代码之家 › 专栏 › 技术社区 › Ryan Porter

计算R中3个变量之间的相关性

correlation igraph r

1

Ryan Porter · 技术社区 · 7 年前

以下是链接结构:

structure(list(from = c("Alabama", "Alabama", "Alabama", "Alabama", "Alabama", "Alabama"), to = c("Alaska", "Arizona", "Arkansas", "California", "Colorado", "Connecticut"), weight = c(423L, 894L, 2057L, 3045L, 2328L, 1102L)), row.names = c(NA, 6L), class = "data.frame")

structure(list(State = c("Hawaii", "Alaska", "South Dakota", "Maine", "Colorado", "Vermont"), Well.Being.Score = c(65.2, 64, 63.7, 63.6, 63.5, 63.5), Social.Rank = c(46L, 1L, 29L, 18L, 21L, 5L), Financial.Rank = c(1L, 2L, 3L, 10L, 19L, 39L)), row.names = c(NA, 6L), class = "data.frame")

1 回复 | 直到 7 年前

1

Julius Vainora 7 年前

我想说有几对变量的相关系数在这里很有趣。特别地:

“至”州的权重与幸福感(WB),
“来自”状态的重量与WB,
重量vs.(“至”的WB-“至”的WB)。

因此,我们可以从进行双重合并开始

m <- merge(merge(links, nodes, by.x = "to", by.y = "State"), 
           nodes, by.x = "from", by.y = "State", suff = c(".to", ".from"))

现在我们所有感兴趣的变量都在同一个地方

with(m, cor(cbind(weight, WB.from = Well.Being.Score.from, 
                  WB.to = Well.Being.Score.to,
                  WB.diff = Well.Being.Score.to - Well.Being.Score.from)))

应该返回一个有趣的相关矩阵(这里没有输出,因为可用数据太少)。