netdiffuseR-graphs: Network data formats
Description
List of accepted graph formats
diffnet objects
In the case of diffnet-class objects, the following arguments can be omitted
when calling fuictions suitable for graph objects:
toa: Time of Adoption vector
adopt: Adoption Matrix
cumadopt: Cumulative Adoption Matrix
undirected: Whether the graph is directed or not
Objects' names
When possible, netdiffuseR will try to reuse graphs dimensional names,
this is, rownames, colnames, dimnames
and names (in the case of dynamic graphs as lists). Otherwise,
when no names are provided, these will be created from scratch.Details
The netdiffuseR package can handle different types of graph
objects. Two general classes are defined accross the package's functions:
static graphs, and dynamic graphs.
- In the case of static graphs, these are represented as adjacency
matrices of size $n * n$ and can be either
matrix
(dense matrices) or dgCMatrix
(sparse matrix from the Matrix package). While
most of the package functions are defined for both classes, the default output
graph is sparse, i.e. dgCMatrix.
- With respect to dynamic graphs, these are represented by either
a
diffnet object, an array of size
$n * n * T$, or a list of size $T$
with sparse matrices (class dgCMatrix) of size $n * n$.
Just like the static graph case, while most of the functions accept both
graph types, the default output is dgCMatrix.