How to covert columns of a data.frame to numeric vectors

Question

data1 <- structure(list(y = c(0.786495547305097, -0.275294469743402, 5.01043554797471, 
3.80786336657872, 2.13918082927747), x1 = c(1.26295428488079, 
-0.326233360705649, 1.3297992629225, 1.2724293214294, 0.414641434456408
), x2 = c(-0.729979080566544, -0.977769604136991, 1.60183697933767, 
0.911948278954854, 0.879294677497606), x3 = c(-0.39000956273028, 
-1.81922223008177, 0.659180710219526, 0.459621672673119, 1.6166263368295
)), row.names = c(NA, 5L), class = "data.frame")

data2 <- structure(list(V8 = c(-1.72904385751157, -1.07084107622855, 0.485233893853134, 
-0.512561121790257, -0.584170977530818), V9 = c(-0.370376604243532, 
-0.264930923072872, 0.815334082977478, -0.188695978779318, -1.58942168300819
), V10 = c(-0.154982593979532, 1.29764064456055, -0.851074132826296, 
0.496249346485277, -1.72196830424353), y = structure(c(-1.33203440383281, 
0.117885549592542, 4.36303461564377, 0.572854292213718, -0.627729381607459
), .Dim = c(5L, 1L), .Dimnames = list(c("1", "2", "3", "4", "5"
), NULL))), row.names = c("1", "2", "3", "4", "5"), class = "data.frame")

I have 2 data sets. Both are of class data.frame

> class(data1)
[1] "data.frame"
> class(data2)
[1] "data.frame"

However, the columns in data1 are of type numeric , whereas those in data2 are of type matrix .

> class(data1$y)
[1] "numeric"
> class(data2$y)
[1] "matrix" "array"

How can I convert the columns in data2 to class numeric like those in data1 ?

Answer 1

It could be a result of scale ing operations creating a matrix column. An easy option is to call data.frame again with do.call

data2 <- do.call(data.frame, data2)
class(data2$y)
#[1] "numeric"

Or convert to matrix and then to data.frame as matrix conversion removes the attributes

as.data.frame(as.matrix(data2))

Or another option is to loop over the columns with lapply , convert to numeric again

data2[] <- lapply(data2, as.numeric)

Answer 2

Here is a way.
If you run

data2[ -ncol(data2) ]
data2[[ ncol(data2) ]]

you will see that the last column is an object of class "matrix" , like it is said in the question. So that column can be binded to the rest of the data.frame.

data3 <- cbind(data2[ -ncol(data2) ], y = data2[[ ncol(data2) ]])
str(data3)
#'data.frame':  5 obs. of  4 variables:
# $ V8 : num  -1.729 -1.071 0.485 -0.513 -0.584
# $ V9 : num  -0.37 -0.265 0.815 -0.189 -1.589
# $ V10: num  -0.155 1.298 -0.851 0.496 -1.722
# $ y  : num  -1.332 0.118 4.363 0.573 -0.628

Answer 3

c oncatenating.

r <- as.data.frame(lapply(data2, c))
str(r)
# 'data.frame': 5 obs. of  4 variables:
# $ V8 : num  -1.729 -1.071 0.485 -0.513 -0.584
# $ V9 : num  -0.37 -0.265 0.815 -0.189 -1.589
# $ V10: num  -0.155 1.298 -0.851 0.496 -1.722
# $ y  : num  -1.332 0.118 4.363 0.573 -0.628

Answer 4

Try the code like below

data2 <- list2DF(Map(c, data2))

or

data2 <- list2DF(Map(unlist, data2))