Question

下面是我的r脚本，用于规范化数据集，然后使用dist（）查找欧氏距离。

    normalize_data <-  function (wine_data1)
      {
        feature_mat <- (wine_data1[,2:14]) #first column is class label hence normalizing only features                              
        #print(feature_mat)
        min_vector=apply(feature_mat, 2, min)
        max_vector=apply(feature_mat, 2, max)
        feature_mat <- sweep(feature_mat, 2, min_vector, FUN="-") 
        feature_mat =sweep(feature_mat, 2,  min_vector-max_vector, FUN="/")
        distance_mat <- dist(feature_mat, method = "euclidean", diag = FALSE, upper= FALSE)
        closest_dist <- apply(distance,1,min)
        return(closest_dist)
      }
    normdata <-normalize_data(winedata)
    normdata1 <- cbind(winedata[,1],normdata) #here i am binding the class labels with the normalized data


My euclidean distance matrix looks like below:
[1, 0, 1.2, 1.3, 1.4]
[1, 1.2, 0, 3.4, 5.1]
[2, 1.3, 1.7, 0, 3.4]
[2, 1.4, 1.9, 2.0, 0]

第一列是类标签。对于每一行，我只想找到距离较小的一行。所以我想要这样的东西：

[class label1，classlabel2，minvalue]

classlabel1是第一个点的标签，classlabel2是行（点）的标签，距离最小，minvalue是一个点与其他点之间的最小距离

但是当我申请时（dist_mat，1，min）我得到全部为零。

非常感谢任何帮助。提前致谢

Answer 1

如果我明白你的要求，你只需要做以下事情：

mat<- matrix(c(0,1,3,4,
           3,0,2,4,
           1,2,0,6,
           4,8,2,0),ncol=4,byrow=T,dimnames =list(c("R1","R2","R3","R4"),
                                                  c("C1","C2","C3","C4")))

apply(mat,1,function(x)colnames(mat)[order(x)[2]])

您将获得列的名称，其最小值不为零：

 R1   R2   R3   R4 
"C2" "C3" "C1" "C3"

如何摆脱R中的对角线

1 个答案: