build_motif_adjacency_matrix: Build a motif adjacency matrix In motifcluster: Motif-Based Spectral Clustering of Weighted Directed Networks

Usage

 ```1 2 3 4 5 6 7``` ```build_motif_adjacency_matrix( adj_mat, motif_name, motif_type = c("struc", "func"), mam_weight_type = c("unweighted", "mean", "poisson"), mam_method = c("sparse", "dense") ) ```

Arguments

 `adj_mat` Adjacency matrix from which to build the motif adjacency matrix. `motif_name` Motif used for the motif adjacency matrix. `motif_type` Type of motif adjacency matrix to build. One of `"func"` or `"struc"`. `mam_weight_type` The weighting scheme to use. One of `"unweighted"`, `"mean"` or `"product"`. `mam_method` Which formulation to use. One of `"dense"` or `"sparse"`. The sparse formulation avoids generating large dense matrices so tends to be faster for large sparse graphs.

Details

Entry (i, j) of a motif adjacency matrix is the sum of the weights of all motifs containing both nodes i and j. The motif is specified by name and the type of motif instance can be one of:

• Functional: motifs should appear as subgraphs.

• Structural: motifs should appear as induced subgraphs.

The weighting scheme can be one of:

• Unweighted: the weight of any motif instance is one.

• Mean: the weight of any motif instance is the mean of its edge weights.

• Product: the weight of any motif instance is the product of its edge weights.

Examples

 ```1 2``` ```adj_mat <- matrix(c(1:9), nrow = 3) build_motif_adjacency_matrix(adj_mat, "M1", "func", "mean") ```

Example output

```3 x 3 sparse Matrix of class "dgCMatrix"

[1,]  1.333333 14.000000 16.66667
[2,] 14.000000  6.666667 19.33333
[3,] 16.666667 19.333333 12.00000
```

motifcluster documentation built on Nov. 15, 2021, 9:06 a.m.