R：reshape2包

最新推荐文章于 2024-01-02 10:39:47 发布

weixin_34292924

最新推荐文章于 2024-01-02 10:39:47 发布

阅读量350

点赞数 1

文章标签： python

原文链接：https://my.oschina.net/u/3093769/blog/842790

版权

2019独角兽企业重金招聘Python工程师标准>>>

1、cast函数

cast函数,把长型数据转换成你想要的任何宽型数据：

dcast(data, formula, fun.aggregate = NULL, ..., margins = NULL, subset = NULL, fill = NULL, drop = TRUE, value.var = guess_value(data))
acast(data, formula, fun.aggregate = NULL, ..., margins = NULL, subset = NULL, fill = NULL, drop = TRUE, value.var = guess_value(data))
#参数：
#data  要进行转换的数据框
#formula  用于转换的公式
#fun.aggregate   聚合函数，表达式为：行变量~列变量~三维变量~
#margins  用于添加边界汇总数据
#subset   用于添加过滤条件，需要载入plyr包

acast,dcast的区别在于输出结果。acast 输出结果为vector/matrix/array，dcast 输出结果为data.frame。

> library(reshape2)
> x<-data.frame(id=1:6,name=c("wang","zhang","li","chen","zhao","song"),shuxue=c(89,85,68,79,96,53),yuwen=c(77,68,86,87,92,63))
> x
  id  name shuxue yuwen
1  1  wang     89    77
2  2 zhang     85    68
3  3    li     68    86
4  4  chen     79    87
5  5  zhao     96    92
6  6  song     53    63
> x1<-melt(x,id=c("id","name"))    #先使用melt函数对数据进行融化操作，把数据已经变成长型数据
> x1
   id  name variable value
1   1  wang   shuxue    89
2   2 zhang   shuxue    85
3   3    li   shuxue    68
4   4  chen   shuxue    79
5   5  zhao   shuxue    96
6   6  song   shuxue    53
7   1  wang    yuwen    77
8   2 zhang    yuwen    68
9   3    li    yuwen    86
10  4  chen    yuwen    87
11  5  zhao    yuwen    92
12  6  song    yuwen    63
> x2<-acast(x1,id~variable)
> x2
  shuxue yuwen
1     89    77
2     85    68
3     68    86
4     79    87
5     96    92
6     53    63
> x3<-dcast(x1,id~variable)
> x3
  id shuxue yuwen
1  1     89    77
2  2     85    68
3  3     68    86
4  4     79    87
5  5     96    92
6  6     53    63    
#从以上两个执行结果来看，可以看出acast和dcast的区别：这里acast输出结果省略了id这个列，而dcast则输出id列。
> x4<-acast(x1,id~name~variable)
> x4
, , shuxue

  chen li song wang zhang zhao
1   NA NA   NA   89    NA   NA
2   NA NA   NA   NA    85   NA
3   NA 68   NA   NA    NA   NA
4   79 NA   NA   NA    NA   NA
5   NA NA   NA   NA    NA   96
6   NA NA   53   NA    NA   NA

, , yuwen

  chen li song wang zhang zhao
1   NA NA   NA   77    NA   NA
2   NA NA   NA   NA    68   NA
3   NA 86   NA   NA    NA   NA
4   87 NA   NA   NA    NA   NA
5   NA NA   NA   NA    NA   92
6   NA NA   63   NA    NA   NA
#三维的情况下acast输出的是一个数组，而dcast则报错，因为dcast输出结果为数据框。
> x5<-dcast(x1,id~variable,mean,margins=T)
> x5
     id   shuxue    yuwen    (all)
1     1 89.00000 77.00000 83.00000
2     2 85.00000 68.00000 76.50000
3     3 68.00000 86.00000 77.00000
4     4 79.00000 87.00000 83.00000
5     5 96.00000 92.00000 94.00000
6     6 53.00000 63.00000 58.00000
7 (all) 78.33333 78.83333 78.58333
#可以看到，边缘多了两列汇总数据是对行列求平均的结果
> x6<-dcast(x1,id~variable,mean,margins=c("id"))
> x6
     id   shuxue    yuwen
1     1 89.00000 77.00000
2     2 85.00000 68.00000
3     3 68.00000 86.00000
4     4 79.00000 87.00000
5     5 96.00000 92.00000
6     6 53.00000 63.00000
7 (all) 78.33333 78.83333
#只对列求平均值，当然也可以只对行求平均值，把id改成variable就可以了
> x7<-dcast(x1,id~variable,mean,margins=c("variable"))    
> x7
  id shuxue yuwen (all)
1  1     89    77  83.0
2  2     85    68  76.5
3  3     68    86  77.0
4  4     79    87  83.0
5  5     96    92  94.0
6  6     53    63  58.0
> library(plyr)
> x8<-dcast(x1,id~variable,mean,subset=.(id==1|id==3))
> x8
  id shuxue yuwen
1  1     89    77
2  3     68    86
#subset的筛选功能强大可以进行各种各样的筛选操作，类似filter的作用

2、melt函数

> x
  id  name shuxue yuwen
1  1  wang     89    77
2  2 zhang     85    68
3  3    li     68    86
4  4  chen     79    87
5  5  zhao     96    92
6  6  song     53    63
> x9<-melt(x,id=1:2,variable.name="课目",value.name="值")
> x9
   id  name   课目 值
1   1  wang shuxue 89
2   2 zhang shuxue 85
3   3    li shuxue 68
4   4  chen shuxue 79
5   5  zhao shuxue 96
6   6  song shuxue 53
7   1  wang  yuwen 77
8   2 zhang  yuwen 68
9   3    li  yuwen 86
10  4  chen  yuwen 87
11  5  zhao  yuwen 92
12  6  song  yuwen 63

转载于:https://my.oschina.net/u/3093769/blog/842790