CODEWITHSUNDEEP
rdataframedplyr
PC-Stats
PC-Stats

Reputation: 95

R: Shift some rows by one column across table

I have data frame X that looks like this. It has 4 columns and 5 rows.

name age gender class
A       12      M      C1
B      10      F      C2
C      M      C1      N/A
D      F      C2      N/A
E      F      C1      N/A

I would like to shift all data from col 2 (age) and row 3 onward by one column to right so that gender and classes align leaving the wrongly filled age data as blank . My resulting set should look like:

name age gender class
A      12      M      C1
B      10      F      C2
C      N/A      M      C1
D      N/A      F      C2
E      N/A      F      C1

Please note: this is a situation from a very large dataset with 4 mil records and 52 cols. Any help will be much appreciated. Thanks in advance!

Upvotes: 2

Views: 3294

Answers (2)

Sowmya S. Manian
Sowmya S. Manian

Reputation: 3833

> df <- data.frame("name" = LETTERS[1:5], 
+            "age" = c(12, 10, "M","F","F"),
+            "gender" = c("M", "F", "C1", "C2", "C1"),
+            "class" = c("C1", "C2", NA,NA,NA))
> df
  name age gender class
1    A  12      M    C1
2    B  10      F    C2
3    C   M     C1  <NA>
4    D   F     C2  <NA>
5    E   F     C1  <NA>

> df[3:nrow(df),3:ncol(df)] <- df[3:nrow(df),2:ncol(df)]
Warning message:
In `[<-.data.frame`(`*tmp*`, 3:nrow(df), 3:ncol(df), value = list( :
provided 3 variables to replace 2 variables

> df
   name age gender class
1    A  12      M    C1
2    B  10      F    C2
3    C   M      M    C1
4    D   F      F    C2
5    E   F      F    C1
> df[3:nrow(df),2] <- NA
> df
   name  age gender class
1    A   12      M    C1
2    B   10      F    C2
3    C <NA>      M    C1
4    D <NA>      F    C2
5    E <NA>      F    C1

Upvotes: 2

user3603486
user3603486

Reputation:

Like this:

nc  <- ncol(dfr)
dfr[-(1:2), 3:nc] <- dfr[-(1:2), 2:(nc-1)]
dfr[-(1:2), 2] <- NA

The negative indices in the rows mean 'everything but rows 1 and 2'.

Upvotes: 6

Related Questions