[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: RE: table of counts of values for N variables
Program -tabm- from -tab_chi- on SSC has this aim.
Row percents come for free, as all -tabm- does is
restructure the data on the fly and then call up
-tab-. That is an upstream solution, rather than
your downstream strategy of picking up output and
> I have about 30 variables that take on integer values from 1 through
> 6. I would like to create a table with the rows being the variables,
> the columns consisting of values 1/6, and the cells consisting of the
> count of the variable equal to that value. There may be missings,
> which I would like to include. Since it would include values that
> would otherwise be left out by -tab-, this is similar to -tabcount-,
> but vertically. I don't need to be able to specify the value range
> (the numlist in tabcount), since they are cleaned within the range.
> vertab var1-var3, missing
> 1 2 3 4 5 6 . N
> var1 2 4 5 2 1 4 0 18
> var2 7 4 2 2 0 3 0 18
> var3 9 8 0 0 0 0 1 18
> The ability to have row percents would also be fantastic.
> I have looked into using -tab, matcell()- but am not sure how to
> recombine the matrices once they have been stored in a manner which
> would be consistent across values (i.e., if var3 has no 3's, then the
> third value is actually for value 4).
> I feel like I am missing some obvious command, but am not sure what.
* For searches and help try: