SuperRanker (version 1.2.1)

sra: Compute the sequential rank agreement

Description

Compute the sequential rank agreement

Usage

sra(object, B, na.strings, nitems, type, epsilon = 0, ...)
# S3 method for default
sra(object, B, na.strings, nitems, type, epsilon = 0, ...)
# S3 method for matrix
sra(
  object,
  B = 1,
  na.strings = NULL,
  nitems = nrow(object),
  type = c("sd", "mad"),
  epsilon = 0,
  ...
)
# S3 method for list
sra(
  object,
  B = 1,
  na.strings = NULL,
  nitems = max(sapply(object, length)),
  type = c("sd", "mad"),
  epsilon = 0,
  ...
)

Value

A vector of the sequential rank agreement

Arguments

object

Either matrix where each column is a ranked list of items or a list of ranked lists of items. Elements are integers between 1 and the length of the lists. The lists should have the same length but censoring can be used by setting the list to zero from a point onwards. See details for more information.

B

An integer giving the number of randomization to sample over in the case of censored observations

na.strings

A vector of strings/values that represent missing values in addition to NA. Defaults to NULL which means only NA are censored values.

nitems

The total number of items in the original lists if we only have partial lists available.

type

The type of measure to use. Either sd (standard deviation - the default) or mad (median absolute deviance around the median)

epsilon

A non-negative numeric vector that contains the minimum limit in proportion of lists that must show the item. Defaults to 0. If a single number is provided then the value will be recycles to the number of items.

...

Arguments passed to methods.

Author

Claus Ekstrøm <[e-mail]> and Thomas A Gerds <[e-mail]>

Examples

mlist <- matrix(cbind(1:8,c(1,2,3,5,6,7,4,8),c(1,5,3,4,2,8,7,6)),ncol=3)
sra(mlist)

mlist <- matrix(cbind(1:8,c(1,2,3,5,6,7,4,8),c(1,5,3,4,2,8,7,6)),ncol=3)
sra(mlist, nitems=20, B=10)

alist <- list(a=1:8,b=sample(1:8),c=sample(1:8))
sra(alist)

blist <- list(x1=letters,x2=sample(letters),x3=sample(letters))
sra(blist)

## censored lists are either too short
clist <- list(x1=c("a","b","c","d","e","f","g","h"),
              x2=c("h","c","f","g","b"),
              x3=c("d","e","a"))
set.seed(17)
sra(clist,na.strings="z",B=10)

## or use a special code for missing elements
Clist <- list(x1=c("a","b","c","d","e","f","g","h"),
              x2=c("h","c","f","g","b","z","z","z"),
              x3=c("d","e","a","z","z","z","z","z"))
set.seed(17)
sra(Clist,na.strings="z",B=10)