vector of categories of each real number in Values |
bootT |
is a number of times of sample with replacement for bootstrapping. The default is 1000. It must be above zero |
alpha |
is a significance level using in both confidence intervals and ordering inference it has the range [0,1]. The default is 0.05. |
methodType |
is an option for bootstrapping methods:either "perc" or "bca". The "perc" is the default option. |
This class constructor returns an object of EDOIF class.
obj an object of EDOIF class that contains the results of ordering inference
that can be print in text mode (print(obj)) or graphic mode (plot(obj)).
The obj consists of the following variables
Values, Group
|
The main inputs of the framework. They are the double and character vectors respectively. |
bootT, alpha, methodType
|
The number of bootstrapping, significance level, and bootstrapping method parameters. |
sortedGroupList |
A list of names of categories ascendingly ordered by their means. |
sortedmeanList |
A list of means of categories that are ascendingly ordered. |
MegDiffList[[i]] |
Mean difference confidence intervals and related information of all categories that have higher means than sortedGroupList[i] category. |
confInvsList[i, ]
|
A mean confidence interval of sortedGroupList[i] category. confInvsList[i,1] is a lower bound and confInvsList[i,2] is an upper bound. |
adjMat[i, j]
|
An element of adjacency matrix: one if sortedGroupList[j] category dominates sortedGroupList[i] using Mann-Whitney test, otherwise zero. |
pValMat[i, j]
|
A p-value of Mann-Whitney test for adjMat[i,j]. |
adjDiffMat[i, j]
|
A lower bound of confidence interval of mean difference for sortedGroupList[j] minus sortedGroupList[i] using methodType bootstrap. |
adjBootMat[i, j]
|
One if adjDiffMat[i,j] is positive, otherwise, zero. |
netDen |
A network density of dominant-distribution network derived from |
gObj |
An object of iGraph of a dominant-distribution network. |
Chainarong Amornbunchornvej, [email protected]
Run vignette("EDOIF_demo", package = "EDOIF") in a terminal to learn more details about how to use our package.
# Generate simulation data nInv<-100 initMean=10 stepMean=20 std=8 simData1<-c() simData1$Values<-rnorm(nInv,mean=initMean,sd=std) simData1$Group<-rep(c("C1"),times=nInv) simData1$Values<-c(simData1$Values,rnorm(nInv,mean=initMean,sd=std) ) simData1$Group<-c(simData1$Group,rep(c("C2"),times=nInv)) simData1$Values<-c(simData1$Values,rnorm(nInv,mean=initMean+2*stepMean,sd=std) ) simData1$Group<-c(simData1$Group,rep(c("C3"),times=nInv) ) simData1$Values<-c(simData1$Values,rnorm(nInv,mean=initMean+3*stepMean,sd=std) ) simData1$Group<-c(simData1$Group, rep(c("C4"),times=nInv) ) simData1$Values<-c(simData1$Values,rnorm(nInv,mean=initMean+4*stepMean,sd=std) ) simData1$Group<-c(simData1$Group, rep(c("C5"),times=nInv) ) # Performing ordering infernce from simData1 resultObj<-EDOIF(simData1$Values,simData1$Group) # Print results in text mode print(resultObj) # Plot results in graphic mode plot(resultObj)# Generate simulation data nInv<-100 initMean=10 stepMean=20 std=8 simData1<-c() simData1$Values<-rnorm(nInv,mean=initMean,sd=std) simData1$Group<-rep(c("C1"),times=nInv) simData1$Values<-c(simData1$Values,rnorm(nInv,mean=initMean,sd=std) ) simData1$Group<-c(simData1$Group,rep(c("C2"),times=nInv)) simData1$Values<-c(simData1$Values,rnorm(nInv,mean=initMean+2*stepMean,sd=std) ) simData1$Group<-c(simData1$Group,rep(c("C3"),times=nInv) ) simData1$Values<-c(simData1$Values,rnorm(nInv,mean=initMean+3*stepMean,sd=std) ) simData1$Group<-c(simData1$Group, rep(c("C4"),times=nInv) ) simData1$Values<-c(simData1$Values,rnorm(nInv,mean=initMean+4*stepMean,sd=std) ) simData1$Group<-c(simData1$Group, rep(c("C5"),times=nInv) ) # Performing ordering infernce from simData1 resultObj<-EDOIF(simData1$Values,simData1$Group) # Print results in text mode print(resultObj) # Plot results in graphic mode plot(resultObj)
getADJNetDen is a support function for calculating a network density of a dominant-distribution network.
getADJNetDen(adjMat)getADJNetDen(adjMat)
adjMat |
is an adjacency matrix of a dominant-distribution network. |
This function returns a value of network density of of a dominant-distribution network for a given adjMat.
# Generate simulation data with 100 samples per categories simData<-SimNonNormalDist(nInv=100) # Performing ordering infernce from simData resultObj<-EDOIF(simData$Values,simData$Group) # Get a network density of an adjacency matrix getADJNetDen(adjMat=resultObj$adjMat)# Generate simulation data with 100 samples per categories simData<-SimNonNormalDist(nInv=100) # Performing ordering infernce from simData resultObj<-EDOIF(simData$Values,simData$Group) # Get a network density of an adjacency matrix getADJNetDen(adjMat=resultObj$adjMat)
getConfInv is a support function for bootstrapping method. Its main purpose is to compute a mean confidence intervals of all distributions.
getConfInv(Values, Group, GroupList, bootT, alpha, methodType)getConfInv(Values, Group, GroupList, bootT, alpha, methodType)
Values |
is a vector of real-number values |
Group |
is a vector of categories of each real number in Values |
GroupList |
is a list of names of categories ascendingly ordered by their means. |
bootT |
is a number of times of sample with replacement for bootstrapping. The default is 1000. It must be above zero |
alpha |
is a significance level using in both confidence intervals and ordering inference it has the range [0,1]. The default is 0.05. |
methodType |
is an option for bootstrapping methods:either "perc" or "bca". The "perc" is the default option. |
This function returns a list of mean confidence intervals.
confInvsList[i, ]
|
The mean confidence interval of sortedGroupList[i] category. confInvsList[i,1] is a lower bound and confInvsList[i,2] is an upper bound. |
getDominantRADJ is a support function for inferring a dominant-distribution network using mean-difference confidence intervals.
getDominantRADJ(MegDiffList, methodType)getDominantRADJ(MegDiffList, methodType)
MegDiffList |
is a list of objects that contains mean-difference confidence intervals inferred by getMegDiffConfInv function. |
methodType |
is an option for bootstrapping methods:either "perc" or "bca". |
This function returns an adjacency matrix of a dominant-distribution network adjMat
and the corresponding lower-bound of mean difference CIs adjDiffMat.
adjDiffMat[i, j]
|
A lower bound of confidence interval of mean difference for j minus i using methodType bootstrap. |
adjMat[i, j]
|
An element of adjacency matrix: One if adjDiffMat[i,j] is positive, otherwise, zero. |
getiGraphNetDen is a support function for calculating a network density of a dominant-distribution network.
getiGraphNetDen(g)getiGraphNetDen(g)
g |
is an object of iGraph class of a dominant-distribution network. |
This function returns a value of network density of of a dominant-distribution network for a given object g.
# Generate simulation data with 100 samples per categories simData<-SimNonNormalDist(nInv=100) # Performing ordering infernce from simData resultObj<-EDOIF(simData$Values,simData$Group) # Get a network density of an iGraph object getiGraphNetDen(g=resultObj$gObj)# Generate simulation data with 100 samples per categories simData<-SimNonNormalDist(nInv=100) # Performing ordering infernce from simData resultObj<-EDOIF(simData$Values,simData$Group) # Get a network density of an iGraph object getiGraphNetDen(g=resultObj$gObj)
getiGraphOBJ is a support function for converting a dominant-distribution network adjacency matrix to an iGraph object.
getiGraphOBJ(adjMat, sortedGroupList)getiGraphOBJ(adjMat, sortedGroupList)
adjMat |
is an adjacency matrix of a dominant-distribution network. |
sortedGroupList |
is a list of names of categories ascendingly ordered by their means. |
This function returns an iGraph object of a dominant-distribution network for a given adjMat.
# Generate simulation data with 100 samples per categories simData<-SimNonNormalDist(nInv=100) # Performing ordering infernce from simData resultObj<-EDOIF(simData$Values,simData$Group) # Get an iGraph object from an adjacency matrix igraphObj<-getiGraphOBJ(adjMat=resultObj$adjMat,sortedGroupList=resultObj$sortedGroupList)# Generate simulation data with 100 samples per categories simData<-SimNonNormalDist(nInv=100) # Performing ordering infernce from simData resultObj<-EDOIF(simData$Values,simData$Group) # Get an iGraph object from an adjacency matrix igraphObj<-getiGraphOBJ(adjMat=resultObj$adjMat,sortedGroupList=resultObj$sortedGroupList)
getMegDiffConfInv is a support function for bootstrapping method. Its main purpose is to compute a mean-difference confidence intervals between all pair of distributions.
getMegDiffConfInv(Values, Group, GroupList, bootT, alpha, methodType)getMegDiffConfInv(Values, Group, GroupList, bootT, alpha, methodType)
Values |
is a vector of real-number values |
Group |
is a vector of categories of each real number in Values |
GroupList |
is a list of names of categories ascendingly ordered by their means. |
bootT |
is a number of times of sample with replacement for bootstrapping. The default is 1000. It must be above zero |
alpha |
is a significance level using in both confidence intervals and ordering inference it has the range [0,1]. The default is 0.05. |
methodType |
is an option for bootstrapping methods:either "perc" or "bca". The "perc" is the default option. |
This function returns a list of mean-difference confidence intervals.
MegDiffList a list of objects that contains mean-difference confidence intervals of all possible pairs of distributions.
It contains MegDiffList[[1]],...,MegDiffList[[length(GroupList)]].
The MegDiffList consists of the following variables
MegDiffList[[i]] |
Mean-difference confidence intervals and related information of all categories that have higher means than sortedGroupList[i] category. |
getOrder is a support function for inferring a linear order of categories ascendingly sorted by their means.
getOrder(Values, Group)getOrder(Values, Group)
Values |
is a vector of real-number values |
Group |
is a vector of categories of each real number in Values |
This function returns two lists: an order list of categories sortedGroupList
and its correspoding list of means sortedmeanList.
sortedGroupList |
The list of names of categories ascendingly ordered by their means. |
sortedmeanList |
The list of means of categories that are ascendingly ordered. |
# Generate simulation data simData<-SimNonNormalDist(nInv=100,noisePer=0.1) # Call the function to get the sorted lists getOrder(Values=simData$Values,Group=simData$Group)# Generate simulation data simData<-SimNonNormalDist(nInv=100,noisePer=0.1) # Call the function to get the sorted lists getOrder(Values=simData$Values,Group=simData$Group)
getttestDominantRADJ is a support function for inferring a dominant-distribution network using Student's t-test.
getttestDominantRADJ(Values, Group, GroupList, alpha)getttestDominantRADJ(Values, Group, GroupList, alpha)
Values |
is a vector of real-number values |
Group |
is a vector of categories of each real number in Values |
GroupList |
is a list of names of categories ascendingly ordered by their means. |
alpha |
is a significance level using in both confidence intervals and ordering inference it has the range [0,1]. |
This function returns an adjacency matrix of a dominant-distribution network adjMat
and the corresponding p-values of all category pairs.
adjMat[i, j]
|
An element of adjacency matrix: one if GroupList[j] category dominates GroupList[i] using Student's t-test, otherwise zero. |
pValMat[i, j]
|
A p-value of Student's t-test for adjMat[i,j]. |
getWilcoxDominantRADJ is a support function for inferring a dominant-distribution network using Mann-Whitney (Wilcoxon) Test.
getWilcoxDominantRADJ(Values, Group, GroupList, alpha)getWilcoxDominantRADJ(Values, Group, GroupList, alpha)
Values |
is a vector of real-number values |
Group |
is a vector of categories of each real number in Values |
GroupList |
is a list of names of categories ascendingly ordered by their means. |
alpha |
is a significance level using in both confidence intervals and ordering inference it has the range [0,1]. |
This function returns an adjacency matrix of a dominant-distribution network adjMat.
and the corresponding p-values of all category pairs.
adjMat[i, j]
|
An element of adjacency matrix: one if GroupList[j] category dominates GroupList[i] using Mann-Whitney test, otherwise zero. |
pValMat[i, j]
|
A p-value of Mann-Whitney test for adjMat[i,j]. |
meanBoot is a support function for bootstrapping method.
Its main purpose is to compute a mean of a given samples from data selected by indices.
meanBoot(data, indices)meanBoot(data, indices)
data |
is a vector of real-number values |
indices |
is a vector of TRUE/FALSE indices. It allows boot to select samples. |
This function returns a mean of values in data that have values TRUE within indices.
plot.EDOIF is a support function for printing all plots of EDOIF framework: dominant-distribution network plot, mean CI plot, and mean-difference CI plot.
## S3 method for class 'EDOIF' plot(x, ..., NList, options, fontSize)## S3 method for class 'EDOIF' plot(x, ..., NList, options, fontSize)
x |
is an object of EDOIF class that contains the results of ordering inference. |
... |
Signature for S3 generic function. |
NList |
is a list of based categories users want to have in mean-difference CI plot. |
options |
is an option of reporting EDOIF plot(s): 0 for reporting all plots, 1 for mean-difference CI plot, 2 for mean CI plot, and 3 for dominant-distribution network plot. |
fontSize |
is a font size of text for all plots. |
# Generate simulation data with 100 samples per categories simData<-SimNonNormalDist(nInv=100) # Performing ordering infernce from simData resultObj<-EDOIF(simData$Values,simData$Group) # Plot results in graphic mode plot(resultObj)# Generate simulation data with 100 samples per categories simData<-SimNonNormalDist(nInv=100) # Performing ordering infernce from simData resultObj<-EDOIF(simData$Values,simData$Group) # Plot results in graphic mode plot(resultObj)
plotGraph is a support function for plotting a dominant-distribution network from an adjacency matrix.
plotGraph(obj, rankFlag = TRUE)plotGraph(obj, rankFlag = TRUE)
obj |
is an object of EDOIF class that contains the results of ordering inference. |
rankFlag |
is an option for including ranks of categories with in the plot: default is TRUE for including ranks. |
This function returns a list of an object of iGraph for a dominant-distribution network and its plot variable.
graphVar |
An object of iGraph for a dominant-distribution network |
# Generate simulation data with 100 samples per categories simData<-SimNonNormalDist(nInv=100) # Performing ordering infernce from simData resultObj<-EDOIF(simData$Values,simData$Group) # Plot a dominant-distribution network and return a list of an iGraph object iGraphList<-plotGraph(obj=resultObj)# Generate simulation data with 100 samples per categories simData<-SimNonNormalDist(nInv=100) # Performing ordering infernce from simData resultObj<-EDOIF(simData$Values,simData$Group) # Plot a dominant-distribution network and return a list of an iGraph object iGraphList<-plotGraph(obj=resultObj)
plotMeanCIs is a support function for plotting mean confidence intervals.
plotMeanCIs(obj, fontSize = 15, rankFlag = TRUE)plotMeanCIs(obj, fontSize = 15, rankFlag = TRUE)
obj |
is an object of EDOIF class that contains the results of ordering inference. |
fontSize |
is a font size of text for all plots. |
rankFlag |
is an option for including ranks of categories with in the plot: default is TRUE for including ranks. |
This function returns a list of an object of ggplot class.
pMeanCI |
An object of ggplot class containing the plot of mean confidence intervals |
# Generate simulation data with 100 samples per categories simData<-SimNonNormalDist(nInv=100) # Performing ordering infernce from simData resultObj<-EDOIF(simData$Values,simData$Group) # Get a list of ggplot object of mean confidence intervals ggplotList<-plotMeanCIs(obj=resultObj) # Plot mean confidence intervals plot(ggplotList$pMeanCI)# Generate simulation data with 100 samples per categories simData<-SimNonNormalDist(nInv=100) # Performing ordering infernce from simData resultObj<-EDOIF(simData$Values,simData$Group) # Get a list of ggplot object of mean confidence intervals ggplotList<-plotMeanCIs(obj=resultObj) # Plot mean confidence intervals plot(ggplotList$pMeanCI)
plotMeanDiffCIs is a support function for plotting difference-mean confidence intervals.
plotMeanDiffCIs(obj, NList, fontSize = 15, rankFlag = TRUE)plotMeanDiffCIs(obj, NList, fontSize = 15, rankFlag = TRUE)
obj |
is an object of EDOIF class that contains the results of ordering inference. |
NList |
is a list of based categories users want to have in mean-difference CI plot. |
fontSize |
is a font size of text for all plots. |
rankFlag |
is an option for including ranks of categories with in the plot: default is TRUE for including ranks. |
This function returns a list of an object of ggplot class.
pDiffCI |
An object of ggplot class containing the plot of mean-difference confidence intervals |
# Generate simulation data with 100 samples per categories simData<-SimNonNormalDist(nInv=100) # Performing ordering infernce from simData resultObj<-EDOIF(simData$Values,simData$Group) # Get a list of ggplot object of mean-difference confidence intervals ggplotList<-plotMeanDiffCIs(obj=resultObj) # Plot mean-difference confidence intervals plot(ggplotList$pDiffCI)# Generate simulation data with 100 samples per categories simData<-SimNonNormalDist(nInv=100) # Performing ordering infernce from simData resultObj<-EDOIF(simData$Values,simData$Group) # Get a list of ggplot object of mean-difference confidence intervals ggplotList<-plotMeanDiffCIs(obj=resultObj) # Plot mean-difference confidence intervals plot(ggplotList$pDiffCI)
print.EDOIF is a support function for printing results of ordering inference in text.
## S3 method for class 'EDOIF' print(x, ...)## S3 method for class 'EDOIF' print(x, ...)
x |
is an object of EDOIF class that contains the results of ordering inference. |
... |
Signature for S3 generic function. |
# Generate simulation data with 100 samples per categories simData<-SimNonNormalDist(nInv=100) # Performing ordering infernce from simData resultObj<-EDOIF(simData$Values,simData$Group) # Print results in text mode print(resultObj)# Generate simulation data with 100 samples per categories simData<-SimNonNormalDist(nInv=100) # Performing ordering infernce from simData resultObj<-EDOIF(simData$Values,simData$Group) # Print results in text mode print(resultObj)
SimMixDist is a support function for generating samples from mixture distribution. The main purpose of this function is to generate samples from non-normal distribution.
SimMixDist(nInv, mean, std, p1, p2)SimMixDist(nInv, mean, std, p1, p2)
nInv |
is a number of samples the function will generate. |
mean |
is a mean of a normal distribution part of mixture distribution. |
std |
is a standard deviation of a normal distribution part of mixture distribution. |
p1 |
is a ratio of a normal distribution within a mixture distribution. |
p2 |
is a ratio of a Cauchy distribution within a mixture distribution. |
This function returns a list of samples V generated by a mixture distribution.
# Generate simulation data with 100 samples with a mixture distribution # The distribution consist of the following distributions: # 1) 10% of uniform distribution range [-400,400]; # 2) 50% of normal distribution with mean = 40 and std =8; and # 3) 40% of Cauchy distribution with location= 45 and scale = 2. V<-SimMixDist(nInv=100,mean=40,std=8,p1=0.1,p2=0.5)# Generate simulation data with 100 samples with a mixture distribution # The distribution consist of the following distributions: # 1) 10% of uniform distribution range [-400,400]; # 2) 50% of normal distribution with mean = 40 and std =8; and # 3) 40% of Cauchy distribution with location= 45 and scale = 2. V<-SimMixDist(nInv=100,mean=40,std=8,p1=0.1,p2=0.5)
SimNonNormalDist is a support function for generating samples from mixture distribution.
There are five categories. Each categories has nInv samples.
Categories C1,C2,C3, and C4 are dominated by C5 but none of them dominate each other.
SimNonNormalDist(nInv, noisePer)SimNonNormalDist(nInv, noisePer)
nInv |
is a number of samples the function will generate for each category. |
noisePer |
is ratio of uniform distribution within a mixture distribution. It is considered as a uniform noise that make an approach to hardly distinguish whether one distribution dominates another. |
The main purpose of this function is to generate samples that contains domination relation among categories.
This function returns a list of samples Values and their category Group generated by a mixture distribution.
Values |
A vector of samples generated by a mixture distribution. |
Group |
A list of categories associated with |
V1, ..., V5
|
Lists of sample vectors separated by categories. |
# Generate simulation data with 100 samples per categories with 10% of uniform noise simData<-SimNonNormalDist(nInv=100,noisePer=0.1)# Generate simulation data with 100 samples per categories with 10% of uniform noise simData<-SimNonNormalDist(nInv=100,noisePer=0.1)