如何创建一个表，其中第一个表中的行也匹配 R 中第二个表的行中的 3 列

Question

I need to create a third table (df3) with the rows of a first table (df1) that match the row values of three of the 5 columns in a second table (df2). Two starting tables df1 and df2 do not have the same number of rows.

Example:

              df1                                                  df2 
chain   freq   color  length  type1  type2       chain   freq    color  length  type1  type2
AC       24     red    100      C      V2         BD      45      blue   73      C      G5
BD       57     green   87      C      G5         YJ      57      green  78      N      Y6
OP       83     yellow  68      R      Q9         TP       8      orange 98      Y      P2
TP       28     blue    74      Y      P2         HP      50      white  87      A      U9
HP       23     yellow  39      A      U9         ZS      87      red    98      P      N8
                                                  XC       8      green  98      T      N8

The resulting table has rows that are in df1 that match the columns chain , type1 , and type2 in df2. In this example it would look like this:

          df3                                
chain   freq    color  length  type1  type2
BD       57     green   87      C      G5
TP       28     blue    74      Y      P2
HP       23     yellow  39      A      U9

I'm trying to do this avoiding loops as much as possible. I've been looking through the dplyr functions, but I'm not quite familiar with the package yet. Any thoughts are appreciated.

Answer 1

We could use semi_join

library(dplyr)
semi_join(df1, df2, by = c('chain', 'type1', 'type2'))
#   chain freq  color length type1 type2
#1    BD   57  green     87     C    G5
#2    TP   28   blue     74     Y    P2
#3    HP   23 yellow     39     A    U9

data

df1 <- structure(list(chain = c("AC", "BD", "OP", "TP", "HP"), freq = c(24L, 
57L, 83L, 28L, 23L), color = c("red", "green", "yellow", "blue", 
"yellow"), length = c(100L, 87L, 68L, 74L, 39L), type1 = c("C", 
"C", "R", "Y", "A"), type2 = c("V2", "G5", "Q9", "P2", "U9")), class = "data.frame", row.names = c(NA, 
-5L))

df2 <- structure(list(chain = c("BD", "YJ", "TP", "HP", "ZS", "XC"), 
    freq = c(45L, 57L, 8L, 50L, 87L, 8L), color = c("blue", "green", 
    "orange", "white", "red", "green"), length = c(73L, 78L, 
    98L, 87L, 98L, 98L), type1 = c("C", "N", "Y", "A", "P", "T"
    ), type2 = c("G5", "Y6", "P2", "U9", "N8", "N8")), 
    class = "data.frame", row.names = c(NA, 
-6L))