R 中带有网络包的桑基图

Question

我正在尝试按照 R Graph Gallery 的说明创建一个简单的 Sankey 图：https://www.r-graph-gallery.com/322-custom-colours-in-sankey-diagram.html。我有一个数据集，每个 ID 有两个 obvs。对于每个时期，我都知道某人是否贫穷。数据集如下所示：

ID   YEAR   POVERTY
1    2018      0
1    2019      1
2    2018      1
2    2019      1
3    2018      0
3    2019      1
4    2018      0
4    2019      0
5    2018      0
5    2018      0

我想我需要将其转换为源-目标-值 table 但我不明白“值”的用途。有人会向我解释吗？我该如何推进它？

非常感谢您的帮助:)

我已经使用了提供的代码：

library("dplyr", warn.conflicts = FALSE)
library("networkD3")

diagram <- SUBSET05%>% 
  dplyr::mutate(Poverty = dplyr::if_else(Poverty==1, "poor", "not poor")) %>% 
  dplyr::transmute(id_nmbr, yr_interview, Poverty = paste(Poverty, yr_interview, sep = "_"))

links <- diagram %>%
  tidyr::pivot_wider(names_from = yr_interview, values_from = Poverty) %>% 
  dplyr::rename(source = `2018`, target = `2019`)

nodes <- data.frame(name = unique(c(links$source, links$target))) %>% 
  tidyr::separate(name, into = c("group", "year"), sep = "_", remove = FALSE)

links$id_nmbrsource <- match(links$source, nodes$name)-1 
links$id_nmbrtarget <- match(links$target, nodes$name)-1
links$value <- 10

sn <- sankeyNetwork(Links = links,
                    Nodes = nodes,
                    NodeID = "name",
                    Source = "id_nmbrsource",
                    Target = "id_nmbrtarget",
                    NodeGroup = "group",
                    Value = "value")
sn

我得到以下图像：

我的数据集有 34034 个观测值，每年 17017 个。因此我必须更改值列吗？是什么导致了丑陋的形象？

Answer 1

我不确定我是否真的理解您希望输出的样子。

不管怎样，我认为“价值”对你来说并不重要。每个连接具有相同的重要性，因此您可以将其设置为任意值。

如果重点只是显示有多少人从贫困走向非贫困，那么出发点应该是你实际上有四个群体：两次“贫困”和“非贫困”期间。

结果会是这样的：

library("dplyr", warn.conflicts = FALSE)
library("networkD3")


df <- tibble::tribble(
  ~ID, ~YEAR, ~POVERTY,
  "1", 2018, 0,
  "1", 2019, 1,
  "2", 2018, 1,
  "2", 2019, 1,
  "3", 2018, 0,
  "3", 2019, 1,
  "4", 2018, 0,
  "4", 2019, 0,
  "5", 2018, 0,
  "5", 2019, 0
) %>% 
  dplyr::mutate(POVERTY = dplyr::if_else(POVERTY==0, "poor", "not poor")) %>% 
  dplyr::transmute(ID, YEAR, POVERTY = paste(POVERTY, YEAR, sep = "_")) 


links <- df %>% 
  tidyr::pivot_wider(names_from = YEAR, values_from = POVERTY) %>% 
  dplyr::rename(source = `2018`, target = `2019`) 

nodes <- data.frame(name = unique(c(links$source, links$target))) %>% 
  tidyr::separate(name, into = c("group", "year"), sep = "_", remove = FALSE)


links$IDsource <- match(links$source, nodes$name)-1 
links$IDtarget <- match(links$target, nodes$name)-1
links$value <- 10


sn <- sankeyNetwork(Links = links,
                    Nodes = nodes,
                    NodeID = "name",
                    Source = "IDsource",
                    Target = "IDtarget",
                    NodeGroup = "group",
                    Value = "value") 

sn

R 中带有网络包的桑基图

Sankey Diagram with Network package in R

r

sankey-diagram