Hi,    We have data in Redshift that contain floating point numbers (e.g. 0.924780).  My understanding is traditional SAS data step or PROC SQL would not be able to handle the data without rounding.  I was wondering if the data would be displayed in SAS datasets when using a simple query (i.e. where xxx = 0.924780) from SAS against Redshift using PROC FEDSQL or DS2?  We are currently using an AWS Redshift ODBC driver to connect between SAS and Redshift.   Also, I have difficulty connecting FEDSQL and DS2 to ODBC connections and keep getting PROC FEDSQL initialization failed errors, which are not very informative at all.   LIBNAME _all_ clear; PROC FEDSQL NOLIBS CONN="DRIVER=FEDSQL;CONOPTS=((DRIVER=ODBC;DB=REDSHIFT101;UID={&uid.};PWD={&pwd.};CATALOG=RS_PROD; SCHEMA=MY_SCHEMA))"; CREATE TABLE temp_gr AS SELECT * FROM RS_PRODC.MYREDSHIFT_DATA; QUIT;   NOTE: Connection string: NOTE: DRIVER=FEDSQL;CONOPTS= ( ( (DRIVER=ODBC;DB=REDSHIFT101;UID={*};PWD={*};CATALOG=RS_PROD;SCHEMA=MY_SCHEMA))) ERROR: ERROR: [ ERROR: PROC FEDSQL initialization failed.           Many thanks   G   ... View more