rewriting a program using Format

[Armand_B]

Hello

i m trying to rewrite a solution( which i found not aesthetic)  here is the solution

libname cert xlsx "/home/u49144401/Cert/heart.xlsx";

data work.heart;
set cert.heart(drop=AgeAtDeath DeathCause);
where Status='Alive';
if AgeCHDiag=. then delete;
length Smoking_Status $17;

if 0<=Smoking<6 then Smoking_Status='Non-smoker(0-5)';

if 6<=Smoking<=15 then Smoking_Status='Moderate(6-15)';

if 16<=Smoking<25 then Smoking_Status='Heavy(16-25)';

if Smoking>25 then Smoking_Status='Very Heavy(16-25)';

else Smoking_Status='Error';
run;

proc freq data=work.heart;
tables AgeCHDiag*Smoking_status/norow nocol nopercent;
run;

Following is the code i m trying to write...But i have spent more than 2 hours to try to understand what is wrong...please help
Regards

proc format;
value smoking
0 -<6 ="None(0-5)"

6- 15 ="Moderate(6-15)"
16-25 ="Heavy(16-25)"
26 - high= "Very Heavy(>25)"
other ="Error";
run;

libname cert xlsx "/home/u49144401/Cert/heart.xlsx";
data work.hearts;
set cert.heart(drop=AgeAtDeath DeathCause);
where Status='Alive';
if AgeCHDiag=. then delete;
length Smoking_Status $17;
format Smoking_Status smoking.;
run;

proc freq data=work.hearts;
tables AgeCHDiag*Smoking_Status/norow nocol nopercent;
format Smoking_Status smoking.;
run;

LOGFILE ERROR: You are trying to use the numéric format SMOKING with the character variable Smoking_Status in data set WORK.HEARTS

[PaigeMiller]

Smoking_status is a character variable. You would have to use a character format on it, but the format is a numeric format. You can't apply numeric formats to character variables.

What you really want to do is to apply format SMOKING to variable SMOKING (and get rid of variable SMOKING_STATUS).

[Armand_B]

will you explain with code i would appreciate

[ghosh]

You dont need to but if you must create the char variable, replace the format statememt with

Smoking_Status=(put (smoking,smoking.));

[Armand_B]

Sorry i dont understand your suggestion

I add it at the proc freq and at  data  i end up with 2 errors

[ghosh]

please show your code and log

[Armand_B]

68

69 proc format;

70 value smoking

71 0 -<6 ="None(0-5)"

72 6- 15 ="Moderate(6-15)"

73 16-25 ="Heavy(16-25)"

74 26 - high= "Very Heavy(>25)"

75 other ="Error";

NOTE: Format SMOKING has been output.

76 run;

NOTE: PROCEDURE FORMAT a utilisé (Durée totale du traitement) :

real time 0.00 seconds

user cpu time 0.01 seconds

system cpu time 0.01 seconds

memory 302.68k

OS Memory 23968.00k

Timestamp 05/01/2022 08:58:12 PM

Step Count 61 Switch Count 0

Page Faults 0

Page Reclaims 41

Page Swaps 0

Voluntary Context Switches 0

Involuntary Context Switches 0

Block Input Operations 0

Block Output Operations 40

77

78 libname cert xlsx "/home/u49144401/Cert/heart.xlsx";

NOTE: Libref CERT was successfully assigned as follows:

Engine: XLSX

Physical Name: /home/u49144401/Cert/heart.xlsx

79 data work.hearts;

80 set cert.heart(drop=AgeAtDeath DeathCause);

81 where Status='Alive';

82 if AgeCHDiag=. then delete;

83 length Smoking_Status $17;

WARNING: Length of character variable Smoking_Status has already been set.

Use the LENGTH statement as the very first statement in the DATA STEP to declare the length of a character variable.

84 Smoking_Status=(put (smoking,smoking.));

85 run;

NOTE: Variable AgeCHDiag is uninitialized.

NOTE: The import data set has 5209 observations and 17 variables.

NOTE: There were 3218 observations read from the data set CERT.heart.

WHERE Status='Alive';

NOTE: The data set WORK.HEARTS has 0 observations and 16 variables.

NOTE: DATA statement a utilisé (Durée totale du traitement) :

real time 0.74 seconds

user cpu time 0.74 seconds

system cpu time 0.00 seconds

memory 2971.40k

OS Memory 26532.00k

Timestamp 05/01/2022 08:58:13 PM

Step Count 62 Switch Count 2

Page Faults 0

Page Reclaims 685

Page Swaps 0

Voluntary Context Switches 18

Involuntary Context Switches 1

Block Input Operations 0

Block Output Operations 264

86

87 proc freq data=work.hearts;

88 tables AgeCHDiag*Smoking_Status/norow nocol nopercent;

89 Smoking_Status=(put (smoking,smoking.));

______________

180

ERROR 180-322: Statement is not valid or it is used out of proper order.

90 run;

NOTE: The SAS System stopped processing this step because of errors.

NOTE: PROCEDURE FREQ a utilisé (Durée totale du traitement) :

real time 0.00 seconds

user cpu time 0.00 seconds

system cpu time 0.00 seconds

memory 445.53k

OS Memory 24740.00k

Timestamp 05/01/2022 08:58:13 PM

Step Count 63 Switch Count 0

Page Faults 0

Page Reclaims 50

Page Swaps 0

Voluntary Context Switches 0

Involuntary Context Switches 0

Block Input Operations 0

Block Output Operations 0

[ghosh]

Smoking_Status=(put (smoking,smoking.));
Only in the data step and not in the proc. I'm assuming you are creating
this var to include in your excel file, otheri you don't it

[Armand_B]

I am ending up with 2 errors

data work.hearts;

79 set work.heart(drop=AgeAtDeath DeathCause);

ERROR: The variable AgeAtDeath in the DROP, KEEP, or RENAME list has never been referenced.

ERROR: The variable DeathCause in the DROP, KEEP, or RENAME list has never been referenced.

80 where Status='Alive';

81 if AgeCHDiag=. then delete;

82

83 length Smoking_Status $17;

84 smoking_Status=(put(smoking,smoking.));

85 run;

NOTE: The SAS System stopped processing this step because of errors.

WARNING: The data set WORK.HEARTS may be incomplete. When this step was stopped there were 0 observations and 3 variables.

WARNING: Data set WORK.HEARTS was not replaced because the step was stopped

[Reeza]

That log is not from the code I posted. Post the log from the code I posted.

[Reeza]

Commenting and formatting your code would really really help you here. 

You can use SAS autoformat options at the least to help make your code legible. 

*creates a format for a numeric variable;

proc format;
	value smoking 0 -<6="None(0-5)" 6- 15="Moderate(6-15)" 16-25="Heavy(16-25)" 
		26 - high="Very Heavy(>25)" other="Error";
run;

*creates a library to read in data;
libname cert xlsx "/home/u49144401/Cert/heart.xlsx";
*creates a data set called hearts;

data hearts;
	*reads in heart data set;
	set cert.heart(drop=AgeAtDeath DeathCause);
	*keeps only records where people are alive;
	where Status='Alive';
	*deletes records where missing Age of diagnosis - could be included in the WHERE clause for more efficient code;

	if AgeCHDiag=. then
		delete;
	*Your variable is smoking - not smoking status;
	*applies format smoking to your smoking variable;
	format Smoking smoking.;
	*creates smoking_status variable as a character variable;
	*this converts the variable. The format is also applied so this step is optional;
	Smoking_Status=put(smoking, smoking.);
run;

title 'Smoking Variable';

proc freq data=hearts;
	tables AgeCHDiag*Smoking / norow nocol nopercent;
run;

title 'Smoking Status Variable';

proc freq data=hearts;
	tables AgeCHDiag*Smoking_STATUS / norow nocol nopercent;
run;

[Armand_B]

Thanks for your time it is well written and with comments  but.......we end up with the same issue

Log file message

NOTE: Variable AgeCHDiag is uninitialized.

NOTE: The import data set has 5209 observations and 17 variables.

NOTE: There were 3218 observations read from the data set CERT.heart.

WHERE Status='Alive';

NOTE: The data set WORK.HEARTS has 0 observations and 16 variables.

NOTE: No observations in data set WORK.HEARTS.

PS 

how comes  ? what went wrong ?

[Reeza · Accepted solution]

The variable is ageCHDDiag - you were missing a D in your code and without checking the actual data set we couldn't know that. 

proc format;
	value smoking 0 -<6="None(0-5)" 6- 15="Moderate(6-15)" 16-25="Heavy(16-25)" 
		26 - high="Very Heavy(>25)" other="Error";
run;

*creates a library to read in data;
libname cert xlsx "/home/u49144401/Cert/heart.xlsx";
*creates a data set called hearts;

data hearts;
	*reads in heart data set;
	set sashelp.heart(drop=AgeAtDeath DeathCause);
	*keeps only records where people are alive;
	where Status='Alive';
	*deletes records where missing Age of diagnosis - could be included in the WHERE clause for more efficient code;

	if AgeCHDdiag=. then
		delete;
	*Your variable is smoking - not smoking status;
	*applies format smoking to your smoking variable;
	format Smoking smoking.;
	*creates smoking_status variable as a character variable;
	*this converts the variable. The format is also applied so this step is optional;
	Smoking_Status=put(smoking, smoking.);
run;

title 'Smoking Variable - with MISSING';

proc freq data=hearts;
	tables AgeCHDDiag*Smoking / norow nocol nopercent missing;
run;

title 'Smoking Status Variable - with ERROR';

proc freq data=hearts;
	tables AgeCHDDiag*Smoking_STATUS / norow nocol nopercent;
run;

[Reeza]

It's worth noting both your errors were due to incorrectly referencing variable names - in one case referring to the wrong variable and in the other misspelling a variable name.

[Armand_B]

Thanks for your comittment...Now the solution work very well

a) i had the SAS Prep Guide it has typing errors.....so i forgot to list the dataset in order to see whether names were properly written

b) i spent 3 hours thinking about rewriting the solution....with your help i was successful

Thanks for saving me this worry

my regards