CODEWITHSUNDEEP
r
Diegoal
Diegoal

Reputation: 329

Display absolute and relative frequencies with tabular

probably someone has an advice.

I m working on a report with knitR and the tables package. The function tabular generates my Latex code.

I want a table with absolute and relative frequencies, but I dont know how to get it.

I tried to apply within tabular an own function that generates the relative values:

relative <- function(x){
len <- length(x)
ll <- length(table(x))
lev <- unique(x)

out <- numeric(length = ll)
for(i in 1:ll){
   pos <- which(x == lev[i])
   out[i] <- length(pos)/len*100
}
return(out)

}

The function I put in the tabular function with Country and Type as two factors. 

  tabular(Country ~ Type*relative, data = cars)

Here the result:

         Type             
         PKW      SUV     
  Country relative relative
  Germany 0          0     
  Japan   0          0     
  Other   0        NaN     
  USA     0          0

The result is just the example of how it doesnt work :) My main target is a table that contains both: absolute AND relative frequencies with tabular.

To clearify the problem: Here the result without the function relative:

tabular(Country ~ Type, data = cars)

        Type    
Country PKW  SUV
Germany 23    1 
Japan   22   17 
Other   12    0 
USA     58   22

Does anyone has an idea how to get the table?

Thanks in advance!

Upvotes: 1

Views: 864

Answers (1)

Diegoal
Diegoal

Reputation: 329

From the vignette:

The last factor, (mean + sd) names two R functions. These are assumed to be functions that operate on a vector and produce a single value, as mean and sd do. The values in the table will be the results of applying those functions to the two dierent variables and the subsets of the dataset.

I have overlooked this part that is the crucial hint, why the function i wrote doesnt work. "relative" returns a vector as long as manifestations the variable x has - allowed is an atomic vector. The problem is now that there is no way to add the basis into the function to calculate the relative values - the returning value is calculated from an subset of x.

Actually this is working but is not the most elegant way. I wrote the function with a hard coded nrow (here: 155).

relative <- function(x){

round(length(x)/155, 2)

}


tabular(Country ~ Type + relative, data = cars)

        Type             
 Country PKW  SUV relative
 Germany 23    1  0.15    
 Japan   22   17  0.25    
 Other   12    0  0.08    
 USA     58   22  0.52

If anyone find a better solution - please help :)

Upvotes: 2

Related Questions