正在通过使用Quanteda（v1.3.4）中的textplot_network（）函数绘制作者句柄（@mentions）来分析Twitter数据。但是，图也显示了一些隔离的（未连接）节点。
想知道是否这是因为没有其他作者句柄（@mention）出现在此类节点之一的基础上，因此函数调用中设置了频率。还有一种方法可以将它们从频率共生矩阵（fcm）中删除，从而使它们不出现在图中。
require(data.table)
require(quanteda)
require(stringr)
not_clause_1 <- fread("final_query_output.csv", header = TRUE, skip = 0L,
stringsAsFactors = FALSE, select = c("Date","Full Text","Page
Type","Country", "Gender", "Author","Author City","Author Country","Author
State","City","County","Topics"))
setnames(not_clause_1, c("Page Type", "Full Text", "Author Country",
"Author State", "Author City"), c("Page_Type", "Full_Text",
"Author_Country", "Author_State", "Author_City"))
corpus1 <- corpus(not_clause_1$Full_Text, docnames =
paste(not_clause_1$Author, not_clause_1$idx, sep = "(")) ## using
quanteda's corpus
tokens1 <- quanteda::tokens(tolower(corpus1), what = "word", remove_numbers
= TRUE, remove_punct = TRUE, remove_symbols = TRUE, remove_separators =
TRUE, remove_url = TRUE, remove_hyphens = TRUE)
tokens1 <- tokens_remove(tokens1, stopwords("english"))
set.seed(12345)
mydfm <- dfm(tokens1, tolower = TRUE)
mentions_dfm <- dfm_select(mydfm, pattern = "@*") ## select only the author
handles beginning with @ ##
mentions_fcm <- fcm(mentions_dfm)
top_mention_names <- names(topfeatures(mentions_dfm, min(150,
length(featnames(mentions_dfm)))))
topmention_fcm <- fcm_select(mentions_fcm, pattern = top_mention_names,
selection = "keep")
size <- log(colSums(fcm_select(topmention_fcm, top_mention_names)))
textplot_network(topmention_fcm, vertex_size = size/max(size)*3, edge_color
= "blue", edge_alpha = 0.8, edge_size = 3, omit_isolated = TRUE)
如问题概述中所述，我想知道某些作者例如：@moonbow_living在不带任何连接节点的fcm中的身影，以及是否有一种方法可以在调用texplot_network（ ）。
添加了一个链接，用于下载具有隔离节点的基础数据和网络图：reference files