Richard Williams replied:
> Yes, but you are using a dinky little 2 X 2 table whereas Joseph's table
> is
> 7 X 2. There is a reason that even when you specify -all- as an option on
> -tabulate- you still have to specify -exact- separately -- the Fisher
> Exact
> Test can take a long time even for a computer. I'll try Joseph's problem
> out on my machine and see if I fair any better.
Fair enough, but running a larger table on the -union- data still poses no
problem on my machine (save the (minor?) error registered at the end):
. webuse union
. tab year union, exact all
interview | 1 if union
year | 0 1 | Total
-----------+----------------------+----------
70 | 1,310 348 | 1,658
71 | 1,412 381 | 1,793
72 | 1,576 382 | 1,958
73 | 1,633 424 | 2,057
77 | 2,180 490 | 2,670
78 | 1,530 463 | 1,993
80 | 1,432 596 | 2,028
82 | 1,846 562 | 2,408
83 | 1,733 461 | 2,194
85 | 1,888 600 | 2,488
87 | 1,981 555 | 2,536
88 | 1,868 549 | 2,417
-----------+----------------------+----------
Total | 20,389 5,811 | 26,200
Pearson chi2(11) = 107.8144 Pr = 0.000
likelihood-ratio chi2(11) = 105.0420 Pr = 0.000
Cram�r's V = 0.0641
gamma = 0.0342 ASE = 0.009
Kendall's tau-b = 0.0192 ASE = 0.005
integer overflow due to large row margin frequencies
r(1401);
Using Dan Blanchette's -etime- from SSC, I approximately timed the command
to take no more than 2 seconds: and my machine is considerably slower than
Joseph's.
CLIVE NICHOLAS |t: 0(044)7903 397793
Politics |e: [email protected]
Newcastle University |http://www.ncl.ac.uk/geps
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
