Sum rows in data.frame or matrix

Question

I have a very large dataframe with rows as observations and columns as genetic markers. I would like to create a new column that contains the sum of a select number of columns for each observation using R.

If I have 200 columns and 100 rows, I would like a to create a new column that has 100 rows with the sum of say columns 43 through 167. The columns have either 1 or 0. With the new column that contains the sum of each row, I will be able to sort the individuals who have the most genetic markers.

I feel it is something close to:

data$new=sum(data$[,43:167])

Answer 1

you can use rowSums

rowSums(data) should give you what you want.

Answer 2

The rowSums function (as Greg mentions) will do what you want, but you are mixing subsetting techniques in your answer, do not use "$" when using "[]", your code should look something more like:

data$new <- rowSums( data[,43:167] )

If you want to use a function other than sum, then look at ?apply for applying general functions accross rows or columns.

Answer 3

I came here hoping to find a way to get the sum across all columns in a data table and run into issues implementing the above solutions. A way to add a column with the sum across all columns uses the cbind function:

cbind(data, total = rowSums(data))

This method adds a total column to the data and avoids the alignment issue yielded when trying to sum across ALL columns using the above solutions (see the post below for a discussion of this issue).

Adding a new column to matrix error

Answer 4

This could also help, however the best option is beyond any doubt the rowSums function:

data$new <- Reduce(function(x, y) {
  x + data[, y]
}, init = data[, 43], 44:167)