Usually, data must be in text form in order for plt to read them. Each non-empty, non-comment line (row) in the input should contain a value for each column that will be plotted; any additional values or other extra text at the end of a row will be ignored. Columns can be separated by any number of spaces or tabs. Commas and single or double quotation marks can also be used as column separators with current versions of plt, though not with older versions. It is not necessary to line up the values in each row. There may also be spaces or tabs at the beginning of a line, and these will also be ignored. For example, here is squares.data, a text data file that can be read by plt:
# This is a text data file. It doesn't need to begin with a
# comment, although this one does. Comments and empty lines
# are ignored in text data files.
-1 1 # This is row 1. It contains two columns.
0 0 # This is row 2. Columns don't need to be lined up neatly.
1 1 # Row 3. Whitespace separates columns and is otherwise ignored.
1.5 2.25# Row 4. Most common formats for base 10 numbers are acceptable.
# Note, however, that only "." can be used to separate
# the integer and fractional parts ("," is treated as a
# column separator).
2 4 # Row 5 (the empty and comment lines aren't counted.)
2.5 6.25 8 # The "8" is ignored, since row 1 has only 2 columns.
3 9
4,16 # Comma-separated value (CSV) format is acceptable.
"5","25"# Quoted CSV format is also acceptable.
6 3.6e1 # Standard notation for floating-point numbers is acceptable.
7 490e-1# Here's another, with a negative exponent.
8 64.
9 +81.0 # The "+" is redundant, but acceptable.
10 1e2
# There can be more comments such as these at the end ....