I'm trying to translate this algorithm from C language to PHP (for study)
This is an example of a perceptron. I copied the example written in c and I'm trying to translate it into PHP. Currently I wrote this code, what am I wrong? As output I only know 101 iterations with result always 1.
This is the C program:
#include <stdio.h>
#include <stdlib.h>
#include <math.h>
#include <time.h>
#define LEARNING_RATE 0.1
#define MAX_ITERATION 100
float randomFloat()
{
return (float)rand() / (float)RAND_MAX;
}
int calculateOutput(float weights[], float x, float y)
{
float sum = x * weights[0] + y * weights[1] + weights[2];
return (sum >= 0) ? 1 : -1;
}
int main(int argc, char *argv[])
{
srand(time(NULL));
float x[208], y[208], weights[3], localError, globalError;
int outputs[208], patternCount, i, p, iteration, output;
FILE *fp;
if ((fp = fopen("test1.txt", "r")) == NULL) {
printf("Cannot open file.\n");
exit(1);
}
i = 0;
while (fscanf(fp, "%f %f %d", &x[i], &y[i], &outputs[i]) != EOF) {
if (outputs[i] == 0) {
outputs[i] = -1;
}
i++;
}
patternCount = i;
weights[0] = randomFloat();
weights[1] = randomFloat();
weights[2] = randomFloat();
iteration = 0;
do {
iteration++;
globalError = 0;
for (p = 0; p < patternCount; p++) {
output = calculateOutput(weights, x[p], y[p]);
localError = outputs[p] - output;
weights[0] += LEARNING_RATE * localError * x[p];
weights[1] += LEARNING_RATE * localError * y[p];
weights[2] += LEARNING_RATE * localError;
globalError += (localError*localError);
}
/* Root Mean Squared Error */
printf("Iteration %d : RMSE = %.4f\n", iteration,
sqrt(globalError/patternCount));
} while (globalError != 0 && iteration<=MAX_ITERATION);
printf("\nDecision boundary (line) equation: %.2f*x + %.2f*y + %.2f = 0\n",
weights[0], weights[1], weights[2]);
return 0;
}
and this is the code that I wrote
<?php
define("LEARNING_RATE", 0.1);
define("MAX_ITERATION", 100);
function randomFloat(){ return (float) mt_rand() / mt_getrandmax(); }
function calculateOutput($weights, $x, $y){
$sum = (float) $x * $weights[0] + $y * $weights[1] + $weights[2];
return ($sum >= 0) ? 1 : -1;
}
srand(time());
$i = 0;
$ars = explode("\n",file_get_contents('https://raw.githubusercontent.com/RichardKnop/ansi-c-perceptron/master/test1.txt'));
foreach($ars as $ar){
$temp = explode("\t", $ar);
$x[$i] = (float) $temp[0];
$y[$i] = (float) $temp[1];
$output[$i] = (int) $temp[2];
if($output[$i] == 0)
$output[$i] = -1;
$i++;
}
$patternCount = $i;
$weights[0] = randomFloat();
$weights[1] = randomFloat();
$weights[2] = randomFloat();
$iteration = 0;
do{
$iteration++;
$globalError = 0;
for ($p = 0; $p < $patternCount; $p++) {
$output = calculateOutput($weights, $x[$p], $y[$p]);
$localError = $outputs[$p] - $output;
$weights[0] += LEARNING_RATE * $localError * $x[$p];
$weights[1] += LEARNING_RATE * $localError * $y[$p];
$weights[2] += LEARNING_RATE * $localError;
$globalError += ($localError*$localError);
}
$r .= "Iteration $iteration : RMSE = " .
sqrt($globalError/$patternCount)."<br>";
}while($globalError != 0 && $iteration<=MAX_ITERATION);
echo $r;
echo "<br><hr><br>";
echo "Decision boundary (line) equation: ".$weights[0]."*x + ".$weights[1]."*y + ".$weights[2]." = 0<br>";
it's practically identical, but why does it not work?
$ars = explode("\n",file_get_contents('…'));
Since the file ends with a \n, this yields an empty string as the last array value, which disrupts the foreach($ars as $ar) loop. To read the file into an array, simply use:
$ars = file('…');
In the foreach($ars as $ar) loop, you used the wrong name $output[$i] instead of $outputs[$i].
$r .= "Iteration $iteration : RMSE = " .
sqrt($globalError/$patternCount)."<br>";
}while($globalError != 0 && $iteration<=MAX_ITERATION);
echo $r;
You didn't initialize $r. Instead of the above, you can use:
echo "Iteration $iteration : RMSE = " .
sqrt($globalError/$patternCount)."<br>";
} while ($globalError != 0 && $iteration<=MAX_ITERATION);
My php_substr function should set kpart to a33c5b4b58b26d9f however it instead sets it to a33c5b4b58b26d9fìÿ?꬯wÿÿÿÿ®w.
Any ideas?
long ks;
char *kpart1;
ks = 16;
php_substr("a33c5b4b58b26d9f78293df1c5d5a3bf", &kpart1, 0, ks); // a33c5b4b58b26d9fìÿ?꬯wÿÿÿÿ®w
char *php_substr(char *str, char **ptr, long f, long l)
{
int str_len = strlen(str);
if (!l) {
l=0;
}
if (l!=0) {
if ((l < 0 && -l > str_len)) {
return FALSE;
} else if (l > str_len) {
l = str_len;
}
} else {
l = str_len;
}
if (f > str_len) {
return FALSE;
} else if (f < 0 && -f > str_len) {
f = 0;
}
if (l < 0 && (l + str_len - f) < 0) {
return FALSE;
}
if (f < 0) {
f = str_len + f;
if (f < 0) {
f = 0;
}
}
if (l < 0) {
l = (str_len - f) + l;
if (l < 0) {
l = 0;
}
}
if (f >= str_len) {
return FALSE;
}
if ((f + l) > str_len) {
l = str_len - f;
}
strncpy(*ptr,str+f,l);
return 0;
}
Edited to use malloc and included full working code!
I tried the zend engine emalloc however this didn't seem to work
long ks;
char *kpart1;
ks = 16;
kpart1 = php_substr("a33c5b4b58b26d9f78293df1c5d5a3bf", 0, ks); // a33c5b4b58b26d9f
char *php_substr(char *str, long f, long l)
{
int str_len = strlen(str);
unsigned char *buffer;
if (l!=0) {
if ((l < 0 && -l > str_len)) {
return FALSE;
} else if (l > str_len) {
l = str_len;
}
} else {
l = str_len;
}
if (f > str_len) {
return FALSE;
} else if (f < 0 && -f > str_len) {
f = 0;
}
if (l < 0 && (l + str_len - f) < 0) {
return FALSE;
}
if (f < 0) {
f = str_len + f;
if (f < 0) {
f = 0;
}
}
if (l < 0) {
l = (str_len - f) + l;
if (l < 0) {
l = 0;
}
}
if (f >= str_len) {
return FALSE;
}
if ((f + l) > str_len) {
l = str_len - f;
}
buffer = (char*)malloc(l+1);
strncpy(buffer,str+f,l);
buffer[l]='\0';
return buffer;
}
After trimming out all the testing of the parameters
After editing into a file that compiles/links/runs
this is the resulting code:
#include <stdio.h>
#include <string.h>
#define FALSE (0)
char *php_substr(char *str, char **ptr, long f, long l)
{
strncpy(*ptr,str+f,(size_t)l);
return 0;
}
int main( void )
{
long int ks = 16;
char *kpart1;
php_substr("a33c5b4b58b26d9f78293df1c5d5a3bf", &kpart1, 0, ks); // a33c5b4b58b26d9fìÿ?꬯wÿÿÿÿ®w
printf( "%s\n", kpart1);
}
Notice that the char pointer: kpart1 is not initialized to point to any specific location.
(in this case it contains whatever trash was on the stack. If it were a global variable, it would contian NULL.)
The call to strncpy() is trying to copy the first 16 bytes of the passed in constant to where ever kpart1 happens to be pointing.
The strncpy() results in undefined behaviour and can lead to a seg fault event.
I did not perform an analysis on the manipulation of the passed in variables f and l.
Suggest, as you found, that the pointer: kpart1 needs to be initialized to point to some allocated memory, where that allocated memory needs to be at least l-f+1 in length.
Is there an easy way in PHP to figure out the source code/content of a built in function?
Say for example I want to know what base64_decode() actually does to the given encoded base64 string to convert it to a plain text. How can I achieve this?
You can browse the source code of PHP here
In your case base64_decode is implemented here (PHP 5.6.0)
Note: This code is in C since that's what PHP is written in. In fact all built-in functions and extensions of PHP are written in C.
PHPAPI unsigned char *php_base64_decode_ex(const unsigned char *str, int length, int *ret_length, zend_bool strict) /* {{{ */
{
const unsigned char *current = str;
int ch, i = 0, j = 0, k;
/* this sucks for threaded environments */
unsigned char *result;
result = (unsigned char *)safe_emalloc(length, 1, 1);
/* run through the whole string, converting as we go */
while ((ch = *current++) != '\0' && length-- > 0) {
if (ch == base64_pad) {
if (*current != '=' && ((i % 4) == 1 || (strict && length > 0))) {
if ((i % 4) != 1) {
while (isspace(*(++current))) {
continue;
}
if (*current == '\0') {
continue;
}
}
efree(result);
return NULL;
}
continue;
}
ch = base64_reverse_table[ch];
if ((!strict && ch < 0) || ch == -1) { /* a space or some other separator character, we simply skip over */
continue;
} else if (ch == -2) {
efree(result);
return NULL;
}
switch(i % 4) {
case 0:
result[j] = ch << 2;
break;
case 1:
result[j++] |= ch >> 4;
result[j] = (ch & 0x0f) << 4;
break;
case 2:
result[j++] |= ch >>2;
result[j] = (ch & 0x03) << 6;
break;
case 3:
result[j++] |= ch;
break;
}
i++;
}
k = j;
/* mop things up if we ended on a boundary */
if (ch == base64_pad) {
switch(i % 4) {
case 1:
efree(result);
return NULL;
case 2:
k++;
case 3:
result[k] = 0;
}
}
if(ret_length) {
*ret_length = j;
}
result[j] = '\0';
return result;
}
I'm trying to reimplement python slice notation in another language (php) and looking for a snippet (in any language or pseudocode) that would mimic the python logic. That is, given a list and a triple (start, stop, step) or a part thereof, determine correct values or defaults for all parameters and return a slice as a new list.
I tried looking into the source. That code is far beyond my c skills, but I can't help but agree with the comment saying:
/* this is harder to get right than you might think */
Also, if something like this is already done, pointers will be greatly appreciated.
This is my test bench (make sure your code passes before posting):
#place your code below
code = """
def mySlice(L, start=None, stop=None, step=None):
or
<?php function mySlice($L, $start=NULL, $stop=NULL, $step=NULL) ...
or
function mySlice(L, start, stop, step) ...
"""
import itertools
L = [0,1,2,3,4,5,6,7,8,9]
if code.strip().startswith('<?php'):
mode = 'php'
if code.strip().startswith('def'):
mode = 'python'
if code.strip().startswith('function'):
mode = 'js'
if mode == 'php':
var, none = '$L', 'NULL'
print code, '\n'
print '$L=array(%s);' % ','.join(str(x) for x in L)
print "function _c($s,$a,$e){if($a!==$e)echo $s,' should be [',implode(',',$e),'] got [',implode(',',$a),']',PHP_EOL;}"
if mode == 'python':
var, none = 'L', 'None'
print code, '\n'
print 'L=%r' % L
print "def _c(s,a,e):\n\tif a!=e:\n\t\tprint s,'should be',e,'got',a"
if mode == 'js':
var, none = 'L', 'undefined'
print code, '\n'
print 'L=%r' % L
print "function _c(s,a,e){if(a.join()!==e.join())console.log(s+' should be ['+e.join()+'] got ['+a.join()+']');}"
print
n = len(L) + 3
start = range(-n, n) + [None, 100, -100]
stop = range(-n, n) + [None, 100, -100]
step = range(-n, n) + [100, -100]
for q in itertools.product(start, stop, step):
if not q[2]: q = q[:-1]
actual = 'mySlice(%s,%s)' % (var, ','.join(none if x is None else str(x) for x in q))
slice_ = 'L[%s]' % ':'.join('' if x is None else str(x) for x in q)
expect = eval(slice_)
if mode == 'php':
expect = 'array(%s)' % ','.join(str(x) for x in expect)
print "_c(%r,%s,%s);" % (slice_, actual, expect)
if mode == 'python':
print "_c(%r,%s,%s);" % (slice_, actual, expect)
if mode == 'js':
print "_c(%r,%s,%s);" % (slice_, actual, expect)
how to use it:
save into a file (test.py)
place your python, php or javascript code between """s
run python test.py | python or python test.py | php or python test.py | node
Here's a straight port of the C code:
def adjust_endpoint(length, endpoint, step):
if endpoint < 0:
endpoint += length
if endpoint < 0:
endpoint = -1 if step < 0 else 0
elif endpoint >= length:
endpoint = length - 1 if step < 0 else length
return endpoint
def adjust_slice(length, start, stop, step):
if step is None:
step = 1
elif step == 0:
raise ValueError("step cannot be 0")
if start is None:
start = length - 1 if step < 0 else 0
else:
start = adjust_endpoint(length, start, step)
if stop is None:
stop = -1 if step < 0 else length
else:
stop = adjust_endpoint(length, stop, step)
return start, stop, step
def slice_indices(length, start, stop, step):
start, stop, step = adjust_slice(length, start, stop, step)
i = start
while (i > stop) if step < 0 else (i < stop):
yield i
i += step
def mySlice(L, start=None, stop=None, step=None):
return [L[i] for i in slice_indices(len(L), start, stop, step)]
This is what I came up with (python)
def mySlice(L, start=None, stop=None, step=None):
answer = []
if not start:
start = 0
if start < 0:
start += len(L)
if not stop:
stop = len(L)
if stop < 0:
stop += len(L)
if not step:
step = 1
if stop == start or (stop<=start and step>0) or (stop>=start and step<0):
return []
i = start
while i != stop:
try:
answer.append(L[i])
i += step
except:
break
return answer
Seems to work - let me know what you think
Hope it helps
This is a solution I came up with in C# .NET, maybe not the prettiest, but it works.
private object[] Slice(object[] list, int start = 0, int stop = 0, int step = 0)
{
List<object> result = new List<object>();
if (step == 0) step = 1;
if (start < 0)
{
for (int i = list.Length + start; i < list.Length - (list.Length + start); i++)
{
result.Add(list[i]);
}
}
if (start >= 0 && stop == 0) stop = list.Length - (start >= 0 ? start : 0);
else if (start >= 0 && stop < 0) stop = list.Length + stop;
int loopStart = (start < 0 ? 0 : start);
int loopEnd = (start > 0 ? start + stop : stop);
if (step > 0)
{
for (int i = loopStart; i < loopEnd; i += step)
result.Add(list[i]);
}
else if (step < 0)
{
for (int i = loopEnd - 1; i >= loopStart; i += step)
result.Add(list[i]);
}
return result.ToArray();
}
I've written a PHP port based on the C code, optimized for step sizes -1 and 1:
function get_indices($length, $step, &$start, &$end, &$size)
{
if (is_null($start)) {
$start = $step < 0 ? $length - 1 : 0;
} else {
if ($start < 0) {
$start += $length;
if ($start < 0) {
$start = $step < 0 ? -1 : 0;
}
} elseif ($start >= $length) {
$start = $step < 0 ? $length - 1 : $length;
}
}
if (is_null($end)) {
$end = $step < 0 ? -1 : $length;
} else {
if ($end < 0) {
$end += $length;
if ($end < 0) {
$end = $step < 0 ? - 1 : 0;
}
} elseif ($end >= $length) {
$end = $step < 0 ? $length - 1 : $length;
}
}
if (($step < 0 && $end >= $start) || ($step > 0 && $start >= $end)) {
$size = 0;
} elseif ($step < 0) {
$size = ($end - $start + 1) / $step + 1;
} else {
$size = ($end - $start - 1) / $step + 1;
}
}
function mySlice($L, $start = NULL, $end = NULL, $step = 1)
{
if (!$step) {
return false; // could throw exception too
}
$length = count($L);
get_indices($length, $step, $start, $end, $size);
// optimize default step
if ($step == 1) {
// apply native array_slice()
return array_slice($L, $start, $size);
} elseif ($step == -1) {
// negative step needs an array reversal first
// with range translation
return array_slice(array_reverse($L), $length - $start - 1, $size);
} else {
// standard fallback
$r = array();
for ($i = $start; $step < 0 ? $i > $end : $i < $end; $i += $step) {
$r[] = $L[$i];
}
return $r;
}
}
This is based on #ecatmur's Python code ported again to PHP.
<?php
function adjust_endpoint($length, $endpoint, $step) {
if ($endpoint < 0) {
$endpoint += $length;
if ($endpoint < 0) {
$endpoint = $step < 0 ? -1 : 0;
}
}
elseif ($endpoint >= $length) {
$endpoint = $step < 0 ? $length - 1 : $length;
}
return $endpoint;
}
function mySlice($L, $start = null, $stop = null, $step = null) {
$sliced = array();
$length = count($L);
// adjust_slice()
if ($step === null) {
$step = 1;
}
elseif ($step == 0) {
throw new Exception('step cannot be 0');
}
if ($start === null) {
$start = $step < 0 ? $length - 1 : 0;
}
else {
$start = adjust_endpoint($length, $start, $step);
}
if ($stop === null) {
$stop = $step < 0 ? -1 : $length;
}
else {
$stop = adjust_endpoint($length, $stop, $step);
}
// slice_indices()
$i = $start;
$result = array();
while ($step < 0 ? ($i > $stop) : ($i < $stop)) {
$sliced []= $L[$i];
$i += $step;
}
return $sliced;
}
I can't say there's no bug in the codes, but it had past your test program :)
def mySlice(L, start=None, stop=None, step=None):
ret = []
le = len(L)
if step is None: step = 1
if step > 0: #this situation might be easier
if start is None:
start = 0
else:
if start < 0: start += le
if start < 0: start = 0
if start > le: start = le
if stop is None:
stop = le
else:
if stop < 0: stop += le
if stop < 0: stop = 0
if stop > le: stop = le
else:
if start is None:
start = le-1
else:
if start < 0: start += le
if start < 0: start = -1
if start >= le: start = le-1
if stop is None:
stop = -1 #stop is not 0 because we need L[0]
else:
if stop < 0: stop += le
if stop < 0: stop = -1
if stop >= le: stop = le
#(stop-start)*step>0 to make sure 2 things:
#1: step != 0
#2: iteration will end
while start != stop and (stop-start)*step > 0 and start >=0 and start < le:
ret.append( L[start] )
start += step
return ret
PHP has a feature where you can use increment operators on strings. It behaves similarly to an odometer, where once you reach the end of a range, it "rolls over".
<?php
$str = 'zy';
$str++;
echo "$str\n"; // zz
$str++;
echo "$str\n"; // aaa
Just curious where in the PHP source code this is. I often look in the source code at functions/extensions, but something like this I have no idea where to look.
A link to the file using their web based SVN would be awesome.
The implementation for this operator is conveniently located in zend_operators.c, in a function that's even more conveniently called increment_string():
static void increment_string(zval *str) /* {{{ */
{
int carry=0;
int pos=Z_STRLEN_P(str)-1;
char *s=Z_STRVAL_P(str);
char *t;
int last=0; /* Shut up the compiler warning */
int ch;
if (Z_STRLEN_P(str) == 0) {
STR_FREE(Z_STRVAL_P(str));
Z_STRVAL_P(str) = estrndup("1", sizeof("1")-1);
Z_STRLEN_P(str) = 1;
return;
}
if (IS_INTERNED(s)) {
s = (char*) emalloc(Z_STRLEN_P(str) + 1);
memcpy(s, Z_STRVAL_P(str), Z_STRLEN_P(str) + 1);
Z_STRVAL_P(str) = s;
}
while (pos >= 0) {
ch = s[pos];
if (ch >= 'a' && ch <= 'z') {
if (ch == 'z') {
s[pos] = 'a';
carry=1;
} else {
s[pos]++;
carry=0;
}
last=LOWER_CASE;
} else if (ch >= 'A' && ch <= 'Z') {
if (ch == 'Z') {
s[pos] = 'A';
carry=1;
} else {
s[pos]++;
carry=0;
}
last=UPPER_CASE;
} else if (ch >= '0' && ch <= '9') {
if (ch == '9') {
s[pos] = '0';
carry=1;
} else {
s[pos]++;
carry=0;
}
last = NUMERIC;
} else {
carry=0;
break;
}
if (carry == 0) {
break;
}
pos--;
}
if (carry) {
t = (char *) emalloc(Z_STRLEN_P(str)+1+1);
memcpy(t+1, Z_STRVAL_P(str), Z_STRLEN_P(str));
Z_STRLEN_P(str)++;
t[Z_STRLEN_P(str)] = '\0';
switch (last) {
case NUMERIC:
t[0] = '1';
break;
case UPPER_CASE:
t[0] = 'A';
break;
case LOWER_CASE:
t[0] = 'a';
break;
}
STR_FREE(Z_STRVAL_P(str));
Z_STRVAL_P(str) = t;
}
}
/* }}} */