This question already has an answer here:
I am trying to get the first decimal place of a float number as an integer by subtracting the integer part, multiplying the remainder with 10 and then casting the result to int or using intval(). I noticed that the result for numbers with x.1 is correctly 1 as float, but after converting it to integer, it becomes sometimes 0, sometimes 1.
I tried to test it with numbers from 1.1 to 9.1:
for ($number = 1; $number < 10; $number++) {
$result = 10 * ($number + 0.1 - $number);
echo "<br/> number = " . ($number + 0.1) . ", result: ";
var_dump($result);
$result_int = intval($result);
var_dump($result_int);
}
Starting with 4.1 as input, the 1 oddly gets converted to 0:
number = 1.1, result: float(1) int(1)
number = 2.1, result: float(1) int(1)
number = 3.1, result: float(1) int(1)
number = 4.1, result: float(1) int(0)
number = 5.1, result: float(1) int(0)
number = 6.1, result: float(1) int(0)
number = 7.1, result: float(1) int(0)
number = 8.1, result: float(1) int(0)
number = 9.1, result: float(1) int(0)
Why at 4.1? That doesn't make any sense to me. Can anyone give me a hint what I am doing wrong?
PS: also tested at http://ideone.com/hr7M0A
</div>
You are seeing these results because floating point arithmetic is not perfectly accurate.
Instead of trying to manually get the first decimal point use fmod
:
$result = substr(fmod($number, 1) * 10, 0, 1)
My php is a bit rusty, so my syntax in probably off, but shouldn't it be simpler to convert to string and take the rightmost digit ?
sprintf($Str, "%.1f", $number);
$digit=$Str[strlen($Str)-1]; // Last digit