This function produces a reduced matrix of capture histories from a complete one by merging together some capture occasions.
It can also be used to change the format of a capture-recapture data set with complete capture histories: it transforms a data set with one row per captured unit to a data set with one row per capture history followed by its frequency.
periodhist(X, dfreq=FALSE, vt, drop=TRUE)
The matrix of the observed capture histories (see
for a description of the accepted formats).
A logical. By default FALSE, which means that
X has one row per unit. If TRUE, it
indicates that the matrix
X contains frequencies in its last column.
A vector containing the numbers of capture occasions for each pooled capture occasions. The length of this vector equals the number of capture occasions in the reduced matrix (noted \(I\)).
A logical, by default TRUE, meaning that in the output matrix the unobserved
capture histories (frequency of 0) are omitted. To keep them,
drop must be set to FALSE.
A \(2^I-1\) by \(I+1\) matrix of all the pooled capture histories,
with their observed frequencies. If
drop=TRUE (the default), the number of rows
of this matrix is in fact \(2^I-1\) minus the number of unobserved capture histories.
This function is useful when using an open population model to analyze a robust design data set. It can be used to reduce the data set to one observation per primary period. A one represents a unit caught at least once during the period and a zero a unit never caught.
It is also useful for experiments with a large number of capture occasions but a limited number of catches,
especially when there is no capture on some occasions. In such cases, one can pool together some capture occasions
If a data matrix produced by the
periodhist function is given as an argument to a Rcapture function,
dfreq argument must be set to
Baillargeon, S. and Rivest, L.P. (2007) Rcapture: Loglinear models for capture-recapture in R. Journal of Statistical Software, 19(5), http://www.jstatsoft.org/v19/i05.