Representing the CME solution
Core.Array
โ MethodArray(p::FspVectorSparse)
Construct a new dense N
-dimensional array from the FSP vector.
NumCME.FspVectorSparse
โ TypeFspVectorSparse{NS, IntT<:Integer, RealT<:AbstractFloat} <: AbstractFspVector
Sparse multi-indexed vector (i.e., sparse "tensor") with entries of type RealT
and multi-indices of type IntT
for storing CME solution and sensitivity vectors output by the Finite State Projection.
NumCME.FspVectorSparse
โ MethodFspVectorSparse(states::Vector{MVector{NS,IntT}}, values::Vector{RealT}; checksizes::Bool = true) where {NS,IntT<:Integer,RealT<:AbstractFloat}
Construct a sparse representation of a FSP-truncated probability distribution with support states
and proability values values
.
NumCME.FspVectorSparse
โ MethodFspVectorSparse(statespace::AbstractStateSpaceSparse{NS,NR,IntT}, statevalpairs::Vector{Pair{VecT,RealT}}) where {NS,NR,IntT<:Integer,VecT<:AbstractVector,RealT<:AbstractFloat}
Construct a sparse representation of a FSP-truncated probability distribution on the state space statespace
and proability values values
.
Base.sum
โ Methodsum(p::FspVectorSparse, dims)
Reduce the FSP vector p
by summing over species specified in dims
and return a new FSP vector over a reduced-dimension state space.