how to transform a data file

Hello R people,

I have a data file with 101 numeric variables: one variable called IDN
(the
individual's unique id number, which I need to retain, and which ranges
from 1000 to 1320; some numbers are obviously skipped), and V1 to V100
(each has a value of 0 or 1; these 100 variables represent sequentially
ordered days and whether a characteristic was present or absent--e.g.,
v1
is day 1 and a "1" means the characteristic is present; v10 is day 10
and
"0" means the characteristic is absent).

This may be child's play for many on this list, but how do I transform
this
data file to two columns, one called "id"  and another column named "c"
with 100 rows? I think it will end up being a 1000 row file. I've read
some
and think that I'm trying to "melt" my existing data. I can transpose
the
v1 to v100 with t(v1 to v100) but then I'm unclear on how to
automatically
generate 100 identical IDN's for each case and variable and then put
them
together.

This may be redundant, but for the sake of clarity, what I'm trying to
do
is get from this:

IDN      V1 V2 V3 ? V100

1       0   1   0  . . .       1

2       1  1   1   . . .       0

4        0  1   0   . . .      1

.

.

100   0  1   0   . . .       1

To this:

id   c

1    0

1    1

1    0

.      .

.    [continue 96 more times for c4 - c100]

1     1

2     1

2     1

2     1

.       .

. [continue 96 more times for c4 - c100]

2     0

.

.

.

[then repeat this for the next 98 cases]

100    0  1   0              1


   Thank you very much.

PJ

[[alternative HTML version deleted]]

Hi PJ,

Try

# some data
id <- 1:20
m <- matrix(sample(0:1, 200, TRUE), ncol = 10)
colnames(m) <- paste('V', 1:10, sep = "")
d <- data.frame(id, m)
d

# option 1
cbind(rep(d$id, each = ncol(d)-1), matrix(unlist(t(d[,-1])), ncol = 1))

# option 2
cbind(rep(d$id, each = ncol(d) - 1), stack(d[,-1])[,-2])

HTH,
Jorge.-


On Mon, Nov 28, 2011 at 10:19 PM, pat j <> wrote:

Hello R people,

I have a data file with 101 numeric variables: one variable called IDN 

individual's unique id number, which I need to retain, and which 

from 1000 to 1320; some numbers are obviously skipped), and V1 to V100
(each has a value of 0 or 1; these 100 variables represent 

ordered days and whether a characteristic was present or absent--e.g., 

is day 1 and a "1" means the characteristic is present; v10 is day 10 

"0" means the characteristic is absent).

This may be child's play for many on this list, but how do I transform 

data file to two columns, one called "id"  and another column named 

with 100 rows? I think it will end up being a 1000 row file. I've read 

and think that I'm trying to "melt" my existing data. I can transpose 

v1 to v100 with t(v1 to v100) but then I'm unclear on how to 

generate 100 identical IDN's for each case and variable and then put 

together.

 This may be redundant, but for the sake of clarity, what I'm trying 

is get from this:

IDN      V1 V2 V3 ? V100

1       0   1   0  . . .       1

2       1  1   1   . . .       0

4        0  1   0   . . .      1

.

.

100   0  1   0   . . .       1

To this:

id   c

1    0

1    1

1    0

.      .

.    [continue 96 more times for c4 - c100]

1     1

2     1

2     1

2     1

.       .

. [continue 96 more times for c4 - c100]

2     0

.

.

.

 [then repeat this for the next 98 cases]

100    0  1   0              1


   Thank you very much.

PJ

       [[alternative HTML version deleted]]

   [[alternative HTML version deleted]]