Input and Output

Created

Output

Simplest way to output is to leave off the semicolon. Which can be fomratted with format [long|short].

Use disp to output a matrix using default formats.

  • Mix values and text into one output with num2str or int2str to convert numbers to strings.
  • e.g. disp(['The value of pi is' num2str(pi)]);

sprintf can format strings without sending them to a file. Sends an an output and has the same syntax as fprintf, except no fid is required.

Input

The input command obtains a value from the user. The input will be stored as a string or number, or enforce a value with

x = input('Please enter a number.');
y = input('Please enter something.', 's');

Data to Files

Save data in the workspace by using the save command.

save filename var1 var2 var3 options
  • You can select the variables you want to save, or omit variable list to save all.
  • the flag -regexp provides a "regular expression" that determines the variables to save. E.g., -regexp my saves all variables with names that starts wiht my like mylist and mynum.


Load data back into memory with the load command.

load filename options var1 var2
  • -ascii and -mat (default) change the file type to be saved as
  • -append appends to the file


Data files can come as a formatted character data which sotres in ASCII format or binary data that is directly copied from memory in the computer's native file.

  • Formatted data can easily be displayed and inspected by different computers but it is a waste of space
  • Binary data is not easy to display but it is faster.


Reference a particular file through a file's file ID. or fid. Obtain this ID by using fopen.

fid = fopen(filename, permission, format)
[fid, message] = fopen(filename, permission, format)
  • format is optional and should only be used when transferring data between computers.
  • add t if the file is going to be text and b if the file is binary
  • fid will be -1 if there is a problem opening the file and there will be a message returned.


Close a file with fclose

status = close('all')
status = close(fid)
  • 'all' closes every file except for standard output and input
  • status is 0 if there is no problem and -1 if there is an error.


Formatted output with fprintf function. It returns the number of bytes written:

fprintf('The value of pi is %6.2g \n', pi );
  • format specifier like%-12.5e includes % the marker*, - modifier, 12 field width (how many spaces to leave before printing the number), .5 precision, and e format descriptor*. Where any marked with * is required.

\

File to Matlab

Read a file:

 C = textscan( fid, 'format', N );

where N is the number of lines to read, 'format' is a string describing the type of data in each column. Example usage:

myf=fopen('class.data');  
Cdata = textscan( myf, '%s %d');
fclose(myf);

textscan also works with parsing strings


These functions reads a line from the file.

line = fgetl(fid)
line = fgets(fid)
  • fgetl excludes the end-of-line characters, while fgetsincludes them.
  • If the file ends then line is returned with the value -1.

Check files

exist('item') looks for a variable, built-in function, directory or file called item and returns a code.

exist('item','kind') only looks for items of type 'kind' where this can be 'var', 'file', 'builtin', or 'dir'.

Codes:

  • 0: not found
  • 1: variable
  • 2: M-file
  • 3: MEX file
  • 4: MDL file
  • 5: function
  • 6: P file
  • 7: directory
  • 8: Java Class

[message, errnum] = ferror(fid) returns the error message and error number.

  • if there are no errors, then errnum is 0 and the message is '..'. errnum output is optional.

eofstat = feof(fid) checks if you have reached the end of the file. 0 means the end file has not been reached while 1 means you have reached the end of the file.

Positioning Controlling where you are reading and writing: position=ftell(fid) determines the position in bytes frewind(fid) resets the file's position to the beginning status = fseek(fid, offset, origin) repositions the indicator to the offset byte location relative to the origin. origin can be bof beginning of file, cof current position, or eof end of file.

Linked from Computational Modeling