比较R数据帧和A的多列中的值更新缺失值 [英] Comparing values in multiple columns in R dataframes & updating missing values

查看:74
本文介绍了比较R数据帧和A的多列中的值更新缺失值的处理方法,对大家解决问题具有一定的参考价值,需要的朋友们下面随着小编来一起学习吧!

问题描述

我有3个数据框。
first df包含一个列-名称-

I have 3 dataframes . first df contains one column - Name -

df 1
    Name 
    A    
    B    
    C    
    D    
    E    
    F    
    G  
    H
    I
    J
    K   

第二个df包含两列-名称和计数,但第一个df可能会或可能不会缺少某些名称

Second df contains two columns - Name and counts but some of the Names may or may not be missing from first df.

df 2 - 
  Name   Counts 
    A    12
    B    23
    C    34
    D    56
    E    34
    K    44

I要比较从第二个df到第一个df的所有名称,如果没有任何一个名称丢失,则可以。
如果缺少任何名称,则必须从第三df填充该名称及其计数。第三个df将始终具有可用的名称和计数。

I want compare all Names from second df to first df , If none of the names are missing , then fine. If any name is missing then that name and its count has to be filled from third df . The third df will always have names and counts available in it.

df 3 - 
 Name   Counts 
    A    34
    B    45
    C    34
    D    56
    E    67
    F    435
    G    45
    H    76
    I    76
    J    88
    K    90

因此在上面的示例中,由于F,G, H,I,J在第二个df中丢失,应从df 3中添加其信息。

So in above example Since F, G, H , I, J are missing in second df , their info should be added from df 3 .


第二个df应该更新为-

and second df should be updated as -

Name   Counts 
    A    12
    B    23
    C    34
    D    56
    E    34
    F    435
    G    45
    H    76
    I    76
    J    88
    K    44

任何帮助

谢谢

推荐答案

可以。

library(data.table)
setDT(DF1); setDT(DF2); setDT(DF3)

DF1[, n := unique(rbind(DF2, DF3), by="Name")[.(.SD$Name), on=.(Name), x.Counts]]

这会向DF1添加一列:

which adds a column to DF1:

    Name   n
 1:    A  12
 2:    B  23
 3:    C  34
 4:    D  56
 5:    E  34
 6:    F 435
 7:    G  45
 8:    H  76
 9:    I  76
10:    J  88
11:    K  44

您可以改为 merge(DF1,unique(rbind (DF2,DF3),by =名称),all.x = TRUE),尽管那样会创建一个新表,而不是在现有表中添加列。此合并的dplyr类似物是 left_join(DF1,bind_rows(DF2,DF3)%>%distinct(Name))

You could instead do merge(DF1, unique(rbind(DF2, DF3), by="Name"), all.x=TRUE), though that would create a new table instead of adding a column to an existing table. The dplyr analogue of this merge is left_join(DF1, bind_rows(DF2, DF3) %>% distinct(Name)).

工作原理


  • DF = rbind(DF2,DF3 )追加两个源表

  • uDF = unique(DF,by = Name)每个名称
  • 的第一行
  • DF1 [,n:= z] 将值 z 的列 n 添加到 DF1

  • z = x [i,on =,xv] 使用 i 查找 x 的上一行,然后返回列 v ,其中...


    • x = uDF

    • v =计数

    • i = .SD $ Name 是在 DF1

    • DF = rbind(DF2, DF3) appends the two source tables
    • uDF = unique(DF, by="Name") keeps the first row for each Name
    • DF1[, n := z] adds column n with values z to DF1
    • z = x[i, on=, x.v] uses i to look up rows of x then returns column v, where...
      • x = uDF
      • v = Counts
      • i = .SD$Name is the vector of names found in DF1

      .SD DT的 j 中是指 DT 本身就是数据子集。

      .SD in j of DT[i, j] refers to DT itself, the "Subset of Data".

      这篇关于比较R数据帧和A的多列中的值更新缺失值的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!

查看全文
登录 关闭
扫码关注1秒登录
发送“验证码”获取 | 15天全站免登陆