Date: Fri, 8 Mar 2002 12:44:11 -0500
Reply-To: howard_schreier@ITA.DOC.GOV
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: howard_schreier@ITA.DOC.GOV
Subject: Re: help: cocatenatign unknown SAS data sets
The difficulty is because you coded the option
(keep=a b)
Get rid of that, or else add the BY variables to the list.
People on this list often provide high-quality solutions to sometimes
difficult problems. But one should not expect everything to be 100% tested
and debugged.
On Fri, 8 Mar 2002 10:42:47 -0500, P. L <li_9@HOTMAIL.COM> wrote:
>Adding By doesn't help. The data b has only regression variables, and has
>no the By variables used in the data a.
>
>
>On Thu, 7 Mar 2002 23:19:04 -0500, howard_schreier@ITA.DOC.GOV wrote:
>
>>A MERGE with no BY is usually suspicious.
>>
>>Try copying the BY statement used in the PROC REG step and pasting it
after
>>the MERGE statement.
>>
>>On Thu, 7 Mar 2002 16:56:18 -0500, P. L <li_9@HOTMAIL.COM> wrote:
>>
>>>>
>>>>%macro reg (year);
>>>>
>>>>data a;
>>>> set income.earning&year;
>>>>if tyr ne 98;
>>>>run;
>>>>
>>>>proc reg data = a outest =b (rename=(intercept=a yrcan=b));
>>>> by lage sex big5 degree lang warea;
>>>> weight numwith;
>>>> model diffcon = yrcan;
>>>>run;
>>>>
>>>>data table&YEAR;
>>>> merge a b (keep=a b);
>>>> c = abs(a/b);
>>>> if c = . then c = 0;
>>>>run;
>>>>
>>>>%mend;
>>>>
>>>>Kind regards,
>>>>
>>>>Dave
>>>>
>>>>.
>>>
>>>Dave,
>>>
>>>Thanks for your message. I've tried your code and it is indeed simpler
and
>>>much faster than what I did before. However, I have some trouble with
your
>>>program. As the following samples show, what I need is the same a b, c
>>>values for observations with the same combination. But, using your code
>>>produces diffirent values for observations of the same combination. I'm
>not
>>>sure why happens. If you would provide suggestion to fix the problem, it
>>>will be greatly appreciated.
>>>
>>>P. L
>>>
>>>THIS IS WHAT I
>>>WANT:
>>>
>>> Obs lyr tyr a b c
>>>
>>> 1 80 82 -23549.42 1666.51 14.1310
>>> 2 80 83 -23549.42 1666.51 14.1310
>>> 3 80 84 -23549.42 1666.51 14.1310
>>> 4 80 85 -23549.42 1666.51 14.1310
>>> 5 80 86 -23549.42 1666.51 14.1310
>>> 6 80 87 -23549.42 1666.51 14.1310
>>> 7 80 88 -23549.42 1666.51 14.1310
>>> 8 80 89 -23549.42 1666.51 14.1310
>>> 9 80 90 -23549.42 1666.51 14.1310
>>> 10 80 91 -23549.42 1666.51 14.1310
>>> 11 80 92 -23549.42 1666.51 14.1310
>>> 12 80 93 -23549.42 1666.51 14.1310
>>> 13 80 94 -23549.42 1666.51 14.1310
>>> 14 80 95 -23549.42 1666.51 14.1310
>>> 15 80 96 -23549.42 1666.51 14.1310
>>>
>>>THIS IS WHAT YOUR CODE PRODUCED:
>>>
>>> Obs lyr tyr a b c
>>>
>>> 1 80 82 -23522.68 1662.69 14.147
>>> 2 80 83 -47216.20 1132.72 41.684
>>> 3 80 84 -26497.64 0.00 0.000
>>> 4 80 85 -35823.65 1687.36 21.231
>>> 5 80 86 -44633.68 2609.90 17.102
>>> 6 80 87 -27862.22 147.74 188.590
>>> 7 80 88 -30085.42 331.21 90.835
>>> 8 80 89 -15052.34 -17.30 870.111
>>> 9 80 90 -19058.76 266.78 71.441
>>> 10 80 91 -16583.35 -288.29 57.523
>>> 11 80 92 -23158.46 39.82 581.556
>>> 12 80 93 -20230.82 -140.87 143.612
>>> 13 80 94 -21842.69 117.37 186.109
>>> 14 80 95 -23397.47 165.95 140.993
>>> 15 80 96 -19256.60 147.53 130.524
| 
