CODEWITHSUNDEEP
r
PrincessJellyfish
PrincessJellyfish

Reputation: 149

Extract from table in R

I have a problem which should be to hard but I can't get it the way i want. My df have about 200 rows and almost 4200 columns. The only thing that exist are 0, 1 and 2 in the df. Something like: 

    Col1 Col2 Col3 ...
Row1  2    1   0
Row2  1    1   1
Row3  2    2   2
.
.

I now want to save how many times number 2 apper in each coloumn so i wrote 

freq<-sapply(df,table)

but here is where I get stuck. I want to save how many 2:s there was in every coloumn and save it in a ny df and keep track of which coloumn had that many 2:s.

Want it to look something like this (if it was done on the example above):

Col1 Col2 Col3
  2    1    1

Sorry that I explaning a bit bad right now. Please tell me if I should add something to make it easier to understad my problem.

Upvotes: 1

Views: 53

Answers (2)

SabDeM
SabDeM

Reputation: 7200

Another way in base R. Just use the table function as you did previously and then subset to obtain only the values you need (in this case 2).

dft <- sapply(df1, table)
lapply(dft, function(x) x[[2]] )
$Col1
[1] 2

$Col2
[1] 1

$Col3
[1] 1

Upvotes: 1

akrun
akrun

Reputation: 887941

You can try colSums

 colSums(df==2)
 #Col1 Col2 Col3 
 # 2    1    1

We create a logical matrix (df==2) and then get the sum of each column by colSums

For big datasets, it may be better to loop as the big matrix from df==2 may cause memory issues

  vapply(df, function(x) sum(x==2), numeric(1))

data

 df <- structure(list(Col1 = c(2L, 1L, 2L), Col2 = c(1L, 1L, 2L),
 Col3 = 0:2), .Names = c("Col1", 
 "Col2", "Col3"), class = "data.frame", row.names = c("Row1", 
 "Row2", "Row3"))

Upvotes: 2

Related Questions