generalized aggregate by row - r

Question

I would like to aggregate by row. I know how to do this and have answered several questions here from others asking for help doing it. However, I want to generalize the aggregate formula and ideally not have the aggregated rows in a different order than they first appear in the original data set.
Here is an example set:
my.data <- read.table(text = '
0 0 0 1
0 0 0 1
2 2 2 2
2 2 2 2
0 4 0 0
0 4 0 0
2 2 0 0
2 2 0 0
2 2 0 0
2 2 0 0
', header = FALSE)
and my desired result:
desired.result <- read.table(text = '
0 0 0 1 2
2 2 2 2 2
0 4 0 0 2
2 2 0 0 4
', header = FALSE)
Here is one way to obtain the answer, albeit the rows are not in their original order:
my.data[,(ncol(my.data)+1)] = 1
aggregate(V5 ~ V1 + V2 + V3 + V4, FUN = sum, data=my.data)
V1 V2 V3 V4 V5
1 2 2 0 0 4
2 0 4 0 0 2
3 0 0 0 1 2
4 2 2 2 2 2
Here is an unsuccessful attempt to generalize the aggregate formula:
with(my.data, aggregate(my.data[,ncol(my.data)], by = list(paste0('V', seq(1, ncol(my.data)-1))), FUN = sum))
The order of the result is less important than the generalization.
Thank you for any advice.

Answer 1

Since it turned out that the desired result is just the frequency counts of unique rows, you could/should use table (as mentioned in the comments). table uses factor on its arguments and factor, if "levels" is not specified, sorts its input's unique (unique does not sort) to specify the levels. So, for table to "see" your levels (i.e. the desired order of rows) you need to call table on an explicitly specified factor.
tmp = do.call(paste, my.data)
as.data.frame(table(tmp))
# tmp Freq
#1 0 0 0 1 2
#2 0 4 0 0 2
#3 2 2 0 0 4
#4 2 2 2 2 2
res = table(factor(tmp, unique(tmp)))
as.data.frame(res)
# Var1 Freq
#1 0 0 0 1 2
#2 2 2 2 2 2
#3 0 4 0 0 2
#4 2 2 0 0 4
Instead of calling as.data.frame.table -where your rows have been concatenated- you could take advantage of unique.data.frame and use a call like:
data.frame(unique(my.data), unclass(res))
# V1 V2 V3 V4 unclass.res.
#1 0 0 0 1 2
#3 2 2 2 2 2
#5 0 4 0 0 2
#7 2 2 0 0 4

Answer 2

It might be useful to mention that the count function in the plyr package can also aggregate this quickly. Although, you still would lose the original order of rows.
> library(plyr)
> x <- count(my.data)
> x
## V1 V2 V3 V4 freq
## 1 0 0 0 1 2
## 2 0 4 0 0 2
## 3 2 2 0 0 4
## 4 2 2 2 2 2
To order the table as desired.result shows (and borrowing a snippet from #alexis_laz),
> pst <- do.call(paste, my.data)
> x[order(x$freq, as.factor(unique(pst))), ]
## V1 V2 V3 V4 freq
## 1 0 0 0 1 2
## 4 2 2 2 2 2
## 2 0 4 0 0 2
## 3 2 2 0 0 4

Answer 3

I like the posted answers, especially the answer by #alexis_laz since I tend to prefer base R. However, here is a general answer using aggregate. The order of the rows in the output differs from the order of their first appearance in the original data set, but at least the rows are tallied:
I borrowed the . in aggregate from #alexis_laz's comment:
my.data <- read.table(text = '
0 0 0 1
0 0 0 1
2 2 2 2
2 2 2 2
0 4 0 0
0 4 0 0
2 2 0 0
2 2 0 0
2 2 0 0
2 2 0 0
', header = FALSE)
my.data
my.count = rep(1, nrow(my.data))
my.count
aggregate(my.count ~ ., FUN = sum, data=my.data)
V1 V2 V3 V4 my.count
1 2 2 0 0 4
2 0 4 0 0 2
3 0 0 0 1 2
4 2 2 2 2 2