SPSSX Discussion

Error # 4651

Classic

List

Threaded

11 messages Options

John F Hall

Error # 4651

This error message threw me until I spotted the 8–88 instead of 8=88. Easy mistake to make, but must be common enough for SPSS to spot as well..

recode v1752 (8-88) (8=99)(1 = 14)

/ v1745 to v1751 (1=7)(2=8)(3=9)(4=10)(5=11)(6=12)(7=13).

Error # 4651 in column 19. Text: )

The symbol ')' was used on the RECODE command incorrectly. This symbol must

follow either the replacement value on the right of the equals sign or the

keyword CONVERT.

Execution of this command stops.

John F Hall (Mr)

[Retired academic survey researcher]

Email: [hidden email]

Website: www.surveyresearch.weebly.com

SPSS start page: www.surveyresearch.weebly.com/1-survey-analysis-workshop

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

John F Hall

Re: Error # 4651

Looks like I tried to recode the same value to two different values as well!

From: John F Hall [mailto:[hidden email]]
Sent: 04 March 2016 17:39
To: '[hidden email]' <[hidden email]>
Subject: Error # 4651

This error message threw me until I spotted the 8–88 instead of 8=88. Easy mistake to make, but must be common enough for SPSS to spot as well..

recode v1752 (8-88) (8=99)(1 = 14)

/ v1745 to v1751 (1=7)(2=8)(3=9)(4=10)(5=11)(6=12)(7=13).

Error # 4651 in column 19. Text: )

The symbol ')' was used on the RECODE command incorrectly. This symbol must

follow either the replacement value on the right of the equals sign or the

keyword CONVERT.

Execution of this command stops.

John F Hall (Mr)

[Retired academic survey researcher]

Email: [hidden email]

Website: www.surveyresearch.weebly.com

SPSS start page: www.surveyresearch.weebly.com/1-survey-analysis-workshop

Bruce Weaver

Re: Error # 4651

Administrator

In reply to this post by John F Hall

Hi John. You already commented in another post about trying to recode 8 to two different values. ;-)

What struck me was that the error message you posted appears to be complaining about the missing equals sign, but not about the '-88'. That made me wonder if SPSS is actually doing the subtraction. As this is an empirical question, I gave it a try, as follows:

DATA LIST free / Old (F5.0).
BEGIN DATA
8 9 -80 88 1 2 3 4 5
END DATA.

RECODE Old (8-88=88) (9=99)(1 = 14) INTO New1.
RECODE Old (8-88=88) (9=99)(1 = 14) (ELSE = COPY) INTO New2.
RECODE Old (8 THRU 88=88) (9=99)(1 = 14) INTO New3.
FORMATS New1 to New3 (F5.0).
LIST.

OUTPUT:

Old New1 New2 New3

8 88 88 88
9 99 99 88
-80 . -80 .
88 . 88 88
1 14 14 14
2 . 2 .
3 . 3 .
4 . 4 .
5 . 5 .

CONCLUSIONS:

1. -80 in Old is not being recoded to 88 in New1, so SPSS is not doing the subtraction in the RECODE command.

2. SPSS is apparently not bothered by the extra '-88' in (8-88=88).

3. 8-88 is not being interpreted as 8 THRU 88. It appears to be treated the same as (8=88).

HTH.

John F Hall wrote

This error message threw me until I spotted the 8-88 instead of 8=88. Easy
mistake to make, but must be common enough for SPSS to spot as well..

recode v1752 (8-88) (8=99)(1 = 14)
/ v1745 to v1751 (1=7)(2=8)(3=9)(4=10)(5=11)(6=12)(7=13).

Error # 4651 in column 19. Text: )
The symbol ')' was used on the RECODE command incorrectly. This symbol must
follow either the replacement value on the right of the equals sign or the
keyword CONVERT.
Execution of this command stops.

John F Hall (Mr)
[Retired academic survey researcher]

Email: <mailto:[hidden email]> [hidden email]
Website: <http://www.surveyresearch.weebly.com/>
www.surveyresearch.weebly.com
SPSS start page:
<http://surveyresearch.weebly.com/1-survey-analysis-workshop.html>
www.surveyresearch.weebly.com/1-survey-analysis-workshop

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

--
Bruce Weaver
bweaver@lakeheadu.ca
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

PLEASE NOTE THE FOLLOWING:
1. My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above.
2. The SPSSX Discussion forum on Nabble is no longer linked to the SPSSX-L listserv administered by UGA (https://listserv.uga.edu/).

John F Hall

Re: Error # 4651

Bruce

The correct syntax should have been:

recode v1752 (8=88) (9=99)(1 = 14)
/ v1745 to v1751 (1=7)(2=8)(3=9)(4=10)(5=11)(6=12)(7=13).

This was part of a complex recode needed to cope with a situation where
multiple response questions had more categories than codes. For instance,
one question asking about benefits received had fourteen possible responses
coded separately using only codes 1-6 (twice) in columns 17/39 to 17/51
plus 0, 7, 8 and 9 17/52, (one for each code) except column 17/52 which
carried codes for 0 (None) 7 (Other) 8 (DK) and 9 (Not applicable). Codes 8
and 9 were already declared as missing, so the recode of the second set of
values from 6-12 to 7 - 13 can't be done before changing 8 and 9 to 88 and
99.

I am writing a tutorial on this and other multiple response problems and
will upload a draft to my site with URL copied to the list.

My page
http://surveyresearch.weebly.com/33-multiple-response-mult-response.html
carries two tutorials on less complex situations:
3.3.3.1 Analysing multiple response 1 - One field per code
3.3.3.2 Analysing multiple response exercise 2 - More codes than fields

The new one will be :
3.3.3.3 Analysing multiple response exercise 3 - More replies than values

followed by:
3.3.3.4 Analysing multiple response 4 - Dichotomous mode

The example used is the 1986 British Social Attitudes survey (quite old, but
I used it for teaching and have rights to the raw data and to my own SPSS
files for the 1986 and 1989 waves). The same kind of coding scheme has been
used throughout the series up to and including 2015. Once I have completed
these two I plan to generate further examples using data from more recent
surveys distributed by UK Data Services (Essex)

Best

John

John F Hall (Mr)
[Retired academic survey researcher]

Email: [hidden email]
Website: www.surveyresearch.weebly.com
SPSS start page: www.surveyresearch.weebly.com/1-survey-analysis-workshop

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Bruce Weaver
Sent: 05 March 2016 00:28
To: [hidden email]
Subject: Re: Error # 4651

Hi John. You already commented in another post about trying to recode 8 to
two different values. ;-)

What struck me was that the error message you posted appears to be
complaining about the missing equals sign, but not about the '-88'. That
made me wonder if SPSS is actually doing the subtraction. As this is an
empirical question, I gave it a try, as follows:

DATA LIST free / Old (F5.0).
BEGIN DATA
8 9 -80 88 1 2 3 4 5
END DATA.

RECODE Old (8-88=88) (9=99)(1 = 14) INTO New1.
RECODE Old (8-88=88) (9=99)(1 = 14) (ELSE = COPY) INTO New2.
RECODE Old (8 THRU 88=88) (9=99)(1 = 14) INTO New3.
FORMATS New1 to New3 (F5.0).
LIST.

OUTPUT:

Old New1 New2 New3

8 88 88 88
9 99 99 88
-80 . -80 .
88 . 88 88
1 14 14 14
2 . 2 .
3 . 3 .
4 . 4 .
5 . 5 .

CONCLUSIONS:

1. -80 in Old is not being recoded to 88 in New1, so SPSS is not doing the
subtraction in the RECODE command.

2. SPSS is apparently not bothered by the extra '-88' in (8-88=88).

3. 8-88 is not being interpreted as 8 THRU 88. It appears to be treated the
same as (8=88).

HTH.

John F Hall wrote

> This error message threw me until I spotted the 8-88 instead of 8=88.
> Easy
> mistake to make, but must be common enough for SPSS to spot as well..
>
>
> recode v1752 (8-88) (8=99)(1 = 14)
> / v1745 to v1751 (1=7)(2=8)(3=9)(4=10)(5=11)(6=12)(7=13).
>
> Error # 4651 in column 19. Text: )
> The symbol ')' was used on the RECODE command incorrectly. This
> symbol must follow either the replacement value on the right of the
> equals sign or the keyword CONVERT.
> Execution of this command stops.
>
> John F Hall (Mr)
> [Retired academic survey researcher]
>
> Email: <mailto:

> johnfhall@

> >

> johnfhall@

>
> Website: <http://www.surveyresearch.weebly.com/>
> www.surveyresearch.weebly.com
> SPSS start page:
> <http://surveyresearch.weebly.com/1-survey-analysis-workshop.html&g
> t; www.surveyresearch.weebly.com/1-survey-analysis-workshop
>
>
>
>
>
>
> =====================
> To manage your subscription to SPSSX-L, send a message to

> LISTSERV@.UGA

> (not to SPSSX-L), with no body text except the command. To leave the
> list, send the command SIGNOFF SPSSX-L For a list of commands to
> manage subscriptions, send the command INFO REFCARD

-----
--
Bruce Weaver
[hidden email]
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

NOTE: My Hotmail account is not monitored regularly.
To send me an e-mail, please use the address shown above.

--
View this message in context:
http://spssx-discussion.1045642.n5.nabble.com/Error-4651-tp5731665p5731667.h
tml
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command SIGNOFF SPSSX-L For a list of
commands to manage subscriptions, send the command INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

David Marso

Re: Error # 4651

Administrator

In reply to this post by John F Hall

Put down the beer ;-)

John F Hall wrote

Looks like I tried to recode the same value to two different values as well!

From: John F Hall [mailto:[hidden email]]
Sent: 04 March 2016 17:39
To: '[hidden email]' <[hidden email]>
Subject: Error # 4651

This error message threw me until I spotted the 8-88 instead of 8=88. Easy
mistake to make, but must be common enough for SPSS to spot as well..

recode v1752 (8-88) (8=99)(1 = 14)
/ v1745 to v1751 (1=7)(2=8)(3=9)(4=10)(5=11)(6=12)(7=13).

Error # 4651 in column 19. Text: )
The symbol ')' was used on the RECODE command incorrectly. This symbol must
follow either the replacement value on the right of the equals sign or the
keyword CONVERT.
Execution of this command stops.

John F Hall (Mr)
[Retired academic survey researcher]

Email: <mailto:[hidden email]> [hidden email]
Website: <http://www.surveyresearch.weebly.com/>
www.surveyresearch.weebly.com
SPSS start page:
<http://surveyresearch.weebly.com/1-survey-analysis-workshop.html>
www.surveyresearch.weebly.com/1-survey-analysis-workshop

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

Please reply to the list and not to my personal email.
Those desiring my consulting or training services please feel free to email me.
---
"Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"

Jon Peck

Re: Error # 4651

The rule is that the first value or interval wins. This can be convenient for a catchall final recode.

On Sat, Mar 5, 2016 at 3:29 PM, David Marso <[hidden email]> wrote:

Put down the beer ;-)

John F Hall wrote
> Looks like I tried to recode the same value to two different values as
> well!
>
> From: John F Hall [mailto:

> johnfhall@

> ]
> Sent: 04 March 2016 17:39
> To: '

> SPSSX-L@.UGA

> ' <

> SPSSX-L@.UGA

> >
> Subject: Error # 4651
>
> This error message threw me until I spotted the 8-88 instead of 8=88.
> Easy
> mistake to make, but must be common enough for SPSS to spot as well..
>
>
> recode v1752 (8-88) (8=99)(1 = 14)
> / v1745 to v1751 (1=7)(2=8)(3=9)(4=10)(5=11)(6=12)(7=13).
>
> Error # 4651 in column 19. Text: )
> The symbol ')' was used on the RECODE command incorrectly. This symbol
> must
> follow either the replacement value on the right of the equals sign or the
> keyword CONVERT.
> Execution of this command stops.
>
> John F Hall (Mr)
> [Retired academic survey researcher]
>
> Email: <mailto:

> johnfhall@

> >

> johnfhall@

>
> Website: <http://www.surveyresearch.weebly.com/>
> www.surveyresearch.weebly.com
> SPSS start page:
> <http://surveyresearch.weebly.com/1-survey-analysis-workshop.html>
> www.surveyresearch.weebly.com/1-survey-analysis-workshop
>
>
>
>
>
>
> =====================
> To manage your subscription to SPSSX-L, send a message to

> LISTSERV@.UGA

> (not to SPSSX-L), with no body text except the
> command. To leave the list, send the command
> SIGNOFF SPSSX-L
> For a list of commands to manage subscriptions, send the command
> INFO REFCARD

-----
Please reply to the list and not to my personal email.
Those desiring my consulting or training services please feel free to email me.
---
"Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"
--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Error-4651-tp5731665p5731669.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

Jon K Peck
[hidden email]

Jon Peck

Re: Error # 4651

And I meant to add that that the first wins rule disambiguates the recoding of the endpoint of an interval. E.g.,

(1 thru 10=1) (10 thru 20=2)

assigns 1 to the value 10.

And, regarding the original question,

(8-88) is interpreted as the pair of values 8 and -88

On Sat, Mar 5, 2016 at 4:02 PM, Jon Peck <[hidden email]> wrote:

The rule is that the first value or interval wins. This can be convenient for a catchall final recode.

On Sat, Mar 5, 2016 at 3:29 PM, David Marso <[hidden email]> wrote:
Put down the beer ;-)

John F Hall wrote
> Looks like I tried to recode the same value to two different values as
> well!
>
> From: John F Hall [mailto:

> johnfhall@

> ]
> Sent: 04 March 2016 17:39
> To: '

> SPSSX-L@.UGA

> ' <

> SPSSX-L@.UGA

> >
> Subject: Error # 4651
>
> This error message threw me until I spotted the 8-88 instead of 8=88.
> Easy
> mistake to make, but must be common enough for SPSS to spot as well..
>
>
> recode v1752 (8-88) (8=99)(1 = 14)
> / v1745 to v1751 (1=7)(2=8)(3=9)(4=10)(5=11)(6=12)(7=13).
>
> Error # 4651 in column 19. Text: )
> The symbol ')' was used on the RECODE command incorrectly. This symbol
> must
> follow either the replacement value on the right of the equals sign or the
> keyword CONVERT.
> Execution of this command stops.
>
> John F Hall (Mr)
> [Retired academic survey researcher]
>
> Email: <mailto:

> johnfhall@

> >

> johnfhall@

>
> Website: <http://www.surveyresearch.weebly.com/>
> www.surveyresearch.weebly.com
> SPSS start page:
> <http://surveyresearch.weebly.com/1-survey-analysis-workshop.html>
> www.surveyresearch.weebly.com/1-survey-analysis-workshop
>
>
>
>
>
>
> =====================
> To manage your subscription to SPSSX-L, send a message to

> LISTSERV@.UGA

> (not to SPSSX-L), with no body text except the
> command. To leave the list, send the command
> SIGNOFF SPSSX-L
> For a list of commands to manage subscriptions, send the command
> INFO REFCARD

-----
Please reply to the list and not to my personal email.
Those desiring my consulting or training services please feel free to email me.
---
"Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"
--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Error-4651-tp5731665p5731669.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

--
Jon K Peck
[hidden email]

Jon K Peck
[hidden email]

Bruce Weaver

Re: Error # 4651

Administrator

Ah, okay. Thanks Jon. It did not occur to me to include a value of -88 in my exploratory syntax (earlier in the thread).

Jon Peck wrote

And I meant to add that that the first wins rule disambiguates the recoding
of the endpoint of an interval. E.g.,
(1 thru 10=1) (10 thru 20=2)
assigns 1 to the value 10.
And, regarding the original question,
(8-88) is interpreted as the pair of values 8 and -88

On Sat, Mar 5, 2016 at 4:02 PM, Jon Peck <[hidden email]> wrote:

> The rule is that the first value or interval wins. This can be convenient
> for a catchall final recode.
>
> On Sat, Mar 5, 2016 at 3:29 PM, David Marso <[hidden email]> wrote:
>
>> Put down the beer ;-)
>>
>>
>> John F Hall wrote
>> > Looks like I tried to recode the same value to two different values as
>> > well!
>> >
>> > From: John F Hall [mailto:
>>
>> > johnfhall@
>>
>> > ]
>> > Sent: 04 March 2016 17:39
>> > To: '
>>
>> > SPSSX-L@.UGA
>>
>> > ' <
>>
>> > SPSSX-L@.UGA
>>
>> > >
>> > Subject: Error # 4651
>> >
>> > This error message threw me until I spotted the 8-88 instead of 8=88.
>> > Easy
>> > mistake to make, but must be common enough for SPSS to spot as well..
>> >
>> >
>> > recode v1752 (8-88) (8=99)(1 = 14)
>> > / v1745 to v1751 (1=7)(2=8)(3=9)(4=10)(5=11)(6=12)(7=13).
>> >
>> > Error # 4651 in column 19. Text: )
>> > The symbol ')' was used on the RECODE command incorrectly. This symbol
>> > must
>> > follow either the replacement value on the right of the equals sign or
>> the
>> > keyword CONVERT.
>> > Execution of this command stops.
>> >
>> > John F Hall (Mr)
>> > [Retired academic survey researcher]
>> >
>> > Email: <mailto:
>>
>> > johnfhall@
>>
>> > >
>>
>> > johnfhall@
>>
>> >
>> > Website: <http://www.surveyresearch.weebly.com/>
>> > www.surveyresearch.weebly.com
>> > SPSS start page:
>> > <http://surveyresearch.weebly.com/1-survey-analysis-workshop.html>
>> ;
>> > www.surveyresearch.weebly.com/1-survey-analysis-workshop
>> >
>> >
>> >
>> >
>> >
>> >
>> > =====================
>> > To manage your subscription to SPSSX-L, send a message to
>>
>> > LISTSERV@.UGA
>>
>> > (not to SPSSX-L), with no body text except the
>> > command. To leave the list, send the command
>> > SIGNOFF SPSSX-L
>> > For a list of commands to manage subscriptions, send the command
>> > INFO REFCARD
>>
>>
>>
>>
>>
>> -----
>> Please reply to the list and not to my personal email.
>> Those desiring my consulting or training services please feel free to
>> email me.
>> ---
>> "Nolite dare sanctum canibus neque mittatis margaritas vestras ante
>> porcos ne forte conculcent eas pedibus suis."
>> Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff
>> in abyssum?"
>> --
>> View this message in context:
>> http://spssx-discussion.1045642.n5.nabble.com/Error-4651-tp5731665p5731669.html
>> Sent from the SPSSX Discussion mailing list archive at Nabble.com.
>>
>> =====================
>> To manage your subscription to SPSSX-L, send a message to
>> [hidden email] (not to SPSSX-L), with no body text except the
>> command. To leave the list, send the command
>> SIGNOFF SPSSX-L
>> For a list of commands to manage subscriptions, send the command
>> INFO REFCARD
>>
>
>
>
> --
> Jon K Peck
> [hidden email]
>
>

--
Jon K Peck
[hidden email]

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

John F Hall

Re: Error # 4651

In reply to this post by Jon Peck

Jon

Is this a change from the 1970s versions?

I seem to remember that, in those days (1 thru 10=1) (10 thru 20=2) changed 10 to 2: only values 1 thru 9.9999 were recoded to 1.

John F Hall (Mr)

[Retired academic survey researcher]

Email: [hidden email]

Website: www.surveyresearch.weebly.com

SPSS start page: www.surveyresearch.weebly.com/1-survey-analysis-workshop

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Jon Peck
Sent: 06 March 2016 00:26
To: [hidden email]
Subject: Re: Error # 4651

And I meant to add that that the first wins rule disambiguates the recoding of the endpoint of an interval. E.g.,

(1 thru 10=1) (10 thru 20=2)

assigns 1 to the value 10.

And, regarding the original question,

(8-88) is interpreted as the pair of values 8 and -88

On Sat, Mar 5, 2016 at 4:02 PM, Jon Peck <[hidden email]> wrote:

The rule is that the first value or interval wins. This can be convenient for a catchall final recode.

On Sat, Mar 5, 2016 at 3:29 PM, David Marso <[hidden email]> wrote:
Put down the beer ;-)

John F Hall wrote
> Looks like I tried to recode the same value to two different values as
> well!
>
> From: John F Hall [mailto:

> johnfhall@

> ]
> Sent: 04 March 2016 17:39
> To: '

> [hidden email]

> ' <

> [hidden email]

> >
> Subject: Error # 4651
>
> This error message threw me until I spotted the 8-88 instead of 8=88.
> Easy
> mistake to make, but must be common enough for SPSS to spot as well..
>
>
> recode v1752 (8-88) (8=99)(1 = 14)
> / v1745 to v1751 (1=7)(2=8)(3=9)(4=10)(5=11)(6=12)(7=13).
>
> Error # 4651 in column 19. Text: )
> The symbol ')' was used on the RECODE command incorrectly. This symbol
> must
> follow either the replacement value on the right of the equals sign or the
> keyword CONVERT.
> Execution of this command stops.
>
> John F Hall (Mr)
> [Retired academic survey researcher]
>
> Email: <mailto:

> johnfhall@

> >

> johnfhall@

>
> Website: <http://www.surveyresearch.weebly.com/>
> www.surveyresearch.weebly.com
> SPSS start page:
> <http://surveyresearch.weebly.com/1-survey-analysis-workshop.html>
> www.surveyresearch.weebly.com/1-survey-analysis-workshop
>
>
>
>
>
>
> =====================
> To manage your subscription to SPSSX-L, send a message to

> [hidden email]

> (not to SPSSX-L), with no body text except the
> command. To leave the list, send the command
> SIGNOFF SPSSX-L
> For a list of commands to manage subscriptions, send the command
> INFO REFCARD

-----
Please reply to the list and not to my personal email.
Those desiring my consulting or training services please feel free to email me.
---
"Nolite dare sanctum canibus neque mittatis margaritas vestras ante porcos ne forte conculcent eas pedibus suis."
Cum es damnatorum possederunt porcos iens ut salire off sanguinum cliff in abyssum?"
--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Error-4651-tp5731665p5731669.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

--
Jon K Peck
[hidden email]

Jon K Peck
[hidden email]

Robert L

SV: Error # 4651

In reply to this post by Jon Peck

I saw quite recently that the syntax from a Visual Binning procedure used RECODE in a way I had not expected, but which naturally works outside the Visual Binning as well: the “first wins rule” works in steps from the lowest value. RECODE X (LO thru 10=1) (LO thru 20=2) (LO thru HI=3) assigns value 1 to LO<=X <= 10, 2 for the interval 10 < X <=20 and 3 to the rest. A neat way to dodge the problem with interval end points.

Robert

Från: SPSSX(r) Discussion [mailto:[hidden email]] För Jon Peck
Skickat: den 6 mars 2016 00:26
Till: [hidden email]
Ämne: Re: Error # 4651

And I meant to add that that the first wins rule disambiguates the recoding of the endpoint of an interval. E.g.,

(1 thru 10=1) (10 thru 20=2)

assigns 1 to the value 10.

And, regarding the original question,

(8-88) is interpreted as the pair of values 8 and -88

On Sat, Mar 5, 2016 at 4:02 PM, Jon Peck <[hidden email]> wrote:

The rule is that the first value or interval wins. This can be convenient for a catchall final recode.

Robert Lundqvist

Jon Peck

Re: SV: Error # 4651

Indeed. I designed the Visual Binner (originally known as the Visual Bander), and that trick was important for dealing with the endpoints. You will see different RECODE strategies depending on the interval endpoint setting. You will also see a (LO THRU HI=...) term as a catchall at the end of the generated RECODE syntax, which really relies on the "first wins" rule.

On Wed, Mar 9, 2016 at 6:52 AM, Robert Lundqvist <[hidden email]> wrote:

I saw quite recently that the syntax from a Visual Binning procedure used RECODE in a way I had not expected, but which naturally works outside the Visual Binning as well: the “first wins rule” works in steps from the lowest value. RECODE X (LO thru 10=1) (LO thru 20=2) (LO thru HI=3) assigns value 1 to LO<=X <= 10, 2 for the interval 10 < X <=20 and 3 to the rest. A neat way to dodge the problem with interval end points.

Robert

Från: SPSSX(r) Discussion [mailto:[hidden email]] För Jon Peck
Skickat: den 6 mars 2016 00:26
Till: [hidden email]
Ämne: Re: Error # 4651

And I meant to add that that the first wins rule disambiguates the recoding of the endpoint of an interval. E.g.,

(1 thru 10=1) (10 thru 20=2)

assigns 1 to the value 10.

And, regarding the original question,

(8-88) is interpreted as the pair of values 8 and -88

On Sat, Mar 5, 2016 at 4:02 PM, Jon Peck <[hidden email]> wrote:

The rule is that the first value or interval wins. This can be convenient for a catchall final recode.

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

Jon K Peck
[hidden email]