peers {Benchmarking} | R Documentation |
Find peer firms and units
Description
The function peers
finds for each firm its peers,
get.number.peers
finds for each peer the number of times
this peer apears as a peer, and get.which.peers
determines
for one or more peers the firms they appear as peers for. Also
include a function get.peers.lambda
to calculate for firms the
importance (lambdas) of peers.
Usage
peers(object, NAMES = FALSE, N=1:dim(object$lambda)[1], LAMBDA=0)
get.number.peers(object, NAMES = FALSE, N=1:dim(object$lambda)[2], LAMBDA=0)
get.which.peers(object, N = 1:dim(object$lambda)[2], LAMBDA=0)
get.peers.lambda(object, N=1:dim(object$lambda)[1], LAMBDA=0)
Arguments
object |
An object of class Farrell as returned by the functions
|
NAMES |
If true then names for the peers are returned if names
are available otherwise the unit index numbers are used. If |
N |
The firm(s) or peer(s) for which to get the results. |
LAMBDA |
Minimum weight for extracted peers, i.e. the extracted peers have
lambda values larger than |
Details
The returned values are index of the firms and can be used by itself, but can also by used as an index for a variable with names of the firms.
The peers
returns a matrix with numbers for the peers for
each firm; for firms with efficiency 1 the peers are just the firm
itself. If there is slack in the evaluation of a firm with
efficiency 1, this can be found with a call to slack
,
either directly or by the argument SLACK
when a function
dea
was called to generate the Farrell object
.
The get.number.peers
returns the number of firms that a peer
serves as a peer for.
The get.peers.lambda
returns a list of firms with the peers
and corresponding value of lambda.
Value
The return values are firm numbers. If the argument
NAMES=TRUE
is used in the function peers
the return
is a list of names of the peers if names for the firms are available
as row names.
Note
Peers are defined as firms where the corresponding lambdas are positive.
Note that peers might change between a Farrell object return from
dea
with SLACK=FALSE
and a call with SLACK=TRUE
or a following call to the function slack
because a peer on
the frontier with slack might by the call to dea
be a
peer for itself whereas this will not happen when slacks are
calculated.
Author(s)
Peter Bogetoft and Lars Otto larsot23@gmail.com
References
Peter Bogetoft and Lars Otto; Benchmarking with DEA, SFA, and R; Springer 2011. Sect. 4.6 page 93
See Also
Examples
x <- matrix(c(100,200,300,500,100,200,600),ncol=1)
y <- matrix(c(75,100,300,400,25,50,400),ncol=1)
e <- dea(x,y)
peers(e)
get.number.peers(e)
# Who are the firms that firm 1 and 4 is peers for
get.which.peers(e, c(1,4))