j2 function

MCMC estimates for the j2 model

Estimates j2 model parameters as described in Zijlstra (2017) doi:10.1080/0022250X.2017.1387858.

j2(net, sender = NULL, receiver = NULL , density = NULL, reciprocity = NULL, 
burnin = NULL, sample = NULL, adapt= NULL, center = NULL, seed = NULL)

Arguments

• net: Directed dichotomous n*n network (digraph).
• sender: Optional sender covariates of lenght n.
• receiver: Optinal receiver covariates of length n.
• density: Optional density covariates of dimensions n*n.
• reciprocity: Optional symmetric reciprocity covariates of dimensions n*n.
• burnin: Optional specification of number of burn-in iterations (default is 10000).
• sample: Optional specification of number of MCMC samples (default is 40000).
• adapt: Optional number of adaptive sequenses (default is 100).
• center: Optional boolean argument for centering predictors (default is TRUE).
• seed: Optonal specification of random seed (delfault is 1).

Returns

Returns a matrix with MCMC means, standard deviations, quantiles and effective sample sizes for j2 parameters.

References

Zijlstra, B.J.H. (2017). Regression of directed graphs on independent effects for density and reciprocity. The Journal of Mathematical Sociology 41 (4).

Author(s)

Bonne J.H. Zijlstra b.j.h.zijlstra@uva.nl

Examples

# create a very small network with covariates for illustrative purposes
S <- c(1,0,1,0,1,1,0,1,0,1)
REC <- c(0,0,1,1,0,0,1,1,0,0)
D1 <- matrix(c(0,1,0,1,0,1,0,1,0,0,
               0,0,1,1,0,1,0,1,0,1,
               1,1,0,0,1,0,0,0,0,0,
               1,1,1,0,1,0,0,0,0,1,
               1,0,1,0,0,1,1,0,1,1,
               0,0,0,0,0,0,1,1,1,1,
               0,0,0,0,0,1,0,1,0,1,
               1,0,0,0,0,1,1,0,1,1,
               0,1,0,1,0,1,0,1,0,0,
               0,0,1,1,1,0,0,0,0,0), ncol=10)
D2 <- abs(matrix(rep(S,10), byrow = FALSE, ncol= 10) -
            matrix(rep(REC,10), byrow = TRUE, ncol= 10))
R <- D1*t(D1)
Y <- matrix(c(0,0,1,1,1,1,0,0,1,1,
              0,0,0,1,1,1,0,0,1,0,
              1,1,0,1,1,1,0,0,1,1,
              0,1,1,0,1,1,0,1,1,0,
              1,1,1,1,0,1,1,0,1,1,
              0,1,1,1,1,0,1,1,1,0,
              1,0,1,0,1,1,0,1,0,1,
              0,1,1,1,0,1,1,0,1,1,
              1,0,1,0,1,0,1,1,0,1,
              1,1,1,0,0,1,1,1,1,0), ncol=10) 

# estimate j2 model
j2(Y,sender= ~ S, receiver =  ~ REC, density = ~ D1 + D2, reciprocity= ~ R,
   burnin = 100, sample = 400, adapt = 10)
# notice: burn-in, sample size and number of adaptive sequenses are 
# much smaller than recommended to keep computation time low.
# recommended code: 
## Not run:

j2(Y,sender= ~ S, receiver =  ~ REC, density = ~ D1 + D2, reciprocity= ~ R)
## End(Not run)