ERROR: Data set WORK.SORT is not sorted in ascending sequence. The current BY group has (...)

[PatrykSAS]

Hello. I have a problem with the title error. What I've done:

PROC SORT data=WORK.SORT;
BY OWNER TARYFA PRODUKT_TARYFA;
run;
PROC MEANS data=WORK.SORT;
BY OWNER TARYFA PRODUKT_TARYFA;
VAR WOLUMEN_MWH_KONWERSJA;
OUTPUT OUT=WORK.TEST Q1= Q3= /autoname;
RUN;

After the above code, I get an error:

ERROR: Data set WORK.SORT is not sorted in ascending sequence. The current BY group has OWNER = Deweloperzy, Wspólnoty i and the next BY group has OWNER = DEW.

Replacing "by" with "class" or adding "notsorted" wouldn't solve problem due to not correct calculations.

Do you have any suggestions?

[data_null__]

Show the log.

[Astounding]

Only an obscure set of conditions could produce this result.

You would see this easily by reading the log.  There should be some sort of message about PROC SORT being skipped.

Always read the log.  Always show the log if you expect help instead of a guess.

[AMSAS]

Appears to work fine with some simple test data

data have ;
	input owner $ ;
	do n=1 to 5 ;
		output ;
	end ;
cards ;
Deweloperzy
DEW
;

proc sort data=have out=srtd ;
	by owner ;
run ;

proc means data=srtd ;
	by owner ;
	var n ;
	output out=WORK.have Q1= Q3= /autoname;
RUN;

Log:

68   data have ;
69       input owner $ ;
70       do n=1 to 5 ;
71           output ;
72       end ;
73   cards ;

NOTE: The data set WORK.HAVE has 10 observations and 2 variables.
NOTE: DATA statement used (Total process time):
      real time           0.01 seconds
      cpu time            0.01 seconds


76   ;
77
78   proc sort data=have out=srtd ;
79       by owner ;
80   run ;

NOTE: There were 10 observations read from the data set WORK.HAVE.
NOTE: The data set WORK.SRTD has 10 observations and 2 variables.
NOTE: PROCEDURE SORT used (Total process time):
      real time           0.00 seconds
      cpu time            0.00 seconds


81
82   proc means data=srtd ;
83       by owner ;
84       var n ;
85       output out=WORK.have Q1= Q3= /autoname;
86   RUN;

NOTE: There were 10 observations read from the data set WORK.SRTD.
NOTE: The data set WORK.HAVE has 2 observations and 5 variables.
NOTE: PROCEDURE MEANS used (Total process time):
      real time           0.02 seconds
      cpu time            0.01 seconds

Can you create a simple input "have" dataset to replicate the issue and provide the code and log