robertness
8/4/2015 - 3:42 PM

Use glm to predict child on its parents

Use glm to predict child on its parents

glm_igraph.Rd
content_copyfile_download
y_var <- V(g)[v]
        x_vars <- V(g)[setdiff(iparents(g, y_var), V(g)[is.bias])]
        fitted_vals <- .data %>%
          select(c(y_var$name, x_vars$name)) %>%
          `names<-`(paste("v", x_vars$name, sep="")) %>%
          mutate(Y = unlist(y_var$output.signal)) %>%
          {suppressWarnings(glm(Y ~ ., family = binomial(link="logit"), data = .data))} %>%
          fitted