Skip to content

relationship between two factors

7 messages · christopher snow, Greg Snow, Marc Schwartz +3 more

Original
christopher snow
# 
I have a dataset with two variables that are factors:

1) Decision Making Satisfaction (DMS), values = A - Completely, B - 
Mostly, C - Partly, D - Not at all
2) IT Satisfaction values (ITS), values = A - Completely, B - Mostly, C 
- Partly, D - Not at all

I would like to produce a table (matrix) and a chart of the factors, 
with counts at the cross sections:

         A   B   C   D
A   
B       counts
C
D

How can I do this in R?

Many thanks,

Chris
Greg Snow
# 
The 'table' function will give you the simple counts.  Plotting a table
with the 'plot' function gives common charts for this.  The 'CrossTable'
function in the gmodels package creates the table along with additional
information.  There are a lot of other functions for creating/working
with tables depending on what you are trying to do.

Hope this helps,
Marc Schwartz
#
On Thu, 2007-12-06 at 12:51 +0000, christopher snow wrote:
See ?table, for example:

  table(DMS, ITS)

You did not indicate the type of chart you want to create, but some
possibilities, using base graphics, would be barplot(), dotchart() and
mosaicplot(). The latter is in the vcd package on CRAN.

HTH,

Marc Schwartz
John Kane
# 
?table should work 

table (DMS, ITS)

I am not clear on what kind of chart you want.

will 
plot(DMS, ITS) do what you want?
--- christopher snow <snowch at coralms.com> wrote:

            

      
      
      
    

        
Looking for the perfect gift? Give the gift of Flickr!

  
  


  


      

    

    

      
      

        


  

        

      Henrique Dallazuanna
    

    

      
      #
    

  


  

      
Try this:

df <- data.frame(DMS=factor(rep(LETTERS[1:4], 10)),
ITS=factor(rep(LETTERS[1:4], 10)))
table(df)
plot(table(df))

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      David Winsemius
    

    

      
      #
    

  


  

      
"Henrique Dallazuanna" <wwwhsd at gmail.com> wrote in 
news:da79af330712061144m738a9bd6ob8ecb2e89e8ff2c3 at mail.gmail.com:

            

      
      
      
    

        
That gave me simply 10's on the diagonals of the table and the plot was 
not very satisfying. I suggest this alternate example:

tbl<-r2dtable(1,c=4*c(15,12,8,5),r=4*c(20,10,5,5))
tbl
#[[1]]
#     [,1] [,2] [,3] [,4]
#[1,]   27   23   21    9
#[2,]   18   11    5    6
#[3,]    8    6    2    4
#[4,]    7    8    4    1

# > class(tbl)
# [1] "list"
ITS=factor(rep(LETTERS[1:4]))
DMS=factor(rep(LETTERS[1:4]))
df<-data.frame(counts=unlist(tbl),expand.grid(DMS,ITS))
names(df)<-c("counts","DMS",   "ITS")
df
(df.tbl<-xtabs(counts~DMS+ITS,df))
#       ITS
#DMS  A  B  C  D
#      A 27 23 21  9
#      B 18 11  5  6
#      C  8  6  2  4
#      D  7  8  4  1
plot(df.tbl)

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      christopher snow
    

    

      
      #
    

  


  

      
Thanks to everyone for their answers. What a helpful community!

        
christopher snow wrote: