Here we discuss several three primary data structures in SEEPS.
Simulators in SEEPS outputs a list with the following fields.
parents - A K by 2 matrix,
where each row denotes an infection/birth event. The first column is the
index of the parent, the second column is the time of the
birth/infection event. This vector is pre-allocated to avoid repeated
array resizing. Values after the last nonzero row should be disregarded
in any downstream analysis.active - A vector of M < K indicies
corresponding to active infections. These indices correspond to
individuals that have not been removed from the population.t_end - The time index at the end of the simulation. A
non-negative integer.total_offspring - The number of newly generated
offspring added in the most previous update step.Phylogenies and genealogies are represented in an array. SEEPS representations are based on enumerating the parent of each node. Here, we detail the interpretation of each column. A tree looks like this example, from the getting started documentation.
## [,1] [,2] [,3] [,4] [,5] [,6] [,7]
## [1,] 1 32 1.0010000 0.3734455 0.0018690916 0 0
## [2,] 2 32 2.0010000 1.3734455 0.0068740830 0 0
## [3,] 3 31 3.0010000 1.9947840 0.0099838767 0 0
## [4,] 4 30 5.0010000 3.4439264 0.0172368222 0 0
## [5,] 5 29 23.0010000 16.6349876 0.0832579696 0 0
## [6,] 6 27 27.0010000 1.7993404 0.0090056832 0 0
## [7,] 7 28 30.0010000 5.4709745 0.0273821801 0 0
## [8,] 8 25 31.0010000 0.9999708 0.0050048454 0 0
## [9,] 9 24 33.0010000 1.6482525 0.0082494895 0 0
## [10,] 10 26 55.0000000 27.2799548 0.1365359389 1 82
## [11,] 11 22 37.0010000 2.4823942 0.0124243614 0 0
## [12,] 12 19 38.0010000 1.6213194 0.0081146896 0 0
## [13,] 13 22 55.0000000 20.4813942 0.1025092015 1 91
## [14,] 14 18 55.0000000 16.1211358 0.0806861461 1 92
## [15,] 15 21 55.0000000 19.8310839 0.0992544039 1 94
## [16,] 16 19 55.0000000 18.6203194 0.0931945382 1 97
## [17,] 17 18 55.0000000 16.1211358 0.0806861461 1 104
## [18,] 18 20 38.8788642 3.4479771 0.0172570959 0 0
## [19,] 19 20 36.3796806 0.9487936 0.0047487037 0 0
## [20,] 20 21 35.4308870 0.2619709 0.0013111620 0 0
## [21,] 21 23 35.1689161 3.6777947 0.0184073309 0 0
## [22,] 22 23 34.5186058 3.0274844 0.0151525334 0 0
## [23,] 23 24 31.4911214 0.1383739 0.0006925602 0 0
## [24,] 24 25 31.3527475 1.3517183 0.0067653387 0 0
## [25,] 25 26 30.0010292 2.2809840 0.0114163051 0 0
## [26,] 26 27 27.7200452 2.5183856 0.0126044982 0 0
## [27,] 27 28 25.2016596 0.6716341 0.0033615227 0 0
## [28,] 28 29 24.5300255 18.1640131 0.0909107293 0 0
## [29,] 29 30 6.3660124 4.8089389 0.0240686975 0 0
## [30,] 30 31 1.5570736 0.5508576 0.0027570373 0 0
## [31,] 31 33 1.0062160 1.0162071 0.0050861079 0 0
## [32,] 32 33 0.6275545 0.6375456 0.0031909103 0 0
## [33,] 33 0 0.0000000 0.0000000 0.0000000000 0 0
## [34,] 0 0 0.0000000 0.0000000 0.0000000000 0 0Column 1 stores the index of the node. Index 0 is the
root node. 1:N specify the tips that were used for reconstruction, the
remaining nodes are internal nodes.Column 2 denotes the parental node. Node 0 is the root
node. Arriving at node 0 implies that you have reached the root of the
tree.Column 3 stores the time (relative to the zero) of the
node. This can also be thought of as the “height” of the node.Column 4 stores the edge distance between the node and
it’s parent. Notice that the last (highest indexed) node is zero
distance from the root 0.Column 5 stores a realized distance. The interpretation
of this vlaue is context dependent. Here we used
geneology_to_phylogeny_bpb so the distance units are in
expected substitutions/site. If we use
stochastify_transmission_history, this column would be
integer valued.Column 6 stores a boolean (0 or 1) denoting if the node
is sampled. Some leaves may included in the tree that do not represent
samples to ensure that the within host due to the transmission history
is properly accounted for.Column 7 stores the leaf index for each sample. Column
7 has the same sparsity pattern as column 6.