`nwdyads`

Dyad census

Syntax

nwdyads 
[netname]
netname may be a list of networks or a wildcard such as lazega_*, as in nwsummarize; the census is shown for each network and the r() results are those of the last one.

Description

Returns the dyad census of a network (or a list of networks). This is a way to characterize a network based on its dyads.

In directed networks, each dyad (pair of nodes i and j) can be one of the following:

1) M: mutually connected: M_ij = M_ji = 1

2) A: asymmetrically connected: M_ij = 1, but M_ji = 0

3) N: not connected at all: M_ij = M_ji = 1

In undirected network, each dyad (pair of nodes i and j) can be one of the following:

1) M: connected: M_ij = M_ji = 1

2) N: unconnected: M_ij = M_ji = 0

The command also returns the reciprocity of the network.

reciprocity = M / (M + A)

Examples

. nwwebuse florentine
. nwdyads flomarriage

    Dyad census:  flomarriage

    Mutual |      Null
-----------+-----------
        20 |       100

. nwwebuse glasgow
. nwdyads glasgow3

    Dyad census:  glasgow3

    Mutual |      Asym |      Null
-----------+-----------+-----------
        45 |        32 |      1148

Supported network types

Binary: yes (only) - the dyad census (mutual/asymmetric/null) is inherently a presence/absence classification. Directed: yes - a directed network gets the full M/A/N census; an undirected network collapses to M/N only (every tie is definitionally mutual). Weighted: not applicable. Signed: not applicable. Two-mode: yes - the null dyads are counted over the pairs of nodes from different modes; reciprocity of an undirected two-mode network has no meaning and is shown as n/a.

Stored results

Scalars:
r(_100) mutual dyads
r(_010) asymmetric dyads
r(_001) null dyads
r(reciprocity) M / (M + A)
Macros:
r(name) name of network

nwdyads