r语言 - 在一组igraph对象上迭代for循环，比较不同的聚类算法

Question

我想知道是否有人可以帮助我这个for循环。或者建议另一种方法获得我想要的东西(我知道代码有点像噩梦，但是我被如何使它更优雅难住了)。

问题:我在一个列表中有一些网络对象;我想比较这些网络的一些社区检测算法。我想一次性获得这些信息(不像这里的玩具示例，我的真实列表中有许多网络)。下面的示例让我接近我想要的信息，但它没有在最终输出中列出网络。我需要做些什么才能在最终数据帧中也包含网络的名称?

library(graph)
library(igraphdata)
library(igraph)
# create a couple of networks
data(karate)
g2 <- erdos.renyi.game(50, .2)
# stick the networks in a list
list_of_networks <- list(karate, g2)
# get all the community detection algorithms 
clusters <- grep("^cluster_", ls("package:igraph"), value=TRUE)[-1] 
# get rid of the algorithms I don't want
clusters <- clusters[!grepl("spinglass|fluid|leiden|walktrap|infomap|prop|optimal", clusters)]
# for loop stuff
# there are two networks in the network list
x <- 1:2
# a container for my for loop results
container <- vector("list", length = 2)
# nested for loop
for (cluster in clusters) {

for (i in x) {

result <- t(c(

modularity(do.call(cluster, list(list_of_networks[[i]]))), 
length(do.call(cluster, list(list_of_networks[[i]]))),
cluster

))

container[[i]] <- rbind(result, container[[i]])
}
}
container %>%
map(~ as_tibble(.)) %>%
bind_rows() %>% 
rename(modularity = 1,
num_communities = 2,
algorithm = 3)

Answer 1

您可以使用map_df的.id参数:

library(tidyverse)
container %>% 
map_df(~ as_tibble(.) %>% 
rename(modularity = 1,
num_communities = 2,
algorithm = 3), 
.id = "name")

这返回

# A tibble: 6 × 4
name  modularity        num_communities algorithm            
<chr> <chr>             <chr>           <chr>                
1 1     0.443854125672307 4               cluster_louvain      
2 1     0.4366391184573   5               cluster_leading_eigen
3 1     0.434521466989    3               cluster_fast_greedy  
4 2     0.224687153369057 5               cluster_louvain      
5 2     0.197245699823462 4               cluster_leading_eigen
6 2     0.193293130653502 4               cluster_fast_greedy  

所以你的网络被命名为1和2。

为了获得更好的信息，我们在你的代码中添加一行

container <- vector("list", length = 2)
# Add names here
names(container) <- unlist(lapply(list_of_networks, (x) x$name))
# or
# names(container) <- sapply(list_of_networks, (x) x$name)

现在

container %>% 
map_df(~ as_tibble(.) %>% 
rename(modularity = 1,
num_communities = 2,
algorithm = 3), 
.id = "name")

返回

# A tibble: 6 × 4
name                          modularity        num_communities algorithm            
<chr>                         <chr>             <chr>           <chr>                
1 Zachary's karate club network 0.444903581267218 4               cluster_louvain      
2 Zachary's karate club network 0.4366391184573   5               cluster_leading_eigen
3 Zachary's karate club network 0.434521466989    3               cluster_fast_greedy  
4 Erdos-Renyi (gnp) graph       0.210112108800672 4               cluster_louvain      
5 Erdos-Renyi (gnp) graph       0.188379542113001 2               cluster_leading_eigen
6 Erdos-Renyi (gnp) graph       0.21151675068263  3               cluster_fast_greedy