@Tom Thanks a lot. I'll  have a look at the use of array to convert data from wide to long. I've used proc transpose (twice) to achieve this, but it's good to know that there's the other way:   proc transpose data=lbs out=lbs1; by Frequency Measure Balance_sheet_position Type_of_instruments Currency_denomination Currency_type_of_reporting_coun Parent_country Type_of_reporting_institutions Reporting_country Counterparty_sector Counterparty_country Position_type Time_Period; var n_:; run; proc sort data=lbs1; by Frequency Measure Type_of_instruments Currency_denomination Currency_type_of_reporting_coun Parent_country Type_of_reporting_institutions Reporting_country Counterparty_sector Counterparty_country Position_type Time_Period _NAME_; run; proc transpose data=lbs1 out=lbs2; by Frequency Measure Type_of_instruments Currency_denomination Currency_type_of_reporting_coun Parent_country Type_of_reporting_institutions Reporting_country Counterparty_sector Counterparty_country Position_type Time_Period _NAME_; id Balance_sheet_position; var col1; run; As regards your suggestion to look at the data first - I fully agree. I knew that first 7 lines were the "unnecesary" description of the dataset. I just din't know how to handle the problem of getting the variable names from the 8th line. I know I can always copy or type them by hand, but this would require modifying the code every quarter, as with every update, a new column (variable) appears. That's why I like using dictionary as sugeested by @Kurt_Bremser But, once again, thanks for help and very usefull material to study the secrets of sas ... View more

@Kurt_Bremser Many thanks! ... View more

@Kurt_Bremser Thanks a lot! One question rearding this part:   tail -n +8 full_WEBSTATS_LBS_D_PUB_DATAFLOW_csv.csv>input.csv I uderstand this is the Unix command, not SAS command, am I right? Can I somehow do it using SAS? If possible I want to avoid using another software, other than SAS ... View more

Hi,   Thanks for the answer.   1. I have only a link to the zip file. I've tried to attach the csv, but it has over 140 MB and it gets rejected. 2. I would also prefer the data in the long format, but this is what I have. The BIS won't change the format for me. Once I get the data into SAS, I'll transpose them in whatever way I want. But I need to have them in SAS in the first place. 3. I know the structure of the data - I can open the csv and check column names. But, still, with using infile I'll have to re-write the code every quarter to add new column get the new data. This is something I want to avoid. That's why I was hoping there is a workaround for this. ... View more

Thanks a lot. It's working now ... View more

Thanks, it works now. ... View more

@Vince_SAS   Sas 9.4 M3 + Windows 7 64 bit ... View more

@Vince_SAS I 've tried. The result is:    ERROR: Internal error: unable to communicate with PC Files Server process. ERROR: Error in the LIBNAME statement ... View more

@error_prone Default name ("Sheet1"). I've tried to import other files where sheets are named differently. Every time, datasets appear in the library but I cannot open them ... View more

@gamotte   Thanks. That's another approach. I promise to myself that one day I'll learn Pearl regular expressions in SAS 🙂  ... View more

Thanks. I did both and it works now! ... View more