CODEWITHSUNDEEP
rmatrixdataframereplace
Tobi1990
Tobi1990

Reputation: 123

Replace values in row of a data.frame by values of corresponding row a vector

I want to replace each value of a row of data.frame, for with a equal sized logical matrix has the value TRUE, by the value of a corresponding row of a vector. Here is an example:

> df1
     [,1] [,2] [,3] [,4]
[1,]    2    8    9    4
[2,]    2    6    6    6
[3,]    4    9    4    8

> a
     [,1]  [,2]  [,3]  [,4]
[1,] TRUE FALSE FALSE  TRUE
[2,] TRUE FALSE FALSE FALSE
[3,] TRUE FALSE  TRUE FALSE

> df2
[1] 5 4 6

So the result should look like this:

> df1
     [,1] [,2] [,3] [,4]
[1,]    5    8    9    5
[2,]    4    6    6    6
[3,]    6    9    6    8

Is there any way to do this without the use of loops? Thanks

Upvotes: 1

Views: 1673

Answers (2)

akrun
akrun

Reputation: 887951

We replicate 'df2' by the row of 'a' or 'df1', based on the logical matrix of 'a', we change the value

ifelse(a, df2[row(a)], df1)
#     [,1] [,2] [,3] [,4]
#[1,]    5    8    9    5
#[2,]    4    6    6    6
#[3,]    6    9    6    8

If it is data.frame, then we need to just change as.matrix(df1) and it should work in both cases

df1[] <- ifelse(a, df2[row(a)], as.matrix(df1))

Or another option is

(df1*!a) + (df2[row(a)]*a)
#     [,1] [,2] [,3] [,4]
#[1,]    5    8    9    5
#[2,]    4    6    6    6
#[3,]    6    9    6    8

NOTE: Both the output are matrix as showed in the expected output and the initial dataset in the OP's example is matrix. But it works with both data.frame and matrix (with the edit)

Upvotes: 1

h3rm4n
h3rm4n

Reputation: 4187

You need some advanced subsetting for this:

df1[a] <- df2[row(a)[a]]

the output:

> df1
  V1 V2 V3 V4
1  5  8  9  5
2  4  6  6  6
3  6  9  6  8

This solution works on both a dataframe or a matrix.

Upvotes: 2

Related Questions