CODEWITHSUNDEEP
rdata.table
Pavan Ebbadi
Pavan Ebbadi

Reputation: 932

R data.table, Replacing each cell value if greater than 0 with 1

I have R data.table There are 25 columns, 1st column is ID, 24 columns are integer variables. There are close to 1Million rows. How do i convert all these non-zero values to 1

Custid  A   B  C
123     0   8  0
124     0   0  6

Should become 

Custid  A   B  C
123     0   1  0
124     0   0  1

Upvotes: 3

Views: 6377

Answers (3)

akrun
akrun

Reputation: 887951

An efficient option is the set function which replace in place. Loop through the columns using a for loop and set the 'value' to 1 where the element is not equal to 0 by specifying the 'i' and 'j' index.

for(j in 2:ncol(dt)){
   set(dt, i= which(dt[[j]]!=0), j=j, value =1)
}
dt
#   Custid A B C
#1:    123 0 1 0
#2:    124 0 0 1

Or another option is with lapply by looping over the Subset of Data.table after specifying the .SDcols

dt[, names(dt)[-1] := lapply(.SD, function(x) as.integer(x!=0)), .SDcols = 2:ncol(dt)]

Upvotes: 4

Hardik Gupta
Hardik Gupta

Reputation: 4790

this should also work

apply(df, 1, function(x) {ifelse(any(x > 0), 1, 0)})

Upvotes: 1

A.Yazdiha
A.Yazdiha

Reputation: 1378

Assuming your data.table is called 'dt',

df = as.data.frame(dt)
df[,-1] = (df[,-1] != 0)*1

works. the -1 index indicates the non-inclusion of the first column, and then inside the parenthesis returns a true or false statement and *1 ensures that the result is returned in numeric form.

If you want you can turn it back to a data.table.

dt = data.table(df)

Upvotes: 5

Related Questions