04-05-2016 10:37 AM
Dear SAS experts,
which ist the easiest way to import a csv specifying:
- in wich row are the headers
- in which row start the data
I tried in several ways (dataeow, firstobs...) But I do not gest what i want.
proc import datafile='test.csv'
Attached an example of the csv input and the data that I would like to get in SAS.
04-05-2016 10:45 AM
Reading variable names from the CSV requires that the name be on a single row. Easiest is to Edit the CSV to remove the entire first line with the column headers like "Platin/EURO".
Then datarow=2 should work fine.
04-05-2016 11:30 AM
Use a data step to write a new copy of the file with the record(s) edited as you see fit. Then run your PROC IMPORT on that.
04-05-2016 11:47 AM
thank but this is what I am currently doing and I would not.
So means that SAS cannot do it?
Show the exact code you run, the Log results and show the data as a TXT file not Excel (it may change values with out telling you).
You should always use a fully qualified path to point to input and out file locations : datafile = "C:\somefolder\otherfolder\inputfile.csv" for example instead of datafile="test.csv" as the later is going to read from what SAS considers the current active folder and is very likely not the location you expect.
04-05-2016 11:52 AM
For a file with three variable just write the data step yourself. Probably will take less code than the the PROC IMPORT statements.
data want ; infile "myfile.csv" dsd firstobs=3 truncover ; length date field1 field2 8 ; informat date mmddyy10.; format date yymmdd10.; input date field1 field2; run;