Skip to content

ftable and data.frame

4 messages · Silvano, arun, John Kane +1 more

arun
# 
Hi,
Try:
library(reshape2)
dcast(as.data.frame(tab1), SEX+ESTCIV~Q1,value.var="Freq") ##not tested.

A.K.
On Friday, December 20, 2013 8:03 AM, "silvano at uel.br" <silvano at uel.br> wrote:
Hi,

I used this command to produce a table:

(tab1 = ftable(SEX, ESTCIV, Q1))

? ? ? ? ? ? ? ? ? ? ? ? ? ? ?  Q1? B? L? M? N
SEXO? ? ESTCIV
? ? ? ? F? ? ? ? ?  A? ? ? ?  11 13? 4? 2
? ? ? ? ? ? ? ? ? ? ? E? ? ? ? ? 1? 0? 0? 0
? ? ? ? M? ? ? ? ?  A? ? ? ? ? 5? 0? 3? 1
? ? ? ? ? ? ? ? ? ? ? E? ? ? ? ? 0? 0? 0? 0

but I need something like:


SEXO? ? ESTCIV? ? ? ? B? L? M? N
? ? ? F? ? ? ? ? ? A? ? ? ?  11 13? 4? 2
? ? ? F? ? ? ? ? ? E? ? ? ? ? 1? 0? 0? 0
? ? ? M? ? ? ? ?  A? ? ? ? ? 5? 0? 3? 1
? ? ? M? ? ? ? ?  E? ? ? ? ? 0? 0? 0? 0

How can I get it?

I need this format to use ordinal logistic regression and I have many tables.

Thanks,

Silvano.

---
Este email est? limpo de v?rus e malwares porque a prote??o do avast! Antiv?rus est? ativa.


??? [[alternative HTML version deleted]]

______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
John Kane
# 
What does your original data look like?  I seems to me that it would be better to give us that information since you may not want to use ftable() at all.
Ideally you should give us the original data or a sample of it. If it's confidential just replace the actual values with fake data.  The structure of the data is more important than the values.

Please use dput() (see ?dput for information) to supply the data.  Just do :  dput(mydata) and then paste the result into the email.

John Kane
Kingston ON Canada
____________________________________________________________
FREE 3D MARINE AQUARIUM SCREENSAVER - Watch dolphins, sharks & orcas on your desktop!
David Winsemius
#
On Dec 20, 2013, at 5:01 AM, <silvano at uel.br> wrote:

            

      
      
      
    

        
It's not actually a 'table'.

            

      
      
      
    

        
is.table(tab1) # will return FALSE

            

      
      
      
    

        
You need to describe the purpose of this effort. 

If is for display, the answer will be to look at the code for `print.ftable` (which eventually passes its arguments to `format.ftable`.

If it is for creating something other than an ftable (such a data.frame or character-matrix), then you will probably need to coerce the ftable to a matrix to get the entry values and then extract the row labels from the 'tab1'-object with:   attr(tab1, "row.vars") and `rep` them approriately. Or you could work with the output from format(ftable(tab1))[ ,1:2].

cbind( format(ftable(Titanic), quote=FALSE)[ ,1:3], 
       format(ftable(Titanic), quote=FALSE)[ ,5:6])
      [,1]    [,2]     [,3]    [,4]  [,5] 
 [1,] "     " "      " "     " " No" "Yes"
 [2,] "Class" "Sex   " "Age  " "   " "   "
 [3,] "1st  " "Male  " "Child" "  0" "  5"
 [4,] "     " "      " "Adult" "118" " 57"
 [5,] "     " "Female" "Child" "  0" "  1"
 [6,] "     " "      " "Adult" "  4" "140"
 [7,] "2nd  " "Male  " "Child" "  0" " 11"
 [8,] "     " "      " "Adult" "154" " 14"
 [9,] "     " "Female" "Child" "  0" " 13"
[10,] "     " "      " "Adult" " 13" " 80"
[11,] "3rd  " "Male  " "Child" " 35" " 13"
[12,] "     " "      " "Adult" "387" " 75"
[13,] "     " "Female" "Child" " 17" " 14"
[14,] "     " "      " "Adult" " 89" " 76"
[15,] "Crew " "Male  " "Child" "  0" "  0"
[16,] "     " "      " "Adult" "670" "192"
[17,] "     " "Female" "Child" "  0" "  0"
[18,] "     " "      " "Adult" "  3" " 20"

            

      
      
      
    

        
That makes me think using ftable is the completely misguided approach. You should describe in more detail you plans to use ordinal logistic regression. I'm also wondering if you have used attach(). It would make more sense to keep these variables in a dataframe. Almost all regression functions expect input in the form of a normalized dataframe rather than in separate variables of aggregated counts.

            

      
      
      
    

        

      
      
    

        
David Winsemius
Alameda, CA, USA