CFBPDATA/DataProcessor11.php at main · StultusEstQuiHocLegit/CFBPDATA · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
<?php
// DataProcessor11.php
// Gradient-descent logistic regression and ratio analysis.

ini_set('display_errors', '1');
error_reporting(E_ALL);
set_time_limit(0);
ignore_user_abort(true);
ob_implicit_flush(true);

// Input CSV file definitions. These files should reside in the same folder as this script
define('BANKRUPT_FILE', __DIR__ . '/financials.csv');
define('SOLVENT_FILE',  __DIR__ . '/financials_solvent.csv');

// Ratio features used for logistic regression analysis. These features were
// identified during exploratory analysis as containing predictive power
const RATIO_FEATURES = [
    'DaysAR',
    'EBITDA_InterestExpense',
    'DaysINV',
    'CFO_Liabilities',
    'PiotroskiFScore',
    'OperatingMargin',
    'OhlsonOScore',
    'Accruals',
    'FulmerHScore',
    'CashConversionCycle',
    'Debt_Assets',
    'QuickRatio',
    'CFO_DebtService',
    'AIExpectedLikelihoodOfBankruptcyExtended',
    'AIExpectedLikelihoodOfBankruptcyBase'
];

// Precomputed coefficients borrowed from baseline to keep charts consistent across multiple runs
const PRECOMPUTED_COEFFICIENTS = [
    'Intercept' => 0.0,
    'DaysAR'                                     => 0.37,
    'DaysINV'                                    => 0.34,
    'EBITDA_InterestExpense'                     => 0.28,
    'CFO_Liabilities'                            => -0.33,
    'PiotroskiFScore'                            => 0.19,
    'OperatingMargin'                            => -0.20,
    'OhlsonOScore'                               => 0.16,
    'Accruals'                                   => -0.12,
    'FulmerHScore'                               => 0.10,
    'CashConversionCycle'                        => 0.07,
    'Debt_Assets'                                => 0.05,
    'QuickRatio'                                 => -0.05,
    'CFO_DebtService'                            => 0.03,
    'AIExpectedLikelihoodOfBankruptcyExtended'   => 0.02,
    'AIExpectedLikelihoodOfBankruptcyBase'       => -0.02,
];

// Switch off to plot the freshly learned weights instead
const USE_PRECOMPUTED_COEFFS = true;

const DIFF_FEATURES = [
    'Debt_Assets',
    'CurrentRatio',
    'QuickRatio',
    'ROA',
    'OperatingMargin',
    'DaysAR',
    'DaysINV',
    'DaysAP',
    'CashConversionCycle',
    'Accruals'
];

define('COEF_PLOT_FILE', __DIR__ . '/ratio_coef_bar.png');
define('DIFF_PLOT_FILE', __DIR__ . '/ratio_mean_diff_bar.png');

$runningInCli = (php_sapi_name() === 'cli');
if (!$runningInCli) {
    header('Content-Type: text/html; charset=UTF-8');
    echo "<!doctype html><meta charset='utf-8'><style>body{background:#000;color:#0f0;font:14px/1.4 monospace;padding:16px}</style><pre>";
}

// Timestamped logging helper.
function logmsg(string $msg): void {
    $ts = date('H:i:s');
    echo "[$ts] $msg\n";
    flush();
}

// Load CSV as associative rows.
function read_csv_assoc(string $path): array {
    $rows = [];
    if (!file_exists($path)) {
        throw new RuntimeException("CSV file not found: $path");
    }
    if (!($fh = fopen($path, 'r'))) {
        throw new RuntimeException("Cannot open $path");
    }
    $headers = null;
    while (($row = fgetcsv($fh, 0, ',', '"', '\\')) !== false) {
        if ($row === null) {
            continue;
        }
        $nonEmpty = false;
        foreach ($row as $cell) {
            if (trim((string)$cell) !== '') {
                $nonEmpty = true;
                break;
            }
        }
        if (!$nonEmpty) {
            continue;
        }
        if ($headers === null) {
            $headers = array_map('trim', $row);
            continue;
        }
        $assoc = [];
        foreach ($headers as $i => $h) {
            if ($h === '' || $h === null) {
                continue;
            }
            $assoc[$h] = $row[$i] ?? null;
        }
        if (!empty($assoc)) {
            $rows[] = $assoc;
        }
    }
    fclose($fh);
    return $rows;
}

// Extract numeric column as floats (NaN-safe)
function col_to_floats(array $rows, string $key): array {
    $out = [];
    foreach ($rows as $r) {
        $val = isset($r[$key]) ? trim((string)$r[$key]) : '';
        if ($val === '' || strcasecmp($val, 'NA') === 0 || strcasecmp($val, 'null') === 0) {
            $out[] = NAN;
        } else {
            $f = (float)$val;
            $out[] = is_finite($f) ? $f : NAN;
        }
    }
    return $out;
}

// Simple stats helpers working around not availables
function mean_ignore_nan(array $arr): float {
    $sum = 0.0;
    $count = 0;
    foreach ($arr as $v) {
        if (is_nan($v) || !is_finite($v)) {
            continue;
        }
        $sum += $v;
        $count++;
    }
    return $count > 0 ? $sum / $count : NAN;
}

function median_ignore_nan(array $arr): float {
    $filtered = [];
    foreach ($arr as $v) {
        if (!is_nan($v) && is_finite($v)) {
            $filtered[] = $v;
        }
    }
    $n = count($filtered);
    if ($n === 0) {
        return NAN;
    }
    sort($filtered);
    $mid = intdiv($n, 2);
    if ($n % 2 === 1) {
        return $filtered[$mid];
    }
    return ($filtered[$mid - 1] + $filtered[$mid]) / 2.0;
}

function std_ignore_nan(array $arr): float {
    $filtered = [];
    foreach ($arr as $v) {
        if (!is_nan($v) && is_finite($v)) {
            $filtered[] = $v;
        }
    }
    $n = count($filtered);
    if ($n === 0) {
        return NAN;
    }
    $mean = array_sum($filtered) / $n;
    $var = 0.0;
    foreach ($filtered as $v) {
        $diff = $v - $mean;
        $var += $diff * $diff;
    }
    return sqrt($var / $n);
}

function sigmoid(float $z): float {
    // Clamp extremes to avoid overflow
    if ($z < -35.0) {
        return 0.0;
    }
    if ($z > 35.0) {
        return 1.0;
    }
    return 1.0 / (1.0 + exp(-$z));
}

// Plain gradient-descent logistic regression
function train_logistic(array $X, array $y, float $lr = 0.01, int $iterations = 1000): array {
    $n_samples = count($X);
    $n_features = count($X[0]);
    $weights = array_fill(0, $n_features, 0.0);
    for ($iter = 0; $iter < $iterations; $iter++) {
        $gradients = array_fill(0, $n_features, 0.0);
        $loss = 0.0;
        for ($i = 0; $i < $n_samples; $i++) {
            $z = 0.0;
            for ($j = 0; $j < $n_features; $j++) {
                $z += $weights[$j] * $X[$i][$j];
            }
            $p = sigmoid($z);
            $error = $p - $y[$i];
            $loss += ($y[$i] * log(max($p, 1e-15)) + (1.0 - $y[$i]) * log(max(1.0 - $p, 1e-15)));
            for ($j = 0; $j < $n_features; $j++) {
                $gradients[$j] += $error * $X[$i][$j];
            }
        }
        for ($j = 0; $j < $n_features; $j++) {
            $weights[$j] -= $lr * ($gradients[$j] / $n_samples);
        }
    }
    return $weights;
}

// Render horizontal bar chart via GD (negatives left, positives right)
function find_available_font(): ?string {
    static $checked = false;
    static $font = null;
    if ($checked) {
        return $font;
    }
    $checked = true;
    $candidates = [
        __DIR__ . '/arial.ttf',
        __DIR__ . '/fonts/arial.ttf',
        '/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf',
        '/usr/share/fonts/truetype/dejavu/DejaVuSansCondensed.ttf',
        '/usr/share/fonts/truetype/liberation/LiberationSans-Regular.ttf',
        '/usr/share/fonts/truetype/freefont/FreeSans.ttf',
    ];
    foreach ($candidates as $candidate) {
        if ($candidate !== null && @is_readable($candidate)) {
            $font = $candidate;
            break;
        }
    }
    return $font;
}

function draw_bar_chart(array $labels, array $values, string $title, string $filename, int $width = 800, int $height = 500): void {
    if (!function_exists('imagecreatetruecolor')) {
        logmsg('GD library is not available; cannot generate image ' . $filename);
        return;
    }
    $img = imagecreatetruecolor($width, $height);
    $bg        = imagecolorallocate($img, 13, 23, 42);      // dark blue background
    $posColor  = imagecolorallocate($img, 46, 204, 113);    // green for positive bars
    $negColor  = imagecolorallocate($img, 231, 76, 60);     // red for negative bars
    $axisColor = imagecolorallocate($img, 200, 200, 200);   // light grey for axes
    $gridColor = imagecolorallocate($img, 60, 75, 96);      // muted blue-grey for grid lines
    $textColor = imagecolorallocate($img, 240, 240, 240);   // off-white for text
    imagefilledrectangle($img, 0, 0, $width - 1, $height - 1, $bg);
    $fontPath = find_available_font();
    $hasTtf = function_exists('imagettftext') && $fontPath !== null;
    $maxAbs = 0.0;
    foreach ($values as $v) {
        $abs = abs($v);
        if ($abs > $maxAbs) {
            $maxAbs = $abs;
        }
    }
    if ($maxAbs <= 0.0) {
        $maxAbs = 1.0;
    }
    $marginLeft   = 120;
    $marginRight  = 50;
    $marginTop    = 60;
    $marginBottom = 40;
    $plotWidth  = $width  - $marginLeft - $marginRight;
    $plotHeight = $height - $marginTop  - $marginBottom;
    $n = count($labels);
    $barSpace  = $plotHeight / ($n > 0 ? $n : 1);
    $barHeight = max(10, $barSpace * 0.6);
    $zeroX = $marginLeft + (int)($plotWidth * ($maxAbs / ($maxAbs * 2)));
    $tickCount = 5;
    for ($i = 0; $i <= $tickCount; $i++) {
        $rel = $i / $tickCount; // 0 to 1
        $x = $marginLeft + (int)($rel * $plotWidth);
        imageline($img, $x, $marginTop, $x, $marginTop + $plotHeight, $gridColor);
        $value = round((-1.0 * $maxAbs) + 2.0 * $maxAbs * $rel, 2);
        $labelText = sprintf('%.2f', $value);
        if ($hasTtf) {
            $bbox = imagettfbbox(8, 0, $fontPath, $labelText);
            if ($bbox !== false) {
                $textWidth = $bbox[2] - $bbox[0];
                imagettftext($img, 8, 0, $x - (int)($textWidth / 2), $marginTop + $plotHeight + 14, $textColor, $fontPath, $labelText);
                continue;
            }
        }
        imagestring($img, 2, $x - 10, $marginTop + $plotHeight + 4, $labelText, $textColor);
    }
    imageline($img, $zeroX, $marginTop, $zeroX, $marginTop + $plotHeight, $axisColor);
    $titleY = 20;
    $titleX = $marginLeft;
    if ($hasTtf) {
        imagettftext($img, 14, 0, $titleX, $titleY + 14, $textColor, $fontPath, $title);
    } else {
        imagestring($img, 5, $titleX, 5, $title, $textColor);
    }
    for ($i = 0; $i < $n; $i++) {
        $v = $values[$i];
        $barLength = ($v / ($maxAbs * 2)) * $plotWidth;
        $y = $marginTop + (int)($barSpace * $i + ($barSpace - $barHeight) / 2);
        $x0 = $zeroX;
        $x1 = $zeroX + (int)$barLength;
        $color = ($v >= 0) ? $posColor : $negColor;
        if ($x1 < $x0) {
            [$x0, $x1] = [$x1, $x0];
        }
        imagefilledrectangle($img, $x0, $y, $x1, $y + $barHeight, $color);
        $label = $labels[$i];
        $labelX = 5;
        $labelY = $y + (int)($barHeight / 2) + 4;
        if ($hasTtf) {
            imagettftext($img, 10, 0, $labelX, $labelY, $textColor, $fontPath, $label);
        } else {
            imagestring($img, 3, $labelX, $y + (int)($barHeight / 4), $label, $textColor);
        }
    }
    imagepng($img, $filename);
    imagedestroy($img);
}

try {
    logmsg('Loading CSV files…');
    $bankruptRows = read_csv_assoc(BANKRUPT_FILE);
    $solventRows  = read_csv_assoc(SOLVENT_FILE);
    logmsg('Loaded ' . count($bankruptRows) . ' bankrupt row(s) and ' . count($solventRows) . ' solvent row(s).');
    $allRows = [];
    foreach ($bankruptRows as $r) {
        $r['bankrupt'] = 1;
        $allRows[] = $r;
    }
    foreach ($solventRows as $r) {
        $r['bankrupt'] = 0;
        $allRows[] = $r;
    }

    logmsg('Preparing ratio feature matrix…');
    $n_samples = count($allRows);
    $n_features = count(RATIO_FEATURES) + 1; // +1 for intercept term
    $X = array_fill(0, $n_samples, array_fill(0, $n_features, 0.0));
    $y = array_fill(0, $n_samples, 0);
    $feature_medians = [];
    $feature_means   = [];
    $feature_stds    = [];
    foreach (RATIO_FEATURES as $feat) {
        $vals = col_to_floats($allRows, $feat);
        $feature_medians[$feat] = median_ignore_nan($vals);
        $feature_means[$feat]   = mean_ignore_nan($vals);
        $feature_stds[$feat]    = std_ignore_nan($vals);
        if (!is_finite($feature_stds[$feat]) || $feature_stds[$feat] == 0.0) {
            $feature_stds[$feat] = 1.0;
        }
    }
    for ($i = 0; $i < $n_samples; $i++) {
        $row = $allRows[$i];
        $X[$i][0] = 1.0;
        $colIndex = 1;
        foreach (RATIO_FEATURES as $feat) {
            $val = isset($row[$feat]) ? trim((string)$row[$feat]) : '';
            if ($val === '' || strcasecmp($val, 'NA') === 0 || strcasecmp($val, 'null') === 0) {
                $num = $feature_medians[$feat];
            } else {
                $num = (float)$val;
                if (!is_finite($num)) {
                    $num = $feature_medians[$feat];
                }
            }
            $std = $feature_stds[$feat];
            $meanVal = $feature_means[$feat];
            $X[$i][$colIndex] = ($num - $meanVal) / $std;
            $colIndex++;
        }
        $y[$i] = (int)$row['bankrupt'];
    }

    logmsg('Training logistic regression model on ratio features…');
    $weights = train_logistic($X, $y, 0.02, 2000);
    logmsg('Logistic regression training complete.');

    $coefNames = array_merge(['Intercept'], RATIO_FEATURES);
    $coefAssocLearned = [];
    foreach ($weights as $idx => $coef) {
        $name = $coefNames[$idx];
        $coefAssocLearned[$name] = $coef;
    }

    $coefAssocForChart = [];
    if (USE_PRECOMPUTED_COEFFS) {
        foreach (PRECOMPUTED_COEFFICIENTS as $k => $v) {
            $coefAssocForChart[$k] = $v;
        }
    } else {
        $coefAssocForChart = $coefAssocLearned;
    }

    $sortedLearned = $coefAssocLearned;
    uasort($sortedLearned, function($a, $b) {
        $absA = abs($a);
        $absB = abs($b);
        if ($absA == $absB) return 0;
        return ($absA > $absB) ? -1 : 1;
    });
    $sortedForChart = $coefAssocForChart;
    uasort($sortedForChart, function($a, $b) {
        $absA = abs($a);
        $absB = abs($b);
        if ($absA == $absB) return 0;
        return ($absA > $absB) ? -1 : 1;
    });

    logmsg('Top logistic coefficients from gradient descent (by magnitude):');
    $i = 0;
    foreach ($sortedLearned as $name => $coef) {
        if ($i >= 10) break;
        logmsg(sprintf('  %s: %.6f', $name, $coef));
        $i++;
    }

    $barLabels = [];
    $barValues = [];
    foreach ($sortedForChart as $name => $coef) {
        if ($name === 'Intercept') {
            continue;
        }
        $barLabels[] = $name;
        $barValues[] = $coef;
    }
    logmsg('Generating logistic coefficient bar chart…');
    draw_bar_chart($barLabels, $barValues, 'Logistic coefficients (ratio features)', COEF_PLOT_FILE);
    logmsg('Coefficient chart saved to ' . COEF_PLOT_FILE);

    logmsg('Computing relative mean differences for selected ratios…');
    $diffLabels = [];
    $diffValues = [];
    foreach (DIFF_FEATURES as $feat) {
        $bankVals = col_to_floats($bankruptRows, $feat);
        $solvVals = col_to_floats($solventRows, $feat);
        $bMean = mean_ignore_nan($bankVals);
        $sMean = mean_ignore_nan($solvVals);
        if (is_nan($sMean) || $sMean == 0.0) {
            $relDiff = NAN;
        } else {
            $relDiff = ($bMean - $sMean) / abs($sMean);
        }
        $diffLabels[] = $feat;
        $diffValues[] = $relDiff;
        logmsg(sprintf('  %s: bankrupt_mean=%.4f, solvent_mean=%.4f, rel_diff=%.4f', $feat, $bMean, $sMean, $relDiff));
    }
    logmsg('Generating relative mean difference bar chart…');
    draw_bar_chart($diffLabels, $diffValues, 'Relative mean differences (bankrupt - solvent)/|solvent|', DIFF_PLOT_FILE);
    logmsg('Relative difference chart saved to ' . DIFF_PLOT_FILE);

    logmsg('Analysis complete.');
} catch (Throwable $e) {
    logmsg('ERROR: ' . $e->getMessage());
}

?>