In contrast to the CSV files seen earlier, a flat file does not contain any separator between the fields. Since there is no separator, all records in a flat file are usually of the same length, as the length of columns is the only way of separating data. Prior to the advent of spreadsheet programs, it was a common practice to use only flat files. Flat files are still used according to the preference of the authors.
In this example, we will be using Python to read in a flat file. The pandas library of routines includes a function to read flat files, read_fwf. Your Python script passes in the column widths and names to read_fwf, and the function returns a DataFrame.
Of course, now that I am looking for a flat file, I can't find one! I took the first 20 records of the preceding baseball data and stored that in a flat file, baseball.txt. There is no header record. Only the first several columns are available. It looks like this:
