frequency of elements in a data frame

Question

I have the following data frame (this is an example and the dataframe can contain more columns)

SelectVar 
     a    b    c    l    p    v   aa   ff
1 Dxa2 Dxa2 Dxa2 Dxa2 Dxa2 Dxb2 Dxb2 Dxb2
2 Dxb2 Dxb2 Dxb2 Dxd2 Dxi2 Dxc2 Dxd2 Dxi2
3 Dxc2 Dxd2 Dxi2 Dxi2 tneg Dxd2 Dxi2 tneg

I would like to count the frequency of the elements without convert it into a vector and using table or by indicating the element like in

length(SelectVar[SelectVar=="Dxa2"])

Is there any other way as to count the frequencies of the elements than the two mentioned in the above paragraph for the sample dataframe.

What is your ultimate goal? If you can explain what you're going to do with the frequency counts, we may be able to suggest a different approach. — Carl Witthoft, Feb 05 '14 at 16:18
I produce repeated dataframes as above where the column names can be different, longer or shorter all of three rows. I would not want to have to have to manipulate the data every time as I would like to link the three most frequent elements to another expression. in creating a vector to use by a table I have to manipulate the data every time due that both the elements and column names may be different — Barnaby, Feb 05 '14 at 16:20
Isn't constructing a simple function helpful for what you describe in your comment? Something like `f = function(mydf) sort(table(unlist(mydf)))`. And, also, maybe you could, somehow, have all your dataframes in a list and manipulate them with `lapply`? — alexis_laz, Feb 05 '14 at 16:30
Out of curiosity, why are using `data.frame` for this? If the columns all contain the same data type, `matrix` seems more appropriate and perhaps easier to manipulate. — dg99, Feb 05 '14 at 16:39

score 2 · Accepted Answer · edited May 23 '17 at 12:07

I think you asked the same question yesterday, Counting the frequency of an element in a data frame

answer modified from, dickoa's answer to previous question instead of data.frame, if you make it matrix, table() does not need vectorization and it should work.

df <- read.table(text = "   b    c    e    f    g    h    j 
1 Dxa2 Dxa2 Dxa2 Dxa2 Dxa2 Dxa2 Dxa2
2 Dxb2 Dxb2 Dxb2 Dxb2 Dxc2 Dxc2 Dxc2
3 Dxd2 Dxi2 tneg tpos Dxd2 Dxi2 tneg", header = TRUE, row.names = 1)

ll<-data.frame(table(as.matrix(df)))

now, you can sort by freq and select top 3

head(ll[order(ll$Freq, decreasing=T),],3)

I did not realize that by converting into matrix I could make the job much easier, thank you — Barnaby, Feb 05 '14 at 17:22

frequency of elements in a data frame

1 Answers1