SPSSX Discussion

DO IF.. ELSE IF statements

Classic

List

Threaded

20 messages Options

Hashmi, Syed S

DO IF.. ELSE IF statements

Hey all,

I'm having a slight problem with some of my DO IF statements. I'm not
getting any error messages in the log but some of the ".. ELSE IF.. "
statements do not seem to run.

For example, in the code below, the first two IF statements run fine but
the final one doesn't run. At then end, the values I get for c
are '1', '3' and missings but no '2'.

DO IF a=1.
COMPUTE c=3.
ELSE IF (a=1 & (b1=1 OR b2=1 OR b3=1)).
COMPUTE c=1.
ELSE IF (a=1 & (b1=2 & b2=2 & b3=2)).
COMPUTE expt1u=2.
END IF.
EXECUTE.

However if I take that final statement and paste it in a new DO IF
statement with only those criteria, it works just fine. The code looks
fine. The criteria are legit (I've selected cases based on those criteria
and they do exist). Is there something else I'm missing here? I initially
noticed the issue a couple of weeks back but and initially it was little
inconvenient to create new DO IF statements for the lines that didn't run,
but now it's just getting annoying. Not to mention that I worry that there
might be something worse going on that I'm not aware of.

Incidentaly, all this is being run in SPSS 15.0 on an XP machine.

Thanks a heap in advance.

Melissa Ives

Re: DO IF.. ELSE IF statements

I recently had a similar experience (with SPSS 14) and a colleague
recommended running the else statements again separately due to
experiencing a similar phenomenon. Re-running worked fine that way. I
too am curious about why the statements did not appear to work in the DO
IF..ELSE IF' loop.

Melissa

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Syed Hashmi
Sent: Wednesday, August 08, 2007 3:02 PM
To: [hidden email]
Subject: [SPSSX-L] DO IF.. ELSE IF statements

Hey all,

I'm having a slight problem with some of my DO IF statements. I'm not
getting any error messages in the log but some of the ".. ELSE IF.. "
statements do not seem to run.

For example, in the code below, the first two IF statements run fine but
the final one doesn't run. At then end, the values I get for c are '1',
'3' and missings but no '2'.

DO IF a=1.
COMPUTE c=3.
ELSE IF (a=1 & (b1=1 OR b2=1 OR b3=1)).
COMPUTE c=1.
ELSE IF (a=1 & (b1=2 & b2=2 & b3=2)).
COMPUTE expt1u=2.
END IF.
EXECUTE.

However if I take that final statement and paste it in a new DO IF
statement with only those criteria, it works just fine. The code looks
fine. The criteria are legit (I've selected cases based on those
criteria and they do exist). Is there something else I'm missing here?
I initially noticed the issue a couple of weeks back but and initially
it was little inconvenient to create new DO IF statements for the lines
that didn't run, but now it's just getting annoying. Not to mention that
I worry that there might be something worse going on that I'm not aware
of.

Incidentaly, all this is being run in SPSS 15.0 on an XP machine.

Thanks a heap in advance.

PRIVILEGED AND CONFIDENTIAL INFORMATION
This transmittal and any attachments may contain PRIVILEGED AND
CONFIDENTIAL information and is intended only for the use of the
addressee. If you are not the designated recipient, or an employee
or agent authorized to deliver such transmittals to the designated
recipient, you are hereby notified that any dissemination,
copying or publication of this transmittal is strictly prohibited. If
you have received this transmittal in error, please notify us
immediately by replying to the sender and delete this copy from your
system. You may also call us at (309) 827-6026 for assistance.

Hashmi, Syed S

Re: DO IF.. ELSE IF statements

Same here. They work fine if run separately. I've done just that and
modified my code but it would be good to know why it happens.

Shahrukh.

> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf
Of
> Melissa Ives
> Sent: Wednesday, August 08, 2007 3:06 PM
> To: [hidden email]
> Subject: Re: DO IF.. ELSE IF statements
>
> I recently had a similar experience (with SPSS 14) and a colleague
> recommended running the else statements again separately due to
> experiencing a similar phenomenon. Re-running worked fine that way.
I
> too am curious about why the statements did not appear to work in the
DO
> IF..ELSE IF' loop.
>
> Melissa
>
> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf
Of

> Syed Hashmi
> Sent: Wednesday, August 08, 2007 3:02 PM
> To: [hidden email]
> Subject: [SPSSX-L] DO IF.. ELSE IF statements
>
> Hey all,
>
> I'm having a slight problem with some of my DO IF statements. I'm not
> getting any error messages in the log but some of the ".. ELSE IF.. "
> statements do not seem to run.
>
> For example, in the code below, the first two IF statements run fine

but
> the final one doesn't run. At then end, the values I get for c are
'1',

> '3' and missings but no '2'.
>
> DO IF a=1.
> COMPUTE c=3.
> ELSE IF (a=1 & (b1=1 OR b2=1 OR b3=1)).
> COMPUTE c=1.
> ELSE IF (a=1 & (b1=2 & b2=2 & b3=2)).
> COMPUTE expt1u=2.
> END IF.
> EXECUTE.
>
> However if I take that final statement and paste it in a new DO IF
> statement with only those criteria, it works just fine. The code looks
> fine. The criteria are legit (I've selected cases based on those
> criteria and they do exist). Is there something else I'm missing

here?
> I initially noticed the issue a couple of weeks back but and initially
> it was little inconvenient to create new DO IF statements for the
lines
> that didn't run, but now it's just getting annoying. Not to mention
that
> I worry that there might be something worse going on that I'm not
aware

> of.
>
> Incidentaly, all this is being run in SPSS 15.0 on an XP machine.
>
> Thanks a heap in advance.
>
>
> PRIVILEGED AND CONFIDENTIAL INFORMATION
> This transmittal and any attachments may contain PRIVILEGED AND
> CONFIDENTIAL information and is intended only for the use of the
> addressee. If you are not the designated recipient, or an employee
> or agent authorized to deliver such transmittals to the designated
> recipient, you are hereby notified that any dissemination,
> copying or publication of this transmittal is strictly prohibited. If
> you have received this transmittal in error, please notify us
> immediately by replying to the sender and delete this copy from your
> system. You may also call us at (309) 827-6026 for assistance.

Gary Rosin

Re: DO IF.. ELSE IF statements

In reply to this post by Hashmi, Syed S

>some of the ".. ELSE IF.. "
>statements do not seem to run.
>
>For example, in the code below, the first two IF statements run
>fine but the final one doesn't run. At then end, the values I get
>for **c** are '1', '3' and missings but no '2'.
>
> DO IF a=1.
> COMPUTE c=3.
> ELSE IF (a=1 & (b1=1 OR b2=1 OR b3=1)).
> COMPUTE c=1.
> ELSE IF (a=1 & (b1=2 & b2=2 & b3=2)).

**
> COMPUTE expt1u=2.
**
> END IF.
> EXECUTE.

(** = emphasis added)

I'm reluctant to comment with so many list-members who are much
more SPSS-adept than I am. Still, I notice that your question asks
about missing "2" values for "c", but your *syntax* assigns the
value of "2" to "exp1u" (and *not* to "c").

Gary

---
Gary Rosin <[hidden email]>

"You're just a wave,
you're not the water."
--Butch Hancock

Hal 9000

Re: DO IF.. ELSE IF statements

It's because everywhere an a = 1 is already dealt with in the first
block - all combinations of a = 1 won't be addressed afterwards
(because they already have been!). If you do them seperately, be SURE
you want the results it gives you....some initial computations will be
overwritten!
-Gary

On 8/8/07, Gary Rosin <[hidden email]> wrote:

> >some of the ".. ELSE IF.. "
> >statements do not seem to run.
> >
> >For example, in the code below, the first two IF statements run
> >fine but the final one doesn't run. At then end, the values I get
> >for **c** are '1', '3' and missings but no '2'.
> >
> > DO IF a=1.
> > COMPUTE c=3.
> > ELSE IF (a=1 & (b1=1 OR b2=1 OR b3=1)).
> > COMPUTE c=1.
> > ELSE IF (a=1 & (b1=2 & b2=2 & b3=2)).
> **
> > COMPUTE expt1u=2.
> **
> > END IF.
> > EXECUTE.
>
> (** = emphasis added)
>
> I'm reluctant to comment with so many list-members who are much
> more SPSS-adept than I am. Still, I notice that your question asks
> about missing "2" values for "c", but your *syntax* assigns the
> value of "2" to "exp1u" (and *not* to "c").
>
> Gary
>
> ---
> Gary Rosin <[hidden email]>
>
> "You're just a wave,
> you're not the water."
> --Butch Hancock
>

Oliver, Richard

Re: DO IF.. ELSE IF statements

In reply to this post by Hashmi, Syed S

Once a condition is met for a case, processing stops for that case, and the DO IF structure moves on to the next case. So for every case where a=1, the first condition is met, c is set to 3, and DO IF moves on to the next case. Since the remaining conditions include a=1, they're never applied to any cases, because any case where a=1 has already met the first condition.

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Hashmi, Syed S
Sent: Wednesday, August 08, 2007 3:24 PM
To: [hidden email]
Subject: Re: DO IF.. ELSE IF statements

Same here. They work fine if run separately. I've done just that and
modified my code but it would be good to know why it happens.

Shahrukh.

> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf
Of
> Melissa Ives
> Sent: Wednesday, August 08, 2007 3:06 PM
> To: [hidden email]
> Subject: Re: DO IF.. ELSE IF statements
>
> I recently had a similar experience (with SPSS 14) and a colleague
> recommended running the else statements again separately due to
> experiencing a similar phenomenon. Re-running worked fine that way.
I
> too am curious about why the statements did not appear to work in the
DO
> IF..ELSE IF' loop.
>
> Melissa
>
> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf
Of

but
> the final one doesn't run. At then end, the values I get for c are
'1',

Hashmi, Syed S

Re: DO IF.. ELSE IF statements

In reply to this post by Gary Rosin

Yes, sorry about that. I had renamed my variables for the post and had
forgotten to change it in that last line. My bad! But c = expt1u = 2.

Shahrukh

> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf
Of

> Gary Rosin
> Sent: Wednesday, August 08, 2007 3:59 PM
> To: [hidden email]
> Subject: Re: DO IF.. ELSE IF statements
>
> >some of the ".. ELSE IF.. "
> >statements do not seem to run.
> >
> >For example, in the code below, the first two IF statements run
> >fine but the final one doesn't run. At then end, the values I get
> >for **c** are '1', '3' and missings but no '2'.
> >
> > DO IF a=1.
> > COMPUTE c=3.
> > ELSE IF (a=1 & (b1=1 OR b2=1 OR b3=1)).
> > COMPUTE c=1.
> > ELSE IF (a=1 & (b1=2 & b2=2 & b3=2)).
> **
> > COMPUTE expt1u=2.
> **
> > END IF.
> > EXECUTE.
>
> (** = emphasis added)
>
> I'm reluctant to comment with so many list-members who are much
> more SPSS-adept than I am. Still, I notice that your question asks
> about missing "2" values for "c", but your *syntax* assigns the
> value of "2" to "exp1u" (and *not* to "c").
>
> Gary
>
> ---
> Gary Rosin <[hidden email]>
>
> "You're just a wave,
> you're not the water."
> --Butch Hancock

Hashmi, Syed S

Re: DO IF.. ELSE IF statements

In reply to this post by Oliver, Richard

So you're saying SPSS will look at the first of perhaps numerous
criteria and if that one has already been satisfied in a previous ..IF..
statement then it'll stop processing any more ..ELSE IF.. statements in
the loop?

Does this mean that SPSS doesn't look at the other criteria in that IF
statement? What if all the criteria are enclosed in ( )??

Furthermore, I have run a bunch of DO IF..ELSE IF loops where the first
criteria for all of them was the same and there followed about 6 more
criteria that were different in each ELSE IF statement. Many of these
loops worked perfectly. Of some that didn't work perfectly, some ELSE IF
statements did, some didn't, although the syntax was the same all
through.

- Shahrukh

> -----Original Message-----
> From: Oliver, Richard [mailto:[hidden email]]
> Sent: Wednesday, August 08, 2007 4:14 PM
> To: Hashmi, Syed S; [hidden email]
> Subject: RE: Re: DO IF.. ELSE IF statements
>
> Once a condition is met for a case, processing stops for that case,
and
> the DO IF structure moves on to the next case. So for every case where
> a=1, the first condition is met, c is set to 3, and DO IF moves on to
the
> next case. Since the remaining conditions include a=1, they're never
> applied to any cases, because any case where a=1 has already met the
first
> condition.
>
> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf
Of

> Hashmi, Syed S
> Sent: Wednesday, August 08, 2007 3:24 PM
> To: [hidden email]
> Subject: Re: DO IF.. ELSE IF statements
>
> Same here. They work fine if run separately. I've done just that and
> modified my code but it would be good to know why it happens.
>
> Shahrukh.
>
>
> > -----Original Message-----
> > From: SPSSX(r) Discussion [mailto:[hidden email]] On

Behalf

> Of
> > Melissa Ives
> > Sent: Wednesday, August 08, 2007 3:06 PM
> > To: [hidden email]
> > Subject: Re: DO IF.. ELSE IF statements
> >
> > I recently had a similar experience (with SPSS 14) and a colleague
> > recommended running the else statements again separately due to
> > experiencing a similar phenomenon. Re-running worked fine that way.
> I
> > too am curious about why the statements did not appear to work in

the
> DO
> > IF..ELSE IF' loop.
> >
> > Melissa
> >
> > -----Original Message-----
> > From: SPSSX(r) Discussion [mailto:[hidden email]] On
Behalf
> Of
> > Syed Hashmi
> > Sent: Wednesday, August 08, 2007 3:02 PM
> > To: [hidden email]
> > Subject: [SPSSX-L] DO IF.. ELSE IF statements
> >
> > Hey all,
> >
> > I'm having a slight problem with some of my DO IF statements. I'm
not
> > getting any error messages in the log but some of the ".. ELSE IF..
"

> > statements do not seem to run.
> >
> > For example, in the code below, the first two IF statements run fine
> but
> > the final one doesn't run. At then end, the values I get for c are
> '1',
> > '3' and missings but no '2'.
> >
> > DO IF a=1.
> > COMPUTE c=3.
> > ELSE IF (a=1 & (b1=1 OR b2=1 OR b3=1)).
> > COMPUTE c=1.
> > ELSE IF (a=1 & (b1=2 & b2=2 & b3=2)).
> > COMPUTE expt1u=2.
> > END IF.
> > EXECUTE.
> >
> > However if I take that final statement and paste it in a new DO IF
> > statement with only those criteria, it works just fine. The code

looks
> > fine. The criteria are legit (I've selected cases based on those
> > criteria and they do exist). Is there something else I'm missing
> here?
> > I initially noticed the issue a couple of weeks back but and
initially

> > it was little inconvenient to create new DO IF statements for the
> lines
> > that didn't run, but now it's just getting annoying. Not to mention
> that
> > I worry that there might be something worse going on that I'm not
> aware
> > of.
> >
> > Incidentaly, all this is being run in SPSS 15.0 on an XP machine.
> >
> > Thanks a heap in advance.
> >
> >
> > PRIVILEGED AND CONFIDENTIAL INFORMATION
> > This transmittal and any attachments may contain PRIVILEGED AND
> > CONFIDENTIAL information and is intended only for the use of the
> > addressee. If you are not the designated recipient, or an employee
> > or agent authorized to deliver such transmittals to the designated
> > recipient, you are hereby notified that any dissemination,
> > copying or publication of this transmittal is strictly prohibited.

If
> > you have received this transmittal in error, please notify us
> > immediately by replying to the sender and delete this copy from your
> > system. You may also call us at (309) 827-6026 for assistance.

Hashmi, Syed S

Re: DO IF.. ELSE IF statements

In reply to this post by Hashmi, Syed S

On Wed, 8 Aug 2007 14:04:26 -0700, Hal 9000 <[hidden email]> wrote:

>It's because everywhere an a = 1 is already dealt with in the first
>block - all combinations of a = 1 won't be addressed afterwards
>(because they already have been!). If you do them seperately, be SURE
>you want the results it gives you....some initial computations will be
>overwritten!
>-Gary
>

Gary,

I just posted another reply to Richard's post which dealth with the same
issue of "has the first criterion been satisfied?". Like I mentioned in
that post, there is probably some other reason why that loop doesn't work
since there are other loops with similar first criterion that worked just
fine.

Also, am I wrong in assuming that DO IF loops with criteria of
DO IF (a=1 & b=1)... ELSE IF (a=1 & b=2)
would be more or less commonplace? Shouldn't the criteria be looked at in
full as opposed with >1 sub-criterion?

I'm just curious about this since the first criterion rule doesn't seem to
apply everytime.

- Shahrukh

Melissa Ives

Re: DO IF.. ELSE IF statements

In reply to this post by Hashmi, Syed S

In the case you mention below, I would do it...
DO IF a=1.
do if (b=1).
<commands>
else if (b=2).
<commands>
end if.
ELSE IF a=2.
<etc.>
END IF.

Melissa

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Syed Hashmi
Sent: Wednesday, August 08, 2007 4:44 PM
To: [hidden email]
Subject: Re: [SPSSX-L] DO IF.. ELSE IF statements

On Wed, 8 Aug 2007 14:04:26 -0700, Hal 9000 <[hidden email]> wrote:

>It's because everywhere an a = 1 is already dealt with in the first
>block - all combinations of a = 1 won't be addressed afterwards
>(because they already have been!). If you do them seperately, be SURE
>you want the results it gives you....some initial computations will be
>overwritten!
>-Gary
>

Gary,

I just posted another reply to Richard's post which dealth with the same
issue of "has the first criterion been satisfied?". Like I mentioned in
that post, there is probably some other reason why that loop doesn't
work since there are other loops with similar first criterion that
worked just fine.

Also, am I wrong in assuming that DO IF loops with criteria of
DO IF (a=1 & b=1)... ELSE IF (a=1 & b=2) would be more or less
commonplace? Shouldn't the criteria be looked at in full as opposed
with >1 sub-criterion?

I'm just curious about this since the first criterion rule doesn't seem
to apply everytime.

- Shahrukh

PRIVILEGED AND CONFIDENTIAL INFORMATION
This transmittal and any attachments may contain PRIVILEGED AND
CONFIDENTIAL information and is intended only for the use of the
addressee. If you are not the designated recipient, or an employee
or agent authorized to deliver such transmittals to the designated
recipient, you are hereby notified that any dissemination,
copying or publication of this transmittal is strictly prohibited. If
you have received this transmittal in error, please notify us
immediately by replying to the sender and delete this copy from your
system. You may also call us at (309) 827-6026 for assistance.

Hal 9000

Re: DO IF.. ELSE IF statements

In reply to this post by Hashmi, Syed S

> Also, am I wrong in assuming that DO IF loops with criteria of
> DO IF (a=1 & b=1)... ELSE IF (a=1 & b=2)

Those are not mutually exclusive distinctions. You have to specify

do if a=1 & b=1 & not (such and such condition).
else if a=1 & b=1 & (such & such condition).

a silly example:

Do if (I like apple pie).
compute X = 1.
Do if(I like apple pie and blueberry pie).
compute X = 2.
end if.

For that row of data, it stops after the first 'do if' because it
finds a match. It doesn't keep looking past the fact that you like
apple pie, because you are telling the program these categories are
mutually exclusive.

This is OK:

do if (I like apple pie and DON'T like blueberry pie).
compute X = 1.
do if (I like apple pie and DO like blueberry pie).
compute X = 2.
end if.

Melissa's post is nice, but it's really important to understand the
logic rather than find something that works (or appears to work, but
actually doesn't).

Hal 9000

Re: DO IF.. ELSE IF statements

Those are not mutually exclusive distinctions

Ooops, they are! Haha

On 8/8/07, Hal 9000 <[hidden email]> wrote:

> > Also, am I wrong in assuming that DO IF loops with criteria of
> > DO IF (a=1 & b=1)... ELSE IF (a=1 & b=2)
>
> Those are not mutually exclusive distinctions. You have to specify
>
> do if a=1 & b=1 & not (such and such condition).
> else if a=1 & b=1 & (such & such condition).
>
> a silly example:
>
> Do if (I like apple pie).
> compute X = 1.
> Do if(I like apple pie and blueberry pie).
> compute X = 2.
> end if.
>
> For that row of data, it stops after the first 'do if' because it
> finds a match. It doesn't keep looking past the fact that you like
> apple pie, because you are telling the program these categories are
> mutually exclusive.
>
> This is OK:
>
> do if (I like apple pie and DON'T like blueberry pie).
> compute X = 1.
> do if (I like apple pie and DO like blueberry pie).
> compute X = 2.
> end if.
>
> Melissa's post is nice, but it's really important to understand the
> logic rather than find something that works (or appears to work, but
> actually doesn't).
>

Melissa Ives

Re: DO IF.. ELSE IF statements

In reply to this post by Hashmi, Syed S

This is OK as well:

do if (I like apple pie) .
compute X = 1.
do if (I like blueberry pie).
compute X = 2.
end if.
end if.

Because it first searches for any liking of apple pie, then once those
are identified (with X=1) looks at that list for who also likes
blueberry and changes X to 2 in the final calculation -- only because
the blueberry DO IF is embedded within the apple DO IF.

Of course, neither syntax does anything if you only like blueberry pie
and not apple or if you prefer lemon over either/both of these :-) (I
love all 3!)

Melissa
(BTW-Hal, you need a second END IF somewhere--where often makes a
difference.)

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Hal 9000
Sent: Wednesday, August 08, 2007 5:01 PM
To: [hidden email]
Subject: Re: [SPSSX-L] DO IF.. ELSE IF statements

> Also, am I wrong in assuming that DO IF loops with criteria of
> DO IF (a=1 & b=1)... ELSE IF (a=1 & b=2)

Those are not mutually exclusive distinctions. You have to specify

do if a=1 & b=1 & not (such and such condition).
else if a=1 & b=1 & (such & such condition).

a silly example:

Do if (I like apple pie).
compute X = 1.
Do if(I like apple pie and blueberry pie).
compute X = 2.
end if.

For that row of data, it stops after the first 'do if' because it finds
a match. It doesn't keep looking past the fact that you like apple pie,
because you are telling the program these categories are mutually
exclusive.

This is OK:

do if (I like apple pie and DON'T like blueberry pie).
compute X = 1.
do if (I like apple pie and DO like blueberry pie).
compute X = 2.
end if.

Melissa's post is nice, but it's really important to understand the
logic rather than find something that works (or appears to work, but
actually doesn't).

PRIVILEGED AND CONFIDENTIAL INFORMATION
This transmittal and any attachments may contain PRIVILEGED AND
CONFIDENTIAL information and is intended only for the use of the
addressee. If you are not the designated recipient, or an employee
or agent authorized to deliver such transmittals to the designated
recipient, you are hereby notified that any dissemination,
copying or publication of this transmittal is strictly prohibited. If
you have received this transmittal in error, please notify us
immediately by replying to the sender and delete this copy from your
system. You may also call us at (309) 827-6026 for assistance.

Reutter, Alex

Re: DO IF.. ELSE IF statements

In reply to this post by Hashmi, Syed S

You need to post examples of the "loops with similar first criterion that worked just fine", because they may not be doing what you think they are doing.

You should also read the DO IF SPSS syntax documentation, and then for comparison look at documentation of control structures in C, Python, Fortran, Java, etc... How ELSE IF functions in SPSS is how the comparable control structures work in other programming languages.

Melissa's suggestion is a very sensible one.

Alex

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Syed Hashmi
Sent: Wednesday, August 08, 2007 4:44 PM
To: [hidden email]
Subject: Re: DO IF.. ELSE IF statements

On Wed, 8 Aug 2007 14:04:26 -0700, Hal 9000 <[hidden email]> wrote:

>It's because everywhere an a = 1 is already dealt with in the first
>block - all combinations of a = 1 won't be addressed afterwards
>(because they already have been!). If you do them seperately, be SURE
>you want the results it gives you....some initial computations will be
>overwritten!
>-Gary
>

Gary,

I just posted another reply to Richard's post which dealth with the same
issue of "has the first criterion been satisfied?". Like I mentioned in
that post, there is probably some other reason why that loop doesn't work
since there are other loops with similar first criterion that worked just
fine.

Also, am I wrong in assuming that DO IF loops with criteria of
DO IF (a=1 & b=1)... ELSE IF (a=1 & b=2)
would be more or less commonplace? Shouldn't the criteria be looked at in
full as opposed with >1 sub-criterion?

I'm just curious about this since the first criterion rule doesn't seem to
apply everytime.

- Shahrukh

Hashmi, Syed S

Re: DO IF.. ELSE IF statements

In reply to this post by Hal 9000

Thanks Hal. (and Melissa and Gary).

That made more sense. I just assumed that it would look at the condition
in entirety as opposed to breaking it down to each single criterion in
the condition. For example, in your example it would compute X=1 for all
those who like apple pie and then compute X=2 for the subset of the same
apple pie 'likers' who ALSO like blueberry pie - but like you mentioned,
the two statements are not mutually exclusive.

Thanks for the explanation. I've already recoded all my work but this
info will probably come in handy the next time round.

- Shahrukh

> -----Original Message-----
> From: Hal 9000 [mailto:[hidden email]]
> Sent: Wednesday, August 08, 2007 5:01 PM
> To: Hashmi, Syed S; [hidden email]
> Subject: Re: DO IF.. ELSE IF statements
>
> > Also, am I wrong in assuming that DO IF loops with criteria of
> > DO IF (a=1 & b=1)... ELSE IF (a=1 & b=2)
>
> Those are not mutually exclusive distinctions. You have to specify
>
> do if a=1 & b=1 & not (such and such condition).
> else if a=1 & b=1 & (such & such condition).
>
> a silly example:
>
> Do if (I like apple pie).
> compute X = 1.
> Do if(I like apple pie and blueberry pie).
> compute X = 2.
> end if.
>
> For that row of data, it stops after the first 'do if' because it
> finds a match. It doesn't keep looking past the fact that you like
> apple pie, because you are telling the program these categories are
> mutually exclusive.
>
> This is OK:
>
> do if (I like apple pie and DON'T like blueberry pie).
> compute X = 1.
> do if (I like apple pie and DO like blueberry pie).
> compute X = 2.
> end if.
>
> Melissa's post is nice, but it's really important to understand the
> logic rather than find something that works (or appears to work, but
> actually doesn't).

Hashmi, Syed S

Re: DO IF.. ELSE IF statements

In reply to this post by Hashmi, Syed S

On Wed, 8 Aug 2007 17:50:34 -0500, Reutter, Alex <[hidden email]> wrote:

>You need to post examples of the "loops with similar first criterion that
worked just fine", because they may not be doing what you think they are
doing.
>
>You should also read the DO IF SPSS syntax documentation, and then for
comparison look at documentation of control structures in C, Python,
Fortran, Java, etc... How ELSE IF functions in SPSS is how the comparable
control structures work in other programming languages.
>
>Melissa's suggestion is a very sensible one.
>
>Alex
>

Alex,

Here's a bit of code that worked fine completely. The whole loop was
within a DO REPEAT command and 'a' to 'g' were different variables defined
in the DO REPEAT command. If you notice, the first four condition in each
IF statement are the same:

DO IF respill=1 /* code 4.51 */
& NOT(MISSING(a))
& b GE 1 & b LE 12
& MISSING(c)
& NOT(MISSING(d))
& MISSING(e)
& MISSING(f)
& NOT(MISSING(g)).
COMPUTE x = 4.51.
ELSE IF respill=1 /* code 4.52 */
& NOT(MISSING(a))
& b GE 1 & b LE 12
& MISSING(c)
& MISSING(d)
& MISSING(e)
& MISSING(f)
& NOT(MISSING(g)).
COMPUTE x = 4.52.
END IF.

On the other hand, here's a bit of code that didn't work completely. Codes
3.11, 3.13 and 3.15 worked, but 3.12 and 3.14 didn't. I can see how
they're different, but the firt three conditions in each IF statement are
again same. They are all mutually exclusive only if you take the whole IF
statement in it's entirety but not if you break it down by sub-conditions.
It's the same as the code that worked above.

DO IF respill=1 /* code 3.11 */
& NOT(MISSING(a))
& b GE 21 & b LE 33
& c GE 91 & c LE 93
& NOT(MISSING(d))
& e GE 1 & e LE 12
& f GE 1 & f LE 31
& MISSING(g).
COMPUTE x = 3.11.
ELSE IF respill=1 /* code 3.12 */
& NOT(MISSING(a))
& b GE 21 & b LE 33
& MISSING(c)
& NOT(MISSING(d))
& e GE 1 & e LE 12
& f GE 1 & f LE 31
& MISSING(g).
COMPUTE x = 3.12.
ELSE IF respill=1 /* code 3.13 */
& NOT(MISSING(a))
& b GE 21 & b LE 33
& MISSING(c)
& NOT(MISSING(d))
& e GE 1 & e LE 12
& MISSING(f)
& NOT(MISSING(g)).
COMPUTE x = 3.13.
ELSE IF respill=1 /* code 3.14 */
& NOT(MISSING(a))
& b GE 21 & b LE 33
& MISSING(c)
& NOT(MISSING(d))
& MISSING(e)
& MISSING(f)
& NOT(MISSING(g)).
COMPUTE x = 3.14.
ELSE IF respill=1 /* code 3.15 */
& NOT(MISSING(a))
& b GE 21 & b LE 33
& c GE 1 & c LE 31
& MISSING(d)
& MISSING(e)
& MISSING(f)
& NOT(MISSING(g)).
COMPUTE x = 3.15.
END IF.

I'll go over the documentation that you had suggested. I had tried similar
code in SAS and it had worked but haven't tried in any of the other
languages. Thanks again for the tip.

- Shahrukh

Oliver, Richard

Re: DO IF.. ELSE IF statements

In reply to this post by Hashmi, Syed S

DO IF-ELSE IF evaluates each condition separately. They don't have to be mutually exclusive conditions, although that may make your code easier to debug. It all depends on what you're trying to accomplish. Sometimes you may want to go from the most restrictive condition to least restrictive, sometimes the reverse. In your original example, you started with the least restrictive condition, causing any cases that met any of your conditions to meet the first condition.

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Hashmi, Syed S
Sent: Wednesday, August 08, 2007 5:58 PM
To: [hidden email]
Subject: Re: DO IF.. ELSE IF statements

Thanks Hal. (and Melissa and Gary).

That made more sense. I just assumed that it would look at the condition
in entirety as opposed to breaking it down to each single criterion in
the condition. For example, in your example it would compute X=1 for all
those who like apple pie and then compute X=2 for the subset of the same
apple pie 'likers' who ALSO like blueberry pie - but like you mentioned,
the two statements are not mutually exclusive.

Thanks for the explanation. I've already recoded all my work but this
info will probably come in handy the next time round.

- Shahrukh

Hal 9000

Re: DO IF.. ELSE IF statements

Is this thread making anybody else hungry?

Hum, good point - I'm so conditioned to covering ALL of my bases using
clearly delimited and mutually exclusive distinctions that the
possibility of layered processing with decreasingly strict conditions
didn't occur to me! For moderately complex combinations it could be a
debugging nightmare in the making, though (for my mind, at least).

Melissa - your code makes/made perfect sense, and even looks pretty. I
only meant that in this case it seems more important to demysify why
it works or doesn't work. I certainly refined my own way of thinking
about this flow control method, so I won't apologize for bumbling
around in the thread! :)

[note to self : do not multitask]

On 8/8/07, Oliver, Richard <[hidden email]> wrote:

> DO IF-ELSE IF evaluates each condition separately. They don't have to be mutually exclusive conditions, although that may make your code easier to debug. It all depends on what you're trying to accomplish. Sometimes you may want to go from the most restrictive condition to least restrictive, sometimes the reverse. In your original example, you started with the least restrictive condition, causing any cases that met any of your conditions to meet the first condition.
>
> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Hashmi, Syed S
> Sent: Wednesday, August 08, 2007 5:58 PM
> To: [hidden email]
> Subject: Re: DO IF.. ELSE IF statements
>
> Thanks Hal. (and Melissa and Gary).
>
> That made more sense. I just assumed that it would look at the condition
> in entirety as opposed to breaking it down to each single criterion in
> the condition. For example, in your example it would compute X=1 for all
> those who like apple pie and then compute X=2 for the subset of the same
> apple pie 'likers' who ALSO like blueberry pie - but like you mentioned,
> the two statements are not mutually exclusive.
>
> Thanks for the explanation. I've already recoded all my work but this
> info will probably come in handy the next time round.
>
> - Shahrukh
>
> > -----Original Message-----
> > From: Hal 9000 [mailto:[hidden email]]
> > Sent: Wednesday, August 08, 2007 5:01 PM
> > To: Hashmi, Syed S; [hidden email]
> > Subject: Re: DO IF.. ELSE IF statements
> >
> > > Also, am I wrong in assuming that DO IF loops with criteria of
> > > DO IF (a=1 & b=1)... ELSE IF (a=1 & b=2)
> >
> > Those are not mutually exclusive distinctions. You have to specify
> >
> > do if a=1 & b=1 & not (such and such condition).
> > else if a=1 & b=1 & (such & such condition).
> >
> > a silly example:
> >
> > Do if (I like apple pie).
> > compute X = 1.
> > Do if(I like apple pie and blueberry pie).
> > compute X = 2.
> > end if.
> >
> > For that row of data, it stops after the first 'do if' because it
> > finds a match. It doesn't keep looking past the fact that you like
> > apple pie, because you are telling the program these categories are
> > mutually exclusive.
> >
> > This is OK:
> >
> > do if (I like apple pie and DON'T like blueberry pie).
> > compute X = 1.
> > do if (I like apple pie and DO like blueberry pie).
> > compute X = 2.
> > end if.
> >
> > Melissa's post is nice, but it's really important to understand the
> > logic rather than find something that works (or appears to work, but
> > actually doesn't).
>

Reutter, Alex

Re: DO IF.. ELSE IF statements

In reply to this post by Hashmi, Syed S

The example that works below does not have similar criteria to the one that doesn't. In the first example, the set of cases for which the DO IF (code 4.51) evaluates to true is mutually exclusive with the set of cases for which the ELSE IF (code 4.52) evaluates to true; to simplify, the expression for 4.51 is essentially (this & NOT(that))** and the expression for 4.52 is (this & that).

In the second example, the set of cases for which the first ELSE IF (code 3.12) evaluates to true is a subset of the set of cases for which the DO IF (code 3.11) evaluates to true; to simplify, the expression for 3.11 is essentially (this) and the expression for 3.12 is (this & that)***. Add & NOT(MISSING(c)) to the DO IF (code 3.11) and 3.12 should "work".

It's all a matter of set logic. Does this help? That is, do you see what needs to be done to make 3.14 work?

Cheers,
Alex

** where "this" is respill=1 & NOT(MISSING(a)) & b GE 1 & b LE 12 & MISSING(c) & MISSING(e) & MISSING(f) & NOT(MISSING(g)) and "that" is MISSING(d)

*** where "this" is respill=1 & NOT(MISSING(a)) & b GE 21 & b LE 33 & c GE 91 & c LE 93 & NOT(MISSING(d)) & e GE 1 & e LE 12 & f GE 1 & f LE 31 & MISSING(g) and "that" is MISSING(c)

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Syed Hashmi
Sent: Wednesday, August 08, 2007 6:52 PM
To: [hidden email]
Subject: Re: DO IF.. ELSE IF statements

On Wed, 8 Aug 2007 17:50:34 -0500, Reutter, Alex <[hidden email]> wrote:

>You need to post examples of the "loops with similar first criterion that
worked just fine", because they may not be doing what you think they are
doing.
>
>You should also read the DO IF SPSS syntax documentation, and then for
comparison look at documentation of control structures in C, Python,
Fortran, Java, etc... How ELSE IF functions in SPSS is how the comparable
control structures work in other programming languages.
>
>Melissa's suggestion is a very sensible one.
>
>Alex
>

Alex,

Here's a bit of code that worked fine completely. The whole loop was
within a DO REPEAT command and 'a' to 'g' were different variables defined
in the DO REPEAT command. If you notice, the first four condition in each
IF statement are the same:

DO IF respill=1 /* code 4.51 */
& NOT(MISSING(a))
& b GE 1 & b LE 12
& MISSING(c)
& NOT(MISSING(d))
& MISSING(e)
& MISSING(f)
& NOT(MISSING(g)).
COMPUTE x = 4.51.
ELSE IF respill=1 /* code 4.52 */
& NOT(MISSING(a))
& b GE 1 & b LE 12
& MISSING(c)
& MISSING(d)
& MISSING(e)
& MISSING(f)
& NOT(MISSING(g)).
COMPUTE x = 4.52.
END IF.

On the other hand, here's a bit of code that didn't work completely. Codes
3.11, 3.13 and 3.15 worked, but 3.12 and 3.14 didn't. I can see how
they're different, but the firt three conditions in each IF statement are
again same. They are all mutually exclusive only if you take the whole IF
statement in it's entirety but not if you break it down by sub-conditions.
It's the same as the code that worked above.

DO IF respill=1 /* code 3.11 */
& NOT(MISSING(a))
& b GE 21 & b LE 33
& c GE 91 & c LE 93
& NOT(MISSING(d))
& e GE 1 & e LE 12
& f GE 1 & f LE 31
& MISSING(g).
COMPUTE x = 3.11.
ELSE IF respill=1 /* code 3.12 */
& NOT(MISSING(a))
& b GE 21 & b LE 33
& MISSING(c)
& NOT(MISSING(d))
& e GE 1 & e LE 12
& f GE 1 & f LE 31
& MISSING(g).
COMPUTE x = 3.12.
ELSE IF respill=1 /* code 3.13 */
& NOT(MISSING(a))
& b GE 21 & b LE 33
& MISSING(c)
& NOT(MISSING(d))
& e GE 1 & e LE 12
& MISSING(f)
& NOT(MISSING(g)).
COMPUTE x = 3.13.
ELSE IF respill=1 /* code 3.14 */
& NOT(MISSING(a))
& b GE 21 & b LE 33
& MISSING(c)
& NOT(MISSING(d))
& MISSING(e)
& MISSING(f)
& NOT(MISSING(g)).
COMPUTE x = 3.14.
ELSE IF respill=1 /* code 3.15 */
& NOT(MISSING(a))
& b GE 21 & b LE 33
& c GE 1 & c LE 31
& MISSING(d)
& MISSING(e)
& MISSING(f)
& NOT(MISSING(g)).
COMPUTE x = 3.15.
END IF.

I'll go over the documentation that you had suggested. I had tried similar
code in SAS and it had worked but haven't tried in any of the other
languages. Thanks again for the tip.

- Shahrukh

Hashmi, Syed S

Re: DO IF.. ELSE IF statements

In reply to this post by Hal 9000

Thanks to Melissa, Richard, Gary, Alex and anybody else who replied to
my DO IF...ELSE IF query. It's a lot clearer to me now and I might have
to either nest the statements like Melissa did or just go from least
restrictive to most restrictive. I'll have to keep all this in mind for
the next time I code. Thanks again all!!

- Shahrukh