ja is the jsfs, part a list of vectors specifying for each dimension how ja should be partitioned. If part_hi!=NULL, it is a list spefifying how ja is to be paritioned on the higher end of each dimension. if part or part_hi is not a list, it is turned into a list of the same length as dim(ja), in which each entry is the original part or part_hi e.g. 2,7,9 partitions into 1:2, 3:7, 8:9, 9:N For example, with part=c(1,3) and part_hi=c(1,3) we get the classical jaatha summary statistics. Note, however, that the order in which they appear will be different than in the original jaatha package.
coarsen_jsfs(ja, part, part_hi = NULL)vector of numbers, which are the sums over the blocks of the jsfs for all combinations of partitions
an array containing the joint site frequency spectrum
a vector of integers or a list of vectors of integers. If
it is a list, the vector part[[i]] specifies that the \(i\)-th dimension
of ja should be partitioned into 1:(part[[i]][1]-1),
part[[i]][1]:(part[[i]][2]-1), and so on. If part is a
vector, it will be used for all dimensions.
NULL or a vector of integers or a list of vector of integers
indicating the partioning at the higher end of each dimension. This means,
if it is a list, the values in the vector dim(ja)[i]-part_hi[[i]]
will be appended to the end of part[[i]]. If part_hi is a
single vector, it will be used for all dimensions. Thus, with the
combination of part=c(1,3) and part_hi=c(1,3), the classical jaatha summary
statistics, plus the two values ja[0] and
ja[length(ja)]. Note that the order in which they appear will
however be different than in the original jaatha summary statistics.
Dirk Metzler & Paul Staab
A. Tellier, P. Pfaffelhuber, B. Haubold, L. Naduvilezhath, L. E. Rose, T. Staedler, W. Stephan, and D. Metzler (2011) Estimating parameters of speciation models based on refined summaries of the joint site-frequency spectrum. PLoS One 6(5): e18155