I think I can t get what I want from number_format so I did this and it works fine :

 public function floatNumber($number)
 {
      $number_array = explode( . , $number);
      $left = $number_array[0];
      $right = $number_array[1];
      return number_format($number, strlen($right));
 }

thanx all for your replies.

Investigate the printf and sprintf functions instead of number_format.

They give the ability to format numbers as you wish.

printf("%d", $int) 

would be appropriate for a decimal integer.

printf("%4.2f", $float) 

would be appropriate for a floating point number with two decimal places.

number_format seems to be intended for internationalisation of currency output, but I don t think that s what you want because you mentioned decimal whole numbers.

I don t know a better way to do this, but you can compare the type, like the code below:

$number = 234.159;
if(is_float($number)) {
 echo number_format($number, 2);
}else {
 echo $number;
}

Whilst it s a bit quick and dirty, you could simply do a...

str_replace( .00 ,   , number_format($potentialFloat, 2));

Far from ideal, but effective.

I think the key problem is defining how many positions are needed. Would you define 13.01 as 13 because the first decimal was a 0? Since printf and number format needs you to know how many decimals, I don t know that that would work for you.

Maybe something like this (which is a lot of functions, but looks for the first 0, and then returns the truncated string). Yes, it is intensive, but it may be the best way for you.

function show_number($number, $max = 8){
  if(strpos($number,  . )){
    $decimal = strpos($number,  . );
    if(strpos($number,  .0 )){
      return substr($number, 0, $decimal);//returns whole if zero is first
    } else {
      if(strpos(substr($number, $decimal, $max),  0 )){
        $zero = strpos(substr($number, $decimal, $max),  0 );
        return substr($number, 0, $decimal+$zero);//returns number w/0 first zero
      } else {
        return substr($number, 0, $decimal+$max+1); //returns number with max places
      }
    }
  } else {
    return $number; //returns number if no decimals
  }
}

I use the following to get rid of trailing zeros:

// Trims a floating point number so there are no trailing zeros. For example:
// 1.00 -> 1
// 1.10 -> 1.1
function trimFloatingPoint($input) {
    $strNum = strval($input);

    while(substr($strNum, -1, 1) ==  0 )
        $strNum = substr($strNum, 0, strlen($strNum) -1);

    // Trailing zeros are gone. How about the trailing period?
    if(substr($strNum, -1, 1) ==  . )
        $strNum = substr($strNum, 0, strlen($strNum) -1);

    return $strNum;
}

$x = $x - floor($x)

will give you the right floating point...