Date:         Mon, 20 Oct 2003 10:35:36 -0400
Reply-To:     Sigurd Hermansen <HERMANS1@WESTAT.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Sigurd Hermansen <HERMANS1@WESTAT.COM>
Subject:      Re: How to convert convert character variables
Comments: To: "Groeneveld, Jim" <jim.groeneveld@VITATRON.COM>
Content-Type: text/plain

Jim:
We can build on your solution and answer another question as well. A SQL
solution handles the type conversion without explicit renaming (and, with
the LEFT() function, aligns the character values consistently):
proc sql;
  create table c as
  (select d1b1 as d1b1,d1b2 as d1b2,d1b3 as d1b3
   from a
   )
    outer union corr
  (select left(put(d1b1,3.)) as d1b1,
          left(put(d1b2,3.)) as d1b2,
          left(put(d1b3,3.)) as d1b3
   from b
   )
  ;
quit;

The SQL solution also answers Tine Andersen's question. Is there a way to
make SAS programs more aesthetically pleasing? Yes, use PROC SQL.  ;>  (I
wonder what it would take to convert SAS Data steps automatically, where
equivalent, to SQL queries?)
Sig

-----Original Message-----
From: Groeneveld, Jim [mailto:jim.groeneveld@VITATRON.COM]
Sent: Monday, October 20, 2003 6:04 AM
To: SAS-L@LISTSERV.UGA.EDU
Subject: Re: How to convert convert character variables


Hi Italian nameless Ebola,

Can't you be a little bit less anonymous? You do not need to be ashamed
asking questions like this. That is where the list and all of us are for.
Back to your problem:

Before getting that error report you get important notes as well, saying:

NOTE: Character values have been converted to numeric values at the places
given by:
      (Line):(Column).
      11:13
NOTE: Numeric values have been converted to character values at the places
given by:
      (Line):(Column).
      11:1
NOTE: Invalid numeric data, 'a' , at line 11 column 13.
NOTE: Invalid numeric data, 'b' , at line 11 column 13.
NOTE: Invalid numeric data, 'c' , at line 11 column 13.
d1b1=. d1b2=. d1b3=. k=4 _ERROR_=1 _N_=1
NOTE: Missing values were generated as a result of performing an operation
on missing values.
      Each place is given by: (Number of times) at (Line):(Column).
      3 at 11:21

From that you already could guess that it is impossible to convert a
variable's type. You would have to create new (numeric) variables (and
possible additional arrays). But it makes no sense to convert the character
a, b and c to numeric, so in order to obtain your frequency table you better
could create a character variable from you numeric ones:

data a;
  d1b1 = "a";
  d1b2 = "b";
  d1b3 = "c";
RUN;

data b;
  d1b1 = 1;
  d1b2 = 2;
  d1b3 = 3;
RUN;

DATA B (DROP=d1b1 d1b2 d1b3);
  SET B;
  LENGTH d0b1 - d0b3 $1;
  d0b1 = PUT (d1b1, 1.);
  d0b2 = PUT (d1b2, 1.);
  d0b3 = PUT (d1b3, 1.);
RUN;

data c;
  set a b (RENAME=(d0b1=d1b1 d0b2=d1b2 d0b3=d1b3));
RUN;

proc freq;
  table d1b1-d1b3;
run;

Regards - Jim.
--
.  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .

Jim Groeneveld, MSc.
Biostatistician
Science Team
Vitatron B.V.
Meander 1051
6825 MJ  Arnhem
Tel: +31/0 26 376 7365
Fax: +31/0 26 376 7305
Jim.Groeneveld@Vitatron.com
www.vitatron.com

I wish I had a webcam to show you the boring view from my window.

[common disclaimer]

-----Original Message-----
From: ebola [mailto:nameless20102003@YAHOO.IT]
Sent: Monday, October 20, 2003 09:46
To: SAS-L@LISTSERV.UGA.EDU
Subject: How to convert convert character variables


I am tring to transform character variables to numeric.
Here is a small exemple

data a;

d1b1 = "a";
d1b2 = "b";
d1b3 = "c";

/* here I tried the transformation */

array varc1 (*) d1b1-d1b3 ;
do     k  = 1 to dim (varc1);
varc1 (k) = varc1(k)+0 ;
/* or this
varc1 (k) = INPUT(varc1(k), 3.); */
end;

data b;

d1b1 = 1;
d1b2 = 2;
d1b3 = 3;

data c;
set a b;

proc freq;
   table

   d1b1-d1b3;

run;


and I get this Log message

336  data c;
337  set a b;
ERROR: Variable d1b1 has been defined as both character and numeric.
ERROR: Variable d1b2 has been defined as both character and numeric.
ERROR: Variable d1b3 has been defined as both character and numeric. 338

NOTE: The SAS System stopped processing this step because of errors.
WARNING: The data set WORK.C may be incomplete.  When this step was stopped
there were 0 observations and 4 variables.
WARNING: Data set WORK.C was not replaced because this step was stopped.
NOTE: DATA statement used:
      real time           0.00 seconds
      cpu time            0.00 seconds


Thankyou in advance for any help.