CODEWITHSUNDEEP
rdplyr
Aaron Parrilla
Aaron Parrilla

Reputation: 542

Is there a way to append values of columns from 2 different data frames

I have two data frames
df1:

DAT1 DAT3     DAT4    ...
 1   this is  this is
 2   this is  this is
 3   this is  this is

df2:

DAT1 DAT3       DAT4      ... 
 1   a comment  a comment
 2   a comment  a comment
 3   a comment  a comment

I want to find a way to append the second dataframe columns (I know both the name and the position of the columns I need to append)to the first one and obtain an updated version of the first one that has:
df3:

DAT1 DAT3               DAT4               ... 
 1   this is a comment  this is a comment  
 2   this is a comment  this is a comment
 3   this is a comment  this is a comment

The thing is that the real dataframes have many rows and columns, so a for() loop would be really inefficient.

Upvotes: 0

Views: 42

Answers (4)

akrun
akrun

Reputation: 886968

We can use base R without looping

cols <- c("DAT3", "DAT4")     
df3 <- df1
df3[cols] <-matrix(paste(as.matrix(df1[-1]), as.matrix(df2[-1])), nrow = nrow(df1))
df3
#  DAT1              DAT3              DAT4
#1    1 this is a comment this is a comment
#2    2 this is a comment this is a comment
#3    3 this is a comment this is a comment

data

df1 <- structure(list(DAT1 = 1:3, DAT3 = c("this is", "this is", "this is"
), DAT4 = c("this is", "this is", "this is")), class = "data.frame",
row.names = c(NA, 
-3L))

df2 <- structure(list(DAT1 = 1:3, DAT3 = c("a comment", "a comment", 
"a comment"), DAT4 = c("a comment", "a comment", "a comment")),
   class = "data.frame", row.names = c(NA, 
-3L))

Upvotes: 2

Ronak Shah
Ronak Shah

Reputation: 388817

We can use Map

cols <- c("DAT3", "DAT4")
df3 <- df1
df3[cols] <- Map(paste, df1[cols], df2[cols])

df3
#  DAT1              DAT3              DAT4
#1    1 this is a comment this is a comment
#2    2 this is a comment this is a comment
#3    3 this is a comment this is a comment

Upvotes: 2

Jason Mathews
Jason Mathews

Reputation: 805

a dplyr version

df1 %>% inner_join(df2, by = "DAT1") %>% rowwise() %>%
  mutate(DAT3 = paste(DAT3.x, DAT3.y, collapse = " "),
         DAT4 = paste(DAT4.x, DAT4.y, collapse = " ")) %>%
  select(everything(), -contains("."))

OutPut

# A tibble: 3 x 3
   DAT1 DAT3              DAT4             
  <dbl> <chr>             <chr>            
1     1 this is a comment this is a comment
2     2 this is a comment this is a comment
3     3 this is a comment this is a comment

Upvotes: 0

LyzandeR
LyzandeR

Reputation: 37879

If your data is ordered, I would do something like this:

#initiate the data.frame with the id
df3 <- data.frame(DAT1 = df1$DAT1)

#then run a for-loop with the names you know you need to concatenate
for (i in c('DAT3', 'DAT4')) {
  df3[[i]] <- paste(df1[[i]], df2[[i]])
}

The for-loop iterates over the names only. The core of the code is paste which is vectorised and fast. So, you won't face any speed issues

df3
#  DAT1              DAT3              DAT4
#1    1 this-is a-comment this-is a-comment
#2    2 this-is a-comment this-is a-comment
#3    3 this-is a-comment this-is a-comment

Upvotes: 1

Related Questions