Macro variable: Apparent symbolic reference not resolved

[Babloo]

I'm trying to create a new numeric variable PODVER via case when statement in DI studio as follows. When I exceute this code, I got error like 'Apparent symbolic reference not resolved'. reporting_date is a macro variable and it has values like 02JAN2018, 12OCT2019.... Could you please help me to resolve the issue?

(case 
  when NEW_RENEWAL_DAY_MONTH_YEAR=REPORTING_DATE_ddmon 
   and KNKATD='7'  
      then input("&reporting_date.",date9.)
   when input(cats(NEW_RENEWAL_DAY_MONTH_YEAR,"2018"),date9.) > today() 
     then input(cats(NEW_RENEWAL_DAY_MONTH_YEAR,"2018"),date9.) 
     else input(cats(NEW_RENEWAL_DAY_MONTH_YEAR,"2019"),date9.) end) as PODVER length = 8 format = DATE9.

Log:

1595 (case when NEW_RENEWAL_DAY_MONTH_YEAR=REPORTING_DATE_ddmon and KNKATD='7' then
1595 ! input("&reporting_date.",date9.)
WARNING: Apparent symbolic reference REPORTING_DATE not resolved.
1596 when input(cats(NEW_RENEWAL_DAY_MONTH_YEAR,"2018"),date9.) > today() then
1596 ! input(cats(NEW_RENEWAL_DAY_MONTH_YEAR,"2018"),date9.)
1597 else input(cats(NEW_RENEWAL_DAY_MONTH_YEAR,"2019"),date9.) end) as PODVER length = 8
1598 format = DATE9.
1599 from &SYSLAST

SYMBOLGEN: Macro variable SYSLAST resolves to WORK.W728Z2TN 
WARNING: Apparent symbolic reference REPORTING_DATE not resolved.

[PaigeMiller]

You are using a macro variable in your code (specifically &reporting_date), but no such macro variable exists.

[Babloo]

It is there. It is a global macro variable which I created at the beginning in user written transformation.

[RW9]

I am afraid it is not:

WARNING: Apparent symbolic reference REPORTING_DATE not resolved.

That warning is clearly telling you that no macro variable REPORTING_DATE is present in scope at the point that code is run. 

Possibly its not in scope, don't know how DI integrates various components.

[PaigeMiller]

---

@Babloo wrote:

It is there. It is a global macro variable which I created at the beginning in user written transformation.

---

When I have the choice between believing a programmer who says he did something, or believing SAS, which says that the programmer didn't do that thing, I believe SAS.

You may think you have created a global macro variable by that name, but you have not done it properly.

[Kurt_Bremser]

---

@Babloo wrote:

It is there. It is a global macro variable which I created at the beginning in user written transformation.

---

No, it isn't. In a dispute, SAS is always right, the programmer never. I can tell you this from 20 years experience.

[ballardw · Accepted solution]

---

@Babloo wrote:

It is there. It is a global macro variable which I created at the beginning in user written transformation.

---

Put this code in your program immediately before the part with the problematic macro variable:

%put _global_;

That will display currently defined global macro variables and the values.

In addition to comments about existence of the variable this may be suboptimal coding in general:

then input("&reporting_date.",date9.)

unless you do not trust what ever is supposed to be making &reporting_date to be proper text in data9 format such as "18SEP2018".

Simpler would be

then "&reporting_date."d

[ballardw]

In addition to comments about existence of the variable this may be suboptimal coding in general:

then input("&reporting_date.",date9.)

unless you do not trust what ever is supposed to be making &reporting_date to be proper text in data9 format such as "18SEP2018".

Simpler would be

then "&reporting_date."d

[miriamjamison]

I know that this is a pretty old topic at this point but I was having the same problem and I resolved it myself.

I realized that I had not actually RUN my %let statements before attempting to use them further down in the program.