PHP Doku:: Logarithmus - function.log.html

Verlauf / Chronik / History: (1) anzeigen

Sie sind hier:
Doku-StartseitePHP-HandbuchFunktionsreferenzMathematische ErweiterungenMathematische FunktionenMathematische Funktionenlog

Ein Service von Reinhard Neidl - Webprogrammierung.

Mathematische Funktionen

<<log1p

max>>

log

(PHP 4, PHP 5)

logLogarithmus

Beschreibung

float log ( float $arg [, float $base = M_E ] )

Berechnet den natürlichen Logarithmus von arg beziehnugsweise den Logarithmus von arg zur Basis base falls der optionale zweite Parameter base gegeben ist.

Parameter-Liste

arg

Der zu verarbeitende Wert

base

Optionale Basis für die Berechnung des Logarithmus, der Vorgabewert ist 'e' für den natürlichen Logarithmus.

Rückgabewerte

Der Logarithmus von arg zur Basis base bzw. der natürliche Logarithmus

Changelog

Version Beschreibung
Ab 4.3.0 Der optionale Parameter base wurde hinzugefügt. Mit älteren Versionen ist die Berechnung des Logarithmus eines Wertes n zur Basis b mit Hilfe folgender Formel möglich: logb(n) = log(n)/log(b), wobei log hier der natürliche Logarithmus ist.

Siehe auch

  • log10() - Dekadischer Logarithmus
  • exp() - Exponentialfunktion
  • pow() - Potenzfunktion


7 BenutzerBeiträge:
- Beiträge aktualisieren...
olafurw [at] gmail.com
4.11.2008 13:56
For those interested. Works with older than 4.3 versions.

<?php
   
function byteConvert($bytes)
    {
       
$s = array('B', 'Kb', 'MB', 'GB', 'TB', 'PB');
       
$e = floor(log($bytes)/log(1024));
     
        return
sprintf('%.2f '.$s[$e], ($bytes/pow(1024, floor($e))));
    }
?>
dingus_76 at hotmail dot com
16.05.2008 4:41
well i been pulling my hair out trying to get log to work with big numbers and i ended up writing a bclog function so to save everyone else the stress here it is

<?php
function bclog($X,$base=10,$decimalplace=12){
   
$integer_value=0;
    while(
$X < 1){
       
$integer_value = $integer_value - 1 ;
       
$X = bcmul($X , base);
    }
    while(
$X >= $base){
       
$integer_value = $integer_value + 1;
       
$X = bcdiv($X , $base );
    }
   
$decimal_fraction = 0.0;
   
$partial = 1.0;
   
# Replace X with X to the 10th power
   
$X = bcpow($X , 10);
    while(
$decimalplace > 0){
       
$partial = bcdiv($partial , 10);
       
$digit=0;
        while(
$X >= $base){
             
$digit = $digit + 1;
             
$X = bcdiv($X , $base);
        }
       
$decimal_fraction = bcadd($decimal_fraction , bcmul($digit , $partial));
       
# Replace X with X to the 10th power
       
$X = bcpow($X , 10);
       
$decimalplace = $decimalplace - 1 ;
    }
    return
$integer_value + $decimal_fraction ;
}
?>
Ulf Wostner
6.08.2006 14:56
<?php

#--------------------------------------------------------
#     How many digits does an integer have?
#--------------------------------------------------------

function digit_count($n, $base=10) {

  if(
$n == 0) return 1;

  if(
$base == 10) {
   
# using the built-in log10(x)
    # might be more accurate than log(x)/log(10).
   
return 1 + floor(log10(abs($n)));
  }else{
   
# here  logB(x) = log(x)/log(B) will have to do.
  
return 1 + floor(log(abs($n))/ log($base));
  }
}

# Example:  How many decimal digits for 2 to the power 24?
echo digit_count(pow(2, 24));

# Example: How many bits to write 1 billion in binary, last century?

if($country_code  == 'US') echo digit_count(pow(10, 9), 2);
if(
$country_code == 'UK') echo digit_count(pow(10, 12), 2);

#--------------------------------------------------------
#     Using log to format columns.
#--------------------------------------------------------

# Suppose we have a dynamically generated list of integers,
# and want to present them as a table. The use of log10 in
# our digit_count helps calculate the proper format string.

function print_list_of_ints($ints, $line_width=40) {

 
# Apply our digit_count to the max int among ints.
 
$field_width = 2 + digit_count(max($ints));

 
# Create format string for printf.
 
$format = "%${field_width}d";

 
$ints_per_line = floor($line_width/$field_width);

 
$border = str_repeat("-", $ints_per_line * $field_width);

  echo
"\n$border\n";

  foreach(
$ints as $count => $int) {
    if(
$count and ($count % $ints_per_line == 0)) echo "\n";
   
printf($format, $int);
  }

  echo
"\n$border\n";
}

# To generate an example, here is a basic function
# returning a list of (pseudo) random numbers.

function rands($how_many) {
  for(
$i=0; $i < $how_many; $i++) $rands[] = rand();
  return
$rands;
}

# Example:  A list of random ints dynamically formatted into columns.

print_list_of_ints(rands(11));

/* Sample output. Numbers and fonts vary. Visualize monospace!

------------------------------------
  1093146637   244503173  1346204527
   638304372   140216732  1054707210
   573915416  1728677954  2038899669
   534854768    12674586
------------------------------------

*/

?>
mcmeijer at yahoo dot com
3.02.2005 16:22
$val = 1000000
$val2 = floor(log($val,10)) gives a value of 5 for $val2 and not 6 as expected.
$val2 = floor(log10($val)) gives the correct value.
c0x at mail dot ru
19.09.2004 12:08
more general version, works fine on negative, very big ($value > 1E+18) and very small ($value < 1E-18) numbers.

function expn($value, $prec = 3, $base = 1000, $prefix = '') {
    $e = array('a', 'f', 'p', 'n', 'u', 'm', '', 'k', 'M', 'G', 'T', 'P', 'E');
    $p = min(max(floor(log(abs($value), $base)), -6), 6);
    return round((float)$value / pow($base, $p), $prec) . $prefx . $e[$p + 6];
}
admin at worldtakeover dot tk
20.06.2004 22:06
In regards to the note about log in base 10 and the round() function. You need to use floor() instead of round() to find out the order of magnitude. That way, you don't have to worry about subtracting 0.5 or whatever.
mightye (at) mightye.org
6.02.2003 21:02
A minor warning:

in PHP < 4.3.0, in order to get the log base 10 of a number, you have to do:
$log10 = log($n)/log(10);

If you want a whole number (to identify the order of magnitude), and you typecast $log10 to (int), you may not get what you expect:
(int)(log(1000)/log(10)) = 2
(log(1000)/log(10)) = 3 (float with no displayed decimal places)

The mathematical error in this causes the typecast to round the result down, even though the error runs out to so many decimal places that it is not displayed, and the float value looks like a whole number.  Instead you may need to do:
round(log($n)/log(10)-0.5,0);

This will give you the order of magnitude of your number.

Presumably in PHP 4.3.0+, a similar result may occur.



PHP Powered Diese Seite bei php.net
The PHP manual text and comments are covered by the Creative Commons Attribution 3.0 License © the PHP Documentation Group - Impressum - mail("TO:Reinhard Neidl",...)