a b/man/runKappa.Rd
1
% Generated by roxygen2: do not edit by hand
2
% Please edit documentation in R/runKappa.R
3
\name{runKappa}
4
\alias{runKappa}
5
\title{Run consistency evaluation using Kappa statistic}
6
\usage{
7
runKappa(
8
  subt1 = NULL,
9
  subt2 = NULL,
10
  subt1.lab = NULL,
11
  subt2.lab = NULL,
12
  fig.path = getwd(),
13
  fig.name = "constheatmap",
14
  width = 5,
15
  height = 5
16
)
17
}
18
\arguments{
19
\item{subt1}{A numeric vector to indicate the first appraisement. Order should be exactly the same with subt2 for each sample.}
20
21
\item{subt2}{A numeric vector to indicate the second appraisement. Order should be exactly the same with subt1 for each sample.}
22
23
\item{subt1.lab}{A string value to indicate the label of the first subtype.}
24
25
\item{subt2.lab}{A string value to indicate the label of the second subtype.}
26
27
\item{fig.path}{A string value to indicate the output path for storing the consistency heatmap.}
28
29
\item{fig.name}{A string value to indicate the name of the consistency heatmap.}
30
31
\item{width}{A numeric value to indicate the width of output figure.}
32
33
\item{height}{A numeric value to indicate the height of output figure.}
34
}
35
\value{
36
A figure of consistency heatmap (.pdf).
37
}
38
\description{
39
Calculate Kappa statistic to measure the consistency between two appraisements
40
}
41
\details{
42
This function evaluates the consistency between two appraisements that targets to the same cohort.
43
For example, the NTP-predicted subtype amd PAM-predicted subtype of external cohort, or the current subtype and predicted subtype of discovery cohort.
44
Therefore, the arguments `subt1` and `subt2` can be the `clust` column of `clust.res` derived from `getMOIC()` with one specified algorithm
45
or `get\%algorithm_name\%` or `getConsensusMOIC()` with a list of multiple algorithms or `runNTP()` or `runPAM()`.
46
However, subtypes identified from different algorithm (i.e., `get\%algorithm_name1\%` and `get\%algorithm_name2\%`) can not be evaluated
47
because the subtype 1 identified from the first algorithm may not be the same subtype 1 from the second algorithm.
48
}
49
\examples{
50
# There is no example and please refer to vignette.
51
}