Info

Note that some of the listed alternatives to Math.floor() use bitwise operators, which convert numbers to a 32-bit sequence.

These alternatives will only work with positive signed 32-bit floats, i.e. numbers from 0 to +2,147,483,647 (2^31-1).

~~2147483647.1; // 2147483647
~~2147483648.1; // -2147483648

Preparation code

<script>
  var n = Math.PI,
       // 3.141592653589793
      floorNew = (function() {
    var f = Math.floor,
        cap = 2 ^ 31;
    return function(n) {
     return n >= 0 && n < cap ? ~~n : f(n);
    }
   })();
</script>

Test runner

Warning! For accurate results, please disable Firebug before running the tests. (Why?)

Java applet disabled.

Math.floor(n); // 3

parseInt(n, 10); // 3

~~n; // 3

n | n; // 3

n | 0; // 3

n & n; // 3

n << 0; // 3

floorNew(n)

Compare results of other browsers

Revisions

You can edit these tests or add even more tests to this page by appending /edit to the URL. Here’s a list of current revisions for this page:

• Revision 1: published by Mathias Bynens on 31st July 2010 and last updated on 1st August 2010
• Revision 2: published on 1st August 2010
• Revision 3: published on 1st August 2010 and last updated on 15th August 2010
• Revision 4: published on 2nd August 2010 and last updated on 15th August 2010
• Revision 5: published by Jordan on 3rd August 2010
• Revision 6: published by NilColor on 13th January 2011
• Revision 11: published on 12th January 2012
• Revision 12: published by Jack Rugile on 26th February 2012
• Revision 13: published on 1st March 2012
• Revision 14: published on 4th April 2012
• Revision 15: published on 15th November 2012
• Revision 18: published by Luin on 26th February 2013

0 comments