Conceptually, all data presented to plt must be organized in rows and columns. Columns are numbered beginning with zero, and each column contains values for a variable that can be used as an abscissa (x coordinate), ordinate (y coordinate), or (with appropriate options discussed later in this book) a grey level, color, or other plot attributes. Rows are numbered beginning with one, and each row contains a value for each column. Within a data file, values are always arranged in row-major order (all elements of row 1, followed by all elements of row 2, etc.).
plt can read data from a file or from a pipe. A command-line argument that is not otherwise recognized by plt is assumed to be the name of a data file. Any legal file name is acceptable; if it begins with a character that would cause plt to interpret it as an option, simply prefix the name with ``./''. If no data file is named, plt reads data from the standard input.