Hey Filipvdr, If you do not specify a width= statement in your define statement, the column headline will break after the longest value in the column. E.g. the variable sex in sashelp.class dataset has a maximum length of 1, so the column length is one. If you add a label you need to define the column width like the width of the label. (Even the unlabeled variable name sex needs a width option to be displayed correctly.) See the following comparision: proc report data=sashelp.class;    column name sex age weight height;    define sex / 'Gender'; run; Sex in the basic dataset is either 'M' or 'F' with the maximum width of 1. So the column header has a width of 1. proc report data=sashelp.class;    column name sex age weight height;    define sex / display 'Gender' width=6; run; The assigned label 'Gender' has a width of 6 characters and with the corresponding width option you can expand the header. Niels ... View more

Hey Milena, I tried my best but I failed. :smileyconfused: I would export the dataset by PROC EXPORT and build the table in EXCEL.. The simple export is: DATA TEST;     SET SASHELP:CLASS; RUN; PROC EXPORT DATA= TEST;      OUTFILE ="YOUR PATH:::\TEST.XLS"      DBMS = EXCEL REPLACE;      SHEET = "EXPORT SAS TO EXCEL"; RUN; Greetz Niels ... View more

** Alternative;   Data prod;     input product_id product $ country $;     datalines;   1001 screws US   1002 nails D   1003 wood GB   1004 iron LT   1001 screws US   1002 nails D   1003 wood GB   1004 iron LT   ; run; proc sort data=prod;   by product_id; run; data overview (drop= countid countprod countcountry);   set prod;       countID=0;    if product_id ne . then       totalID = countID+1;       countPROD=0;    if product ne '' then       totalPROD = countPROD+1;    countcountry =0;    if country ne '' then      totcountry = countcountry+1;    label totalid = "Sum of unique product Id's"    totalprod = "Sum of unique products"    totcountry ="Sum of unique countries"; run; options nobyline; title "Accumulated counts for #BYVAL(product_id)"; proc freq data=overview ;   by product_id;   table totalid totalprod totcountry / nocum nopercent; run; ... View more

Hi Allan, I think a very smart way to count the appearance of a lot of variables and values is to use the ACROSS option in a proc report. Please take a look at the program code: ** Sample data with 3 variables wich contain repeated values ; data prod;   input product_id product $ country $;   datalines; 1001 screws US 1002 nails D 1003 wood GB 1004 iron LT 1001 screws US 1002 nails D 1003 wood GB 1004 iron LT ; run; proc report data = prod nowd headline headskip;    column product_id product country;    define product_id  / order across;    define product      / display across;    define  country     / display country; run; As you can see in the output window all frequencies of values were accumulated. Or do you also need the counted values in the dataset? Not only a report? ... View more