# MCMC estimates for the (multilevel) p2 model Estimates the (multilevel) p2 model parameters,as described in Zijlstra, Van Duijn and Snijders (2006) . ```r p2ML(nets, sender = NULL, receiver = NULL, density =~ 1, reciprocity =~ 1, adapt = NULL, burnin = NULL, center = NULL, separate= NULL, seed = NULL) ``` ## Arguments - `nets`: List with n dichotomous dependent directed networks. - `sender`: Optional matrix with a stacked actor-level sender covariate, corresponding to the n networks. Multiple sender covariates can be added as a formula object, see example below - `receiver`: Optional matrix with a stacked actor-level receiver covariate, corresponding to the n networks. Multiple receiver covariates can be added as a formula object - `density`: Optional stacked matrix with a density covariate, with dimensions similar to the n dependent networks. Multiple density covariates can be added as a formula object, see example below - `reciprocity`: Optional stacked matrix with a symmetric reciprocity covariate, with dimensions similar to the n dependent networks. Multiple reciprocity covariates can be added as a formula object - `adapt`: Optional number of adaptive sequenses (default is 125). - `burnin`: Optional specification of number of burn-in iterations (default is 2500). - `center`: Optional argument for centering predictors (default is TRUE). - `separate`: Optional argument for estimating separate coefficients for the n dependent networks (default is FALSE). - `seed`: Optonal specification of random seed (delfault is 1). ## Returns Returns a fitted model of class 2ML, to be opened with the function summary(). ## References Zijlstra, B. J., Van Duijn, M. A., & Snijders, T. A. (2006). The Multilevel p2 Model A random effects model for the analysis of multiple social networks. **Methodology: European Journal of Research Methods for the Behavioral and Social Sciences**, 2(1), 42. ## Author(s) Bonne J.H. Zijlstra b.j.h.zijlstra@uva.nl ## Examples ```r # create two very small networks with covariates for illustrative purposes Y1 <- 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) Y2 <- matrix(c(0,0,1,0,1,0,0,1,0,0, 0,0,0,0,0,0,0,1,1,0, 0,0,0,1,0,1,0,1,0,1, 0,0,1,0,0,0,1,1,0,0, 1,0,0,1,0,0,1,0,0,1, 0,0,1,0,0,0,1,1,0,0, 0,1,0,0,1,0,0,0,0,0, 1,0,1,0,1,1,1,0,0,1, 0,1,0,1,0,0,0,0,0,0, 0,1,0,1,0,0,0,1,0,0), ncol=10) Y <- list(Y1, Y2) Sa1 <- c(1,0,1,0,1,1,0,1,0,1) Sa2 <- c(1,0,0,1,0,0,1,1,0,1) Sa <- list(Sa1, Sa2) Sat <- Rat <- do.call(plyr::rbind.fill.matrix, Sa) Sb1 <- c(0,1,1,0,1,0,1,0,1,0) Sb2 <- c(1,0,1,0,0,1,0,1,0,1) Sb <- list(Sb1, Sb2) Sbt <- do.call(plyr::rbind.fill.matrix, Sb) Da1 <- abs(matrix(rep(Sa1,10), byrow = FALSE, ncol= 10) - matrix(rep(Sa1,10), byrow = TRUE, ncol= 10)) Da2 <- abs(matrix(rep(Sa2,10), byrow = FALSE, ncol= 10) - matrix(rep(Sa2,10), byrow = TRUE, ncol= 10)) Da <- list(Da1, Da2) Dat <- do.call(plyr::rbind.fill.matrix, Da) # estimate p2 model for two networks M1 <- p2ML(Y, sender= ~ Sat + Sbt, receiver= ~ Rat, density = ~ Dat, adapt = 10, burnin = 100) summary(M1) # Notice: burn-in, and number of adaptive sequenses are # much smaller than recommended to keep computation time low. # recommended code: ## Not run: M1 <- p2ML(Y,sender= ~ Sat + Sbt, receiver= ~ Rat, density = ~ Dat) summary(M1) ## End(Not run) # estimate p2 model for a single network M2 <- p2ML(list(Y[[1]]),sender= ~ Sat[1:10,] + Sbt[1:10,], receiver= ~ Rat[1:10,], density = ~ Dat[1:10,], adapt = 10, burnin = 100) summary(M2) # Notice: burn-in, and number of adaptive sequenses are # much smaller than recommended to keep computation time low. # recommended code: ## Not run: M2 <- p2ML(list(Y[[1]]),sender= ~ Sat[1:10,] + Sbt[1:10,], receiver= ~ Rat[1:10,], density = ~ Dat[1:10,]) summary(M2) ## End(Not run) ```

p2ML function