Dear Forum Members, 

I hope someone can help me out or point me in the right direction. I have a data file with 300 subjects and quite a few variables, most of them discretely coded 1-6. Here is what I would like to do:

1) if any variable of let's DP1, DP2, DP3, .... DP7 is 1, 3, or 5 make target_var1 = 1
2) if all of the above (DP1-DP7) are $sysmis then make target_var1 = $sysmis
3) if none of these are true make target_var = 0

Ideally, I would like to be able to adapt item 2) to something that says: "if more than i.e. 4 of DP1-7 contain $sysmis then make i.e. target_var 2 = $sysmis" or some other way to designate too few data points.

so finally, at the most complex I might also need to have something like:
IF in DP1-7 at least two have the value 4 and no more than i.e. 3 are $sysmis, then give target_var3 = 1


If anyone could help out, I would be very very grateful.

All the best,

J.

------------------------------
johannes achenbach
------------------------------

#SPSSStatistics

Original Message:
Sent: 11/2/2021 2:06:00 PM
From: johannes achenbach
Subject: conditional syntax; what is the best approach....

Dear Forum Members, 

I hope someone can help me out or point me in the right direction. I have a data file with 300 subjects and quite a few variables, most of them discretely coded 1-6. Here is what I would like to do:

1) if any variable of let's DP1, DP2, DP3, .... DP7 is 1, 3, or 5 make target_var1 = 1
2) if all of the above (DP1-DP7) are $sysmis then make target_var1 = $sysmis
3) if none of these are true make target_var = 0

Ideally, I would like to be able to adapt item 2) to something that says: "if more than i.e. 4 of DP1-7 contain $sysmis then make i.e. target_var 2 = $sysmis" or some other way to designate too few data points.

so finally, at the most complex I might also need to have something like:
IF in DP1-7 at least two have the value 4 and no more than i.e. 3 are $sysmis, then give target_var3 = 1


If anyone could help out, I would be very very grateful.

All the best,

J.

------------------------------
johannes achenbach
------------------------------

#SPSSStatistics

See if this does what you want.

DO REPEAT existVar=dp1 to dp7 /newVar=new1 TO new7.
COMPUTE newVar=0.
COMPUTE newVar=any(existVar,1,3,5).
END REPEAT.
COUNT miss=dp1 to dp7 (MISSING,SYSMIS).
EXECUTE.

DO IF miss=7.
RECODE new1 to new7 (LO THRU HI,MISSING=SYSMIS).
END IF.
EXECUTE.

The DO IF command could be modified from 7 to whatever number you want, or some range, etc.

------------------------------
Rick Marcantonio
Quality Assurance
IBM
------------------------------

Original Message:
Sent: Tue November 02, 2021 02:06 PM
From: johannes achenbach
Subject: conditional syntax; what is the best approach....

Dear Forum Members, 

I hope someone can help me out or point me in the right direction. I have a data file with 300 subjects and quite a few variables, most of them discretely coded 1-6. Here is what I would like to do:

1) if any variable of let's DP1, DP2, DP3, .... DP7 is 1, 3, or 5 make target_var1 = 1
2) if all of the above (DP1-DP7) are $sysmis then make target_var1 = $sysmis
3) if none of these are true make target_var = 0

Ideally, I would like to be able to adapt item 2) to something that says: "if more than i.e. 4 of DP1-7 contain $sysmis then make i.e. target_var 2 = $sysmis" or some other way to designate too few data points.

so finally, at the most complex I might also need to have something like:
IF in DP1-7 at least two have the value 4 and no more than i.e. 3 are $sysmis, then give target_var3 = 1


If anyone could help out, I would be very very grateful.

All the best,

J.

------------------------------
johannes achenbach
------------------------------

#SPSSStatistics

Thanks guys. @Rick Marcantonio
I couldn't quite get that to work because I didn't need 7 new variables but wanted to have one new one that contains the new information.

For people who might run into similar problems I did the following (probably not elegant and too long winded):

compute new_var= 0.
if any  (dp1, 1,3,5) new_var=1.
if any  (dp2, 1,3,5) new_var=1.
and so on til dp7.    [is there any way to combine the seven different if clauses???]

compute mis_var = nmiss(dp1 to dp7).
if (mis_var GE 5 AND new_var NE 1) new_var = $sysmis.
execute.​

looking forward to your feedback.....

all the best,

J.
​

------------------------------
johannes achenbach
------------------------------

Original Message:
Sent: Tue November 02, 2021 03:16 PM
From: Rick Marcantonio
Subject: conditional syntax; what is the best approach....

See if this does what you want.

DO REPEAT existVar=dp1 to dp7 /newVar=new1 TO new7.COMPUTE newVar=0.COMPUTE newVar=any(existVar,1,3,5).END REPEAT.COUNT miss=dp1 to dp7 (MISSING,SYSMIS).EXECUTE.DO IF miss=7.RECODE new1 to new7 (LO THRU HI,MISSING=SYSMIS).END IF.EXECUTE.

The DO IF command could be modified from 7 to whatever number you want, or some range, etc.

------------------------------
Rick Marcantonio
Quality Assurance
IBM

Original Message:
Sent: Tue November 02, 2021 02:06 PM
From: johannes achenbach
Subject: conditional syntax; what is the best approach....

Dear Forum Members, 

I hope someone can help me out or point me in the right direction. I have a data file with 300 subjects and quite a few variables, most of them discretely coded 1-6. Here is what I would like to do:

1) if any variable of let's DP1, DP2, DP3, .... DP7 is 1, 3, or 5 make target_var1 = 1
2) if all of the above (DP1-DP7) are $sysmis then make target_var1 = $sysmis
3) if none of these are true make target_var = 0

Ideally, I would like to be able to adapt item 2) to something that says: "if more than i.e. 4 of DP1-7 contain $sysmis then make i.e. target_var 2 = $sysmis" or some other way to designate too few data points.

so finally, at the most complex I might also need to have something like:
IF in DP1-7 at least two have the value 4 and no more than i.e. 3 are $sysmis, then give target_var3 = 1


If anyone could help out, I would be very very grateful.

All the best,

J.

------------------------------
johannes achenbach
------------------------------

#SPSSStatistics

Sorry I misunderstood. This will combine the 7 separate statements:

COMPUTE newVar=0.
DO REPEAT existVar=dp1 to dp7.
IF newVar=0 newVar=any(existVar,1,3,5).
END REPEAT.
EXECUTE.​

------------------------------
Rick Marcantonio
Quality Assurance
IBM
------------------------------

Original Message:
Sent: Sun November 07, 2021 04:27 PM
From: johannes achenbach
Subject: conditional syntax; what is the best approach....

Thanks guys. @Rick Marcantonio
I couldn't quite get that to work because I didn't need 7 new variables but wanted to have one new one that contains the new information.

For people who might run into similar problems I did the following (probably not elegant and too long winded):

compute new_var= 0.if any  (dp1, 1,3,5) new_var=1.if any  (dp2, 1,3,5) new_var=1.and so on til dp7.    [is there any way to combine the seven different if clauses???]compute mis_var = nmiss(dp1 to dp7).if (mis_var GE 5 AND new_var NE 1) new_var = $sysmis.execute.​

looking forward to your feedback.....

all the best,

J.
​

------------------------------
johannes achenbach

Original Message:
Sent: Tue November 02, 2021 03:16 PM
From: Rick Marcantonio
Subject: conditional syntax; what is the best approach....

See if this does what you want.

DO REPEAT existVar=dp1 to dp7 /newVar=new1 TO new7.COMPUTE newVar=0.COMPUTE newVar=any(existVar,1,3,5).END REPEAT.COUNT miss=dp1 to dp7 (MISSING,SYSMIS).EXECUTE.DO IF miss=7.RECODE new1 to new7 (LO THRU HI,MISSING=SYSMIS).END IF.EXECUTE.

The DO IF command could be modified from 7 to whatever number you want, or some range, etc.

------------------------------
Rick Marcantonio
Quality Assurance
IBM

Original Message:
Sent: Tue November 02, 2021 02:06 PM
From: johannes achenbach
Subject: conditional syntax; what is the best approach....

Dear Forum Members, 

I hope someone can help me out or point me in the right direction. I have a data file with 300 subjects and quite a few variables, most of them discretely coded 1-6. Here is what I would like to do:

1) if any variable of let's DP1, DP2, DP3, .... DP7 is 1, 3, or 5 make target_var1 = 1
2) if all of the above (DP1-DP7) are $sysmis then make target_var1 = $sysmis
3) if none of these are true make target_var = 0

Ideally, I would like to be able to adapt item 2) to something that says: "if more than i.e. 4 of DP1-7 contain $sysmis then make i.e. target_var 2 = $sysmis" or some other way to designate too few data points.

so finally, at the most complex I might also need to have something like:
IF in DP1-7 at least two have the value 4 and no more than i.e. 3 are $sysmis, then give target_var3 = 1


If anyone could help out, I would be very very grateful.

All the best,

J.

------------------------------
johannes achenbach
------------------------------

#SPSSStatistics