intaleq/backend/bot/ai_formula_solver.php

<?php
/**
 * ai_formula_solver.php
 * مكتشف خوارزميات المنافسين (AI Competitor Formula Solver)
 * يستخدم الانحدار الخطي المتعدد (Multiple Linear Regression) لاكتشاف
 * أجرة فتح العداد، وسعر الكيلومتر، وسعر الدقيقة لكل تطبيق منافس.
 */

require_once __DIR__ . '/../core/bootstrap.php';
require_once __DIR__ . '/../functions.php';

try {
    $con = Database::get('main');
} catch (Exception $e) {
    die("Database connection failed: " . $e->getMessage() . "\n");
}

echo "Starting AI Formula Discovery Engine...\n";

// نجلب التطبيقات التي لديها بيانات (مسافة ووقت وسعر)
$sqlApps = "SELECT DISTINCT competitor_name, country_code 
            FROM scraped_competitor_prices 
            WHERE distance_km > 0 AND duration_min > 0 AND price_amount > 0";
$stmtApps = $con->query($sqlApps);
$apps = $stmtApps->fetchAll(PDO::FETCH_ASSOC);

if (empty($apps)) {
    echo "No sufficient data (Distance/Duration) found to perform regression.\n";
    exit;
}

// دالة لحل نظام معادلات خطية (Gaussian Elimination)
function solveLinearSystem($A, $B) {
    $n = count($A);
    for ($i = 0; $i < $n; $i++) {
        // Search for maximum in this column
        $maxEl = abs($A[$i][$i]);
        $maxRow = $i;
        for ($k = $i + 1; $k < $n; $k++) {
            if (abs($A[$k][$i]) > $maxEl) {
                $maxEl = abs($A[$k][$i]);
                $maxRow = $k;
            }
        }
        
        // Swap maximum row with current row
        for ($k = $i; $k < $n; $k++) {
            $tmp = $A[$maxRow][$k];
            $A[$maxRow][$k] = $A[$i][$k];
            $A[$i][$k] = $tmp;
        }
        $tmp = $B[$maxRow];
        $B[$maxRow] = $B[$i];
        $B[$i] = $tmp;
        
        // Make all rows below this one 0 in current column
        for ($k = $i + 1; $k < $n; $k++) {
            if ($A[$i][$i] == 0) continue;
            $c = -$A[$k][$i] / $A[$i][$i];
            for ($j = $i; $j < $n; $j++) {
                if ($i == $j) {
                    $A[$k][$j] = 0;
                } else {
                    $A[$k][$j] += $c * $A[$i][$j];
                }
            }
            $B[$k] += $c * $B[$i];
        }
    }
    
    // Solve equation Ax=b for an upper triangular matrix A
    $x = array_fill(0, $n, 0);
    for ($i = $n - 1; $i >= 0; $i--) {
        if ($A[$i][$i] == 0) continue;
        $x[$i] = $B[$i] / $A[$i][$i];
        for ($k = $i - 1; $k >= 0; $k--) {
            $B[$k] -= $A[$k][$i] * $x[$i];
        }
    }
    return $x;
}

foreach ($apps as $app) {
    $competitor = $app['competitor_name'];
    $countryCode = $app['country_code'];
    
    echo "Analyzing: $competitor ($countryCode)...\n";
    
    // سحب أحدث 5000 رحلة لتكوين نموذج رياضي دقيق
    $sqlData = "SELECT distance_km, duration_min, price_amount 
                FROM scraped_competitor_prices 
                WHERE competitor_name = :comp 
                AND country_code = :country 
                AND distance_km > 0 AND duration_min > 0 AND price_amount > 0
                ORDER BY id DESC LIMIT 5000";
    $stmtData = $con->prepare($sqlData);
    $stmtData->execute([':comp' => $competitor, ':country' => $countryCode]);
    $samples = $stmtData->fetchAll(PDO::FETCH_ASSOC);
    
    $N = count($samples);
    if ($N < 10) {
        echo " -> Not enough samples ($N). Skipping.\n";
        continue;
    }
    
    // بناء مصفوفات Least Squares (X^T X) * Beta = (X^T Y)
    // Beta = [Base_Fare, KM_Price, Min_Price]
    $sum_x1 = 0; $sum_x2 = 0; $sum_y = 0;
    $sum_x1_sq = 0; $sum_x2_sq = 0; $sum_x1_x2 = 0;
    $sum_x1_y = 0; $sum_x2_y = 0;
    
    foreach ($samples as $s) {
        $x1 = (float)$s['distance_km'];
        $x2 = (float)$s['duration_min'];
        $y = (float)$s['price_amount'];
        
        $sum_x1 += $x1;
        $sum_x2 += $x2;
        $sum_y += $y;
        
        $sum_x1_sq += ($x1 * $x1);
        $sum_x2_sq += ($x2 * $x2);
        $sum_x1_x2 += ($x1 * $x2);
        
        $sum_x1_y += ($x1 * $y);
        $sum_x2_y += ($x2 * $y);
    }
    
    $matrixA = [
        [$N,       $sum_x1,    $sum_x2],
        [$sum_x1,  $sum_x1_sq, $sum_x1_x2],
        [$sum_x2,  $sum_x1_x2, $sum_x2_sq]
    ];
    
    $matrixB = [
        $sum_y,
        $sum_x1_y,
        $sum_x2_y
    ];
    
    // حل المصفوفة
    try {
        $beta = solveLinearSystem($matrixA, $matrixB);
        
        $baseFare = round(max(0, $beta[0]), 3); // Base fare cannot be negative
        $kmPrice  = round(max(0, $beta[1]), 3);
        $minPrice = round(max(0, $beta[2]), 3);
        
        echo " -> [DISCOVERED] Base Fare: $baseFare, KM: $kmPrice, Min: $minPrice\n";
        
        // حفظ في جدول المعادلات السرية
        $sqlUpsert = "INSERT INTO competitor_secret_formulas 
                      (competitor_name, country_code, base_fare, price_per_km, price_per_min, sample_size) 
                      VALUES (:comp, :country, :base, :km, :min, :size)
                      ON DUPLICATE KEY UPDATE 
                      base_fare = :base, price_per_km = :km, price_per_min = :min, sample_size = :size, last_updated = NOW()";
        
        $stmtUp = $con->prepare($sqlUpsert);
        $stmtUp->execute([
            ':comp' => $competitor,
            ':country' => $countryCode,
            ':base' => $baseFare,
            ':km' => $kmPrice,
            ':min' => $minPrice,
            ':size' => $N
        ]);
        
        echo " -> Saved successfully.\n";
        
    } catch (Exception $e) {
        echo " -> Error solving matrix: " . $e->getMessage() . "\n";
    }
}

echo "Done.\n";
?>