Date:         Thu, 30 Nov 2006 15:12:50 -0800
Reply-To:     rss <rslotpole@FMR.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         rss <rslotpole@FMR.COM>
Organization: http://groups.google.com
Subject:      Re: basic syntex for Proc Tabulate
Comments: To: sas-l@uga.edu
In-Reply-To:  <445d9dbe0611301259w70ee11acq32e3361c93066a96@mail.gmail.com>
Content-Type: text/plain; charset="us-ascii"

Yu,

Here's a portion of my data set:  Dataset name a33

yy      mm      N       CORR
1997    11/1/1997       2       1
1997    12/1/1997       3       -0.5
1998    1/1/1998        3       -0.5
1998    2/1/1998        15      0.25483871
1998    3/1/1998        20      0.477099237
1998    4/1/1998        33      0.336905966
1998    5/1/1998        41      0.601691042
1998    6/1/1998        31      0.36128692
1998    7/1/1998        33      0.390909091
1998    8/1/1998        27      -0.071959755
1998    9/1/1998        24      -0.286901271
1998    10/1/1998       24      0.292759706
1998    11/1/1998       67      0.55485575
1998    12/1/1998       79      0.10101476
1999    1/1/1999        91      0.103458364
1999    2/1/1999        127     0.125940977
1999    3/1/1999        153     0.051398853
1999    4/1/1999        219     0.142683906
1999    5/1/1999        267     0.028117021
1999    6/1/1999        288     0.080195742
1999    7/1/1999        311     0.120760693
1999    8/1/1999        331     0.152256117
1999    9/1/1999        342     0.114990615
1999    10/1/1999       366     0.137903518
1999    11/1/1999       397     0.141527962
1999    12/1/1999       407     0.203815996
2000    1/1/2000        406     0.140783109
2000    2/1/2000        402     0.119713309
2000    3/1/2000        441     0.186155764
2000    4/1/2000        460     0.130206348
2000    5/1/2000        506     0.10201116
2000    6/1/2000        555     0.1818349
2000    7/1/2000        590     0.179158797
2000    8/1/2000        607     0.184940269

Now, here is some code that almost does everything I need:

proc tabulate data = a33 noseps order=data;
        class yy mm;
        var corr ;

        table yy=''all='AVG',(mm='information Coefficients'
all='AVG')*corr=''*mean=''  *format=5.3/rts= 8 box='YEAR';
run;


Here are the three problems I'm still having:

1) My data starts out in Nov 1997.  The table prints out but the months
start in Nov instead of Jan.  I tried to run the following:

data a34;
set a33;
do yy=1997 to 2006 by 1;
do mm = 1 to 12 by 1;
    abc = ranuni(1234);
    output;
end;
end;
run;


but that create a file of 13080 records instead of the 120.  mm is a
numeric.  Even if this worked how would i relabel the variables 1
through 12.  anyway, all sorts of problems trying to get this to come
up Jan through Dec.

2) the print out very nicely places a column line for the last column.
Is there a way to suppress the lines for rows  (using noseps) but have
1 line just above the last row so we can offset and Highlight the row
of averages.  Also, could we change the font or color for observations
for some cells and not others?  How would we reference a particular
number?  would we use yy and mm?

3) The title "YEAR" in the top left hand corner.  Is there a way to
print this in the middle of the box (in terms of vertical)?

Nailing those three items would just about wrap this up. Thanks a lot
for everyone for their help.  Slowly but surly we are making progress.