Thank you all for your inputs:

joe jim gerhard nat

tom

-----Original Message----- From: Joe Matise <snoopy369@GMAIL.COM> To: SAS-L@LISTSERV.UGA.EDU Sent: Thu, Feb 18, 2010 11:36 am Subject: Re: Convert Character into numeric

BEST. is really BEST12., if I remember correctly. -Joe On Thu, Feb 18, 2010 at 11:33 AM, Gerhard Hellriegel < erhard.hellriegel@t-online.de> wrote: > seems that there is a default for best. I get the following with very long numbers:

23 data _null_; 24 a_c = "12345678953535350"; 25 a_n = input(a_c,best.); 26 put a_c=; 27 put a_n=; 28 run;

a_c=12345678953535350 a_n=123456789535

anyway tehre is no good way to use very long numbers as keys, e.g as a accounting number:

data _null_; a_c = "1234567890123456789"; a_n = input(a_c,32.); put a_c; put a_n 19.; run;

the result is:

1234567890123456789 1234567890123456768

That is simply a problem of the number of significant digits in a 8-byte numeric variable. It is too long! If it is necessary to have longer account_numbers, you must use a character variable for that. In that case it might be also a better way to make it shorter and use a, b, c, ... for making it unique. Or you start with numerics (it's easier to generate a unique key by simply adding 1....) and hope that until you reach the limit there is a SI- solution for the "short" numbers, e.g. a 16- or 32-byte representation of numbers. Think 16 byte would be very much, 1 byte for exponent, 15*8-1 bit for the number = 2**120-1 as biggest number ... SAS says that this is something around 1.3e36 so I think 35 digits should be reliable then. Ok, to be more sure, perhaps 32 byte, 2 for exponent, 30 for mantissa (30*8-1 means 1.7e72, so 70 digits should be ok, enough for the next few years to contain the national dept of Germany - not sure about that...)

Gerhard

On Thu, 18 Feb 2010 12:03:19 -0500, Nathaniel Wooding <nathaniel.wooding@DOM.COM> wrote:

>I agree with Gerhard about avoiding short informats. A further solution would be to simply use the Best. Informat with no length specified. > > >Data t; >cnum = ' 12345678'; >num = input( cnum , best. ); >run; > > >Nat Wooding > >-----Original Message----- >From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU] On Behalf Of Tom White >Sent: Thursday, February 18, 2010 10:52 AM >To: SAS-L@LISTSERV.UGA.EDU >Subject: Convert Character into numeric > >Hello SAS-L, > >I am strugling with converting an ACCOUNT_NUMBER variable with format and informat $9. >Wjen I count the length of this variable it varies from 4 digits to 9 digits. >Yet, when I look at a 4-digit account_number [L=length(account_number); L=4] >the actual account_number has only 3 digits. Similarly for the 9-digit. The actual account_number >has only 8 digits. > >This leads me to believe that maybe we have a trailing zero in account_number. > >I try to conver the account_number (say, account_number=12345678 which has length=9 and not 8 >according to above remark). > >ACCT_NBR=input(account_number, best.) > >This doesn't convert the account numbers. >Instead, I get . for all of them except for the last observation in the account_umber (wired ???) > >when I try > >ACCT_NBR=input(account_number, 9.) [The maximum lenght L=9 per note above] >I get missing account numbers again (.) except for the last obs in account_number variable (wired ???) > >when I try > >ACCT_NBR=input(account_number, 8.) the accounts are converted but not all. > >SAS picks and chooses which ones to convert ! > >It appears though that the 8-digit acount numbers (L=9) are being converted. > >Those account numbers with fewer than 8-digits, some are convertd, som are not convertd. > >What's going on? > >tom > > > > > > > > > > > > > > > >= >CONFIDENTIALITY NOTICE: This electronic message contains >information which may be legally confidential and or privileged and >does not in any case represent a firm ENERGY COMMODITY bid or offer >relating thereto which binds the sender without an additional >express written confirmation to that effect. The information is >intended solely for the individual or entity named above and access >by anyone else is unauthorized. If you are not the intended >recipient, any disclosure, copying, distribution, or use of the >contents of this information is prohibited and may be unlawful. If >you have received this electronic transmission in error, please >reply immediately to the sender that you have received the message >in error, and delete it. Thank you.

-----Original Message----- From: Joe Matise <snoopy369@GMAIL.COM> To: SAS-L@LISTSERV.UGA.EDU Sent: Thu, Feb 18, 2010 11:36 am Subject: Re: Convert Character into numeric

BEST. is really BEST12., if I remember correctly. -Joe On Thu, Feb 18, 2010 at 11:33 AM, Gerhard Hellriegel < erhard.hellriegel@t-online.de> wrote: > seems that there is a default for best. I get the following with very long numbers:

23 data _null_; 24 a_c = "12345678953535350"; 25 a_n = input(a_c,best.); 26 put a_c=; 27 put a_n=; 28 run;

a_c=12345678953535350 a_n=123456789535

anyway tehre is no good way to use very long numbers as keys, e.g as a accounting number:

data _null_; a_c = "1234567890123456789"; a_n = input(a_c,32.); put a_c; put a_n 19.; run;

the result is:

1234567890123456789 1234567890123456768

That is simply a problem of the number of significant digits in a 8-byte numeric variable. It is too long! If it is necessary to have longer account_numbers, you must use a character variable for that. In that case it might be also a better way to make it shorter and use a, b, c, ... for making it unique. Or you start with numerics (it's easier to generate a unique key by simply adding 1....) and hope that until you reach the limit there is a SI- solution for the "short" numbers, e.g. a 16- or 32-byte representation of numbers. Think 16 byte would be very much, 1 byte for exponent, 15*8-1 bit for the number = 2**120-1 as biggest number ... SAS says that this is something around 1.3e36 so I think 35 digits should be reliable then. Ok, to be more sure, perhaps 32 byte, 2 for exponent, 30 for mantissa (30*8-1 means 1.7e72, so 70 digits should be ok, enough for the next few years to contain the national dept of Germany - not sure about that...)

Gerhard

On Thu, 18 Feb 2010 12:03:19 -0500, Nathaniel Wooding <nathaniel.wooding@DOM.COM> wrote:

>I agree with Gerhard about avoiding short informats. A further solution would be to simply use the Best. Informat with no length specified. > > >Data t; >cnum = ' 12345678'; >num = input( cnum , best. ); >run; > > >Nat Wooding > >-----Original Message----- >From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU] On Behalf Of Tom White >Sent: Thursday, February 18, 2010 10:52 AM >To: SAS-L@LISTSERV.UGA.EDU >Subject: Convert Character into numeric > >Hello SAS-L, > >I am strugling with converting an ACCOUNT_NUMBER variable with format and informat $9. >Wjen I count the length of this variable it varies from 4 digits to 9 digits. >Yet, when I look at a 4-digit account_number [L=length(account_number); L=4] >the actual account_number has only 3 digits. Similarly for the 9-digit. The actual account_number >has only 8 digits. > >This leads me to believe that maybe we have a trailing zero in account_number. > >I try to conver the account_number (say, account_number=12345678 which has length=9 and not 8 >according to above remark). > >ACCT_NBR=input(account_number, best.) > >This doesn't convert the account numbers. >Instead, I get . for all of them except for the last observation in the account_umber (wired ???) > >when I try > >ACCT_NBR=input(account_number, 9.) [The maximum lenght L=9 per note above] >I get missing account numbers again (.) except for the last obs in account_number variable (wired ???) > >when I try > >ACCT_NBR=input(account_number, 8.) the accounts are converted but not all. > >SAS picks and chooses which ones to convert ! > >It appears though that the 8-digit acount numbers (L=9) are being converted. > >Those account numbers with fewer than 8-digits, some are convertd, som are not convertd. > >What's going on? > >tom > > > > > > > > > > > > > > > >= >CONFIDENTIALITY NOTICE: This electronic message contains >information which may be legally confidential and or privileged and >does not in any case represent a firm ENERGY COMMODITY bid or offer >relating thereto which binds the sender without an additional >express written confirmation to that effect. The information is >intended solely for the individual or entity named above and access >by anyone else is unauthorized. If you are not the intended >recipient, any disclosure, copying, distribution, or use of the >contents of this information is prohibited and may be unlawful. If >you have received this electronic transmission in error, please >reply immediately to the sender that you have received the message >in error, and delete it. Thank you.