1

I am new to programming in R so I looking for help to start me off. I have a "long" dataset with 4 columns (which I reshaped earlier in R):

d=matrix(c(1,2,3,4,5,6,7,"A","A","A","B","B","B","B",0,0,99,0,0,99,0,99,0,99,99,0,0,99), ncol=4)
colnames(d) = c("ID","S.ID","ValA","ValB"); dt=as.table(d)

I'd like to create, for each S.ID separately (ie A and B), a contingency table for counts of ValA and ValB

For example (here) --

enter image description here

In my actual example I have 17 S.IDs and I'd like to write a program that will generate the contingency tables for each of these.

Additionally, Will this be easier to do in R Studio? If so, please can you also detail how?

Thank you very much!

-Thanks to all who answered so prompltly! :)

EskCargo
  • 35
  • 5
  • Possible duplicate of http://stackoverflow.com/questions/13841599/calculate-frequency-of-occurrence-in-an-array-using-r – akrun Feb 15 '17 at 12:00

3 Answers3

4

You can use table on multiple variables:

with(as.data.frame(d), table(ValA, ValB, S.ID))

# , , S.ID = A
#    ValB
#ValA 0 99
#  0  1  1
#  99 0  1

#, , S.ID = B
#    ValB
#ValA 0 99
#  0  1  2
#  99 1  0
Cath
  • 23,906
  • 5
  • 52
  • 86
1

Using tapply or by needs conversion to a data.frame: 

d<-data.frame(d)
by(d, d$S.ID, function(grouped.d) table(grouped.d$ValA, grouped.d$ValB))
Bernhard
  • 4,272
  • 1
  • 13
  • 23
0

You can try this: 

d <- as.data.frame(d)
l <- lapply(unique(d$S.ID), function(x) xtabs(~ValB+ValA, subset(d, S.ID==x)))
names(l) <- unique(d$S.ID)
l
#$A
#    ValA
#ValB 0 99
#  0  1  0
#  99 1  1

#$B
#    ValA
#ValB 0 99
#  0  1  1
#  99 2  0
Sandipan Dey
  • 21,482
  • 2
  • 51
  • 63