| Date: | Thu, 20 Apr 2000 18:15:14 GMT |
|---|
| Reply-To: | haworthla@MY-DEJA.COM |
|---|
| Sender: | "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU> |
|---|
| From: | haworthla@MY-DEJA.COM |
|---|
| Organization: | Deja.com - Before you buy. |
|---|
| Subject: | Re: ODS HTML Body files in TABULATE with BY and PAGE |
|---|
Ray,
I think your solution lies in a combination of the responses to your
original message.
If you use your %BY code to generate the BY group macro variables,
Ian's suggestion to use PROC SUMMARY first to slim down the data, and
my ODS macro to create the output, you should get what you wanted with
reasonable efficiency.
... Either that or beg a copy of 8.1!
-- Lauren
In article <4.2.2.20000420102020.00bd8f00@postoffice.worldnet.att.net>,
Ray Pass <raypass@WORLDNET.ATT.NET> wrote:
> --=====================_4890422==_.ALT
> Content-Type: text/plain; charset="us-ascii"; format=flowed
>
> Lauren,
>
> At 09:13 PM 4/19/00 -0700, Lauren E Haworth wrote:
> >Ray,
> >
> >How delightful, a problem that combines my two favorite topics,
TABULATE
> >and ODS.
>
> two of my favorites as well.
>
> >I'd tackle this by creating a macro to run the table for each value
of
> >the BY variable. You said you didn't want to run separate TABULATEs,
but
> >I don't see any way around that. The only difference between this
code
> >and a single TABULATE version would be the handling of missing data
> >across the tables. You might need to do some DATA step work if this
is an
> >issue. Was there some other reason you didn't want to run multiple
> >TABULATEs?
>
> The reason that I did not want to take this approach is that I'm
dealing
> with LARGE files, many TABULATES and a multi-value BY variable. I
wanted
> to keep it in the realm of BY group processing without having to
start all
> those TABULATEs with WHERE statements and read through the entire
data set
> so many times. BTW, check out my SUGI %BY Coders Corner paper for a
> data-driven method of doing what you propose. My method does not
require
> you to know a priori what the levels of the bygroup are.
>
> On another note, one of my flock tells me that V8.1 will have a
working ODS
> HTML NEWFILE=BYGROUP option to do exactly what I need. It's in V8.0,
but
> isn't working properly.
>
> Thanks for the response. Now back to work :-)
>
> Ray
>
> *------------------------------------------------*
> | Ray Pass voice: (914) 693-5553 |
> | Ray Pass Consulting eFax: (914) 206-3780 |
> | 5 Sinclair Place |
> | Hartsdale, NY 10530 e-mail: raypass@att.net |
> *------------------------------------------------*
>
> --=====================_4890422==_.ALT
> Content-Type: text/html; charset="us-ascii"
>
> <html>
> Lauren,<br>
> <br>
> At 09:13 PM 4/19/00 -0700, Lauren E Haworth wrote:<br>
> <blockquote type=cite cite>Ray,<br>
> <br>
> How delightful, a problem that combines my two favorite topics,
> TABULATE<br>
> and ODS.</blockquote><br>
> two of my favorites as well.<br>
> <br>
> <br>
> <blockquote type=cite cite>I'd tackle this by creating a macro to run
the
> table for each value of<br>
> the BY variable. You said you didn't want to run separate TABULATEs,
> but<br>
> I don't see any way around that. The only difference between this
> code<br>
> and a single TABULATE version would be the handling of missing
data<br>
> across the tables. You might need to do some DATA step work if this is
> an<br>
> issue. Was there some other reason you didn't want to run multiple<br>
> TABULATEs?</blockquote><br>
> The reason that I did not want to take this approach is that I'm
dealing
> with LARGE files, many TABULATES and a multi-value BY variable.
I
> wanted to keep it in the realm of BY group processing without having
to
> start all those TABULATEs with WHERE statements and read through the
> entire data set so many times. BTW, check out my SUGI %BY Coders
> Corner paper for a data-driven method of doing what you
propose. My
> method does not require you to know a priori what the levels of the
> bygroup are.<br>
> <br>
> On another note, one of my flock tells me that V8.1 will have a
working
> ODS HTML NEWFILE=BYGROUP option to do exactly what I need. It's
in
> V8.0, but isn't working properly. <br>
> <br>
> Thanks for the response. Now back to work :-)<br>
> <br>
> Ray<br>
> <br>
> <br>
> <div>*------------------------------------------------*</div>
> <div>| Ray
>
Pass &n
bsp;
> voice: (914) 693-5553 |</div>
> <div>| Ray Pass Consulting eFax: (914)
> 206-3780 |</div>
> <div>| 5 Sinclair
>
Place &
nbsp; &
nbsp;
> |</div>
> <div>| Hartsdale, NY 10530 e-mail: raypass@att.net
> |</div>
> <div>*------------------------------------------------* </div>
> </html>
>
> --=====================_4890422==_.ALT--
>
Sent via Deja.com http://www.deja.com/
Before you buy.
|
