The most common file format for datasets is a comma separated value (CSV) file. A CSV may have a header record followed by a variable number of data records.
The header record may be the first record in the file. In that record, the separated values are headings or column names for each of the columns of data in the file. The column names are all character string values. We can use these column names for variable names in our scripts, corresponding to column names in a dataset.
Each subsequent data record will have a separated value in that record for every column. The value may be an empty string or no value, but the comma separation of the record will correspond to the columns in the header record.
If there is no header record, you may have to find out what the column layout is for the file. There is normally a descriptor in the same location as the CSV file that describes each of the columns. In this case, you have to manually assign column names to your working dataset...
