1
|
|
|
<?php |
2
|
|
|
namespace HuasoFoundries\Histogram; |
3
|
|
|
|
4
|
|
|
class Histogram extends AbstractHistogram |
5
|
|
|
{ |
6
|
|
|
|
7
|
|
|
/** |
8
|
|
|
* Constructor |
9
|
|
|
* |
10
|
|
|
* @access public |
11
|
|
|
* @param optional int $type one of HISTOGRAM_SIMPLE or HISTOGRAM_CUMMULATIVE |
12
|
|
|
* @param optional int $nbins number of bins to use |
13
|
|
|
* @param optional float $rangeLow lowest value to use for bin frequency calculation |
14
|
|
|
* @param optional float $rangeHigh highest value to use for bin frequency calculation |
15
|
|
|
* @return object Histogram |
16
|
|
|
* |
17
|
|
|
* @see setBinOptions() |
18
|
|
|
* @see Math_AbstractHistogram::setType() |
19
|
|
|
* @see Math_AbstractHistogram |
20
|
|
|
*/ |
21
|
|
|
public function __construct($type = self::HISTOGRAM_SIMPLE, $nbins = -1, $rangeLow = null, $rangeHigh = null) |
22
|
|
|
{ |
23
|
|
|
$this->setType($type); |
24
|
|
|
$this->setBinOptions($nbins, $rangeLow, $rangeHigh); |
25
|
|
|
} |
26
|
|
|
|
27
|
|
|
/** |
28
|
|
|
* Sets the binning options. Overrides parent's method. |
29
|
|
|
* |
30
|
|
|
* @access public |
31
|
|
|
* @param int $nbins the number of bins to use for computing the histogram |
32
|
|
|
* @param optional float $rangeLow lowest value to use for bin frequency calculation |
33
|
|
|
* @param optional float $rangeHigh highest value to use for bin frequency calculation |
34
|
|
|
* @return void |
35
|
|
|
*/ |
36
|
|
|
public function setBinOptions($nbins, $rangeLow = null, $rangeHigh = null) |
37
|
|
|
{ |
38
|
|
|
$this->_nbins = (is_int($nbins) && $nbins > 2) ? $nbins : 10; |
39
|
|
|
$this->_rangeLow = $rangeLow; |
40
|
|
|
$this->_rangeHigh = $rangeHigh; |
41
|
|
|
} |
42
|
|
|
|
43
|
|
|
/** |
44
|
|
|
* Returns an associative array with the bin options |
45
|
|
|
* |
46
|
|
|
* @access public |
47
|
|
|
* @return array Associative array of bin options: |
48
|
|
|
* array( 'nbins'=>$nbins, |
49
|
|
|
* 'rangeLow'=>$rangeLow, |
50
|
|
|
* 'rangeHigh'=>$rangeHigh); |
51
|
|
|
*/ |
52
|
|
|
public function getBinOptions() |
53
|
|
|
{ |
54
|
|
|
return array( |
55
|
|
|
'nbins' => $this->_nbins, |
56
|
|
|
'rangeLow' => $this->_rangeLow, |
57
|
|
|
'rangeHigh' => $this->_rangeHigh, |
58
|
|
|
); |
59
|
|
|
} |
60
|
|
|
|
61
|
|
|
/** |
62
|
|
|
* Sets the data to be processed. The data will be validated to |
63
|
|
|
* be a simple uni-dimensional numerical array |
64
|
|
|
* |
65
|
|
|
* @access public |
66
|
|
|
* @param array $data the numeric array |
67
|
|
|
* @return mixed boolean true on success, a \PEAR_Error object otherwise |
68
|
|
|
* |
69
|
|
|
* @see _clear() |
70
|
|
|
* @see Math_AbstractHistogram::getData() |
71
|
|
|
* @see Math_AbstractHistogram |
72
|
|
|
* @see getHistogramData() |
73
|
|
|
*/ |
74
|
|
|
public function setData($data) |
75
|
|
|
{ |
76
|
|
|
$this->_clear(); |
77
|
|
|
if (!is_array($data)) { |
|
|
|
|
78
|
|
|
throw new \PEAR_Exception("array of numeric data expected"); |
79
|
|
|
} |
80
|
|
|
|
81
|
|
|
foreach ($data as $item) { |
82
|
|
|
if (!is_numeric($item)) { |
83
|
|
|
throw new \PEAR_Exception("non-numeric item in array"); |
84
|
|
|
} |
85
|
|
|
} |
86
|
|
|
|
87
|
|
|
$this->_data = $data; |
88
|
|
|
if (is_null($this->_rangeLow)) { |
89
|
|
|
$this->_rangeLow = min($this->_data); |
90
|
|
|
} |
91
|
|
|
|
92
|
|
|
if (is_null($this->_rangeHigh)) { |
93
|
|
|
$this->_rangeHigh = max($this->_data); |
94
|
|
|
} |
95
|
|
|
|
96
|
|
|
sort($this->_data); |
97
|
|
|
return true; |
98
|
|
|
} |
99
|
|
|
|
100
|
|
|
/** |
101
|
|
|
* Calculates the histogram bins and frequencies |
102
|
|
|
* |
103
|
|
|
* @access public |
104
|
|
|
* @param optional $statsMode calculate basic statistics (STATS_BASIC) or full (STATS_FULL) |
|
|
|
|
105
|
|
|
* @return mixed boolean true on success, a \PEAR_Error object otherwise |
106
|
|
|
* |
107
|
|
|
* @see Math_Stats |
108
|
|
|
*/ |
109
|
|
|
public function calculate($statsMode = \HuasoFoundries\Math\Stats::STATS_BASIC) |
110
|
|
|
{ |
111
|
|
|
$this->_stats = new \HuasoFoundries\Math\Stats(); |
112
|
|
|
|
113
|
|
|
$this->_statsMode = $statsMode; |
|
|
|
|
114
|
|
|
$delta = ($this->_rangeHigh - $this->_rangeLow) / $this->_nbins; |
115
|
|
|
$lastpos = 0; |
|
|
|
|
116
|
|
|
$cumm = 0; |
117
|
|
|
$data = $this->_histogramData(); |
118
|
|
|
$ndata = count($data); |
119
|
|
|
$ignoreList = array(); |
120
|
|
|
|
121
|
|
|
for ($i = 0; $i < $this->_nbins; $i++) { |
122
|
|
|
$loBin = $this->_rangeLow + $i * $delta; |
123
|
|
|
$hiBin = $loBin + $delta; |
124
|
|
|
$this->_bins[$i]["low"] = $loBin; |
125
|
|
|
$this->_bins[$i]["high"] = $hiBin; |
126
|
|
|
$this->_bins[$i]["mid"] = ($hiBin + $loBin) / 2; |
127
|
|
|
if ($this->_type == self::HISTOGRAM_CUMMULATIVE) { |
128
|
|
|
$this->_bins[$i]["count"] = $cumm; |
129
|
|
|
} else { |
130
|
|
|
$this->_bins[$i]["count"] = 0; |
131
|
|
|
} |
132
|
|
|
|
133
|
|
|
for ($j = 0; $j < $ndata; $j++) { |
134
|
|
|
if (!empty($ignoreList) && in_array($j, $ignoreList)) { |
135
|
|
|
continue; |
136
|
|
|
} |
137
|
|
|
|
138
|
|
|
if ($j == 0) { |
139
|
|
|
$inRange = ($loBin <= $data[$j] && $hiBin >= $data[$j]); |
140
|
|
|
} else { |
141
|
|
|
$inRange = ($loBin < $data[$j] && $hiBin >= $data[$j]); |
142
|
|
|
} |
143
|
|
|
|
144
|
|
|
if ($inRange) { |
145
|
|
|
$this->_bins[$i]["count"]++; |
146
|
|
|
if ($this->_type == self::HISTOGRAM_CUMMULATIVE) { |
147
|
|
|
$cumm++; |
148
|
|
|
} |
149
|
|
|
|
150
|
|
|
$ignoreList[] = $j; |
151
|
|
|
} |
152
|
|
|
} |
153
|
|
|
} |
154
|
|
|
return true; |
155
|
|
|
} |
156
|
|
|
|
157
|
|
|
/** |
158
|
|
|
* Returns the statistics for the data set |
159
|
|
|
* |
160
|
|
|
* @access public |
161
|
|
|
* @return mixed an associative array on success, a \PEAR_Error object otherwise |
162
|
|
|
*/ |
163
|
|
|
public function getDataStats() |
164
|
|
|
{ |
165
|
|
|
if ($this->isCalculated()) { |
166
|
|
|
$this->_stats->setData($this->_data); |
167
|
|
|
|
168
|
|
|
return $this->_stats->calc($this->_statsMode); |
169
|
|
|
} else { |
170
|
|
|
throw new \PEAR_Exception("histogram has not been calculated"); |
171
|
|
|
} |
172
|
|
|
} |
173
|
|
|
|
174
|
|
|
/** |
175
|
|
|
* Returns the statistics for the data set, filtered using the bin range |
176
|
|
|
* |
177
|
|
|
* @access public |
178
|
|
|
* @return mixed an associative array on success, a \PEAR_Error object otherwise |
179
|
|
|
*/ |
180
|
|
|
public function getHistogramDataStats() |
181
|
|
|
{ |
182
|
|
|
if ($this->isCalculated()) { |
183
|
|
|
$this->_stats->setData($this->_histogramData()); |
184
|
|
|
return $this->_stats->calc($this->_statsMode); |
185
|
|
|
} else { |
186
|
|
|
throw new \PEAR_Exception("histogram has not been calculated"); |
187
|
|
|
} |
188
|
|
|
} |
189
|
|
|
|
190
|
|
|
/** |
191
|
|
|
* Returns the bins and frequencies calculated using the given |
192
|
|
|
* bin mode and separator |
193
|
|
|
* |
194
|
|
|
* @access public |
195
|
|
|
* @param int $mode one of HISTOGRAM_LO_BINS, HISTOGRAM_MID_BINS (default), or HISTOGRAM_HI_BINS |
196
|
|
|
* @param string $separator the separator, default ", " |
197
|
|
|
* @return mixed a string on success, a \PEAR_Error object otherwise |
198
|
|
|
*/ |
199
|
|
|
public function toSeparated($mode = self::HISTOGRAM_MID_BINS, $separator = ", ") |
200
|
|
|
{ |
201
|
|
|
$bins = $this->getBins($mode); |
202
|
|
|
if (\PEAR::isError($bins)) { |
203
|
|
|
return $bins; |
204
|
|
|
} |
205
|
|
|
|
206
|
|
|
$nbins = count($bins); |
|
|
|
|
207
|
|
|
$out = array("# bin{$separator}frequency"); |
208
|
|
|
foreach ($bins as $bin => $freq) { |
209
|
|
|
$out[] = "{$bin}{$separator}{$freq}"; |
210
|
|
|
} |
211
|
|
|
|
212
|
|
|
return implode("\n", $out) . "\n"; |
213
|
|
|
} |
214
|
|
|
|
215
|
|
|
/** |
216
|
|
|
* Static method to check that an object is a Histogram instance |
217
|
|
|
* |
218
|
|
|
* @static |
219
|
|
|
* @access public |
220
|
|
|
* @param object Histogram $hist An instance of the Histogram class |
221
|
|
|
* @return boolean TRUE on success, FALSE otherwise |
222
|
|
|
*/ |
223
|
|
|
public static function isValidHistogram(&$hist) |
224
|
|
|
{ |
225
|
|
|
return (is_object($hist) && is_a($hist, '\HuasoFoundries\Histogram\Histogram')); |
226
|
|
|
} |
227
|
|
|
|
228
|
|
|
/** |
229
|
|
|
* Method to interrogate if a histogram has been calculated |
230
|
|
|
* |
231
|
|
|
* @access public |
232
|
|
|
* @return boolean TRUE if the histogram was calculated, FALSE otherwise |
233
|
|
|
* @see calculate() |
234
|
|
|
*/ |
235
|
|
|
public function isCalculated() |
236
|
|
|
{ |
237
|
|
|
return !empty($this->_bins); |
238
|
|
|
} |
239
|
|
|
|
240
|
|
|
/** |
241
|
|
|
* Generates a plot using the appropriate printer object |
242
|
|
|
* |
243
|
|
|
* @param object $printer A Histogram_Printer_* object |
244
|
|
|
* @return string|PEAR_Error A string on success, a \PEAR_Error otherwise |
|
|
|
|
245
|
|
|
*/ |
246
|
|
|
public function generatePlot(&$printer) |
247
|
|
|
{ |
248
|
|
|
if (is_object($printer) && is_a($printer, '\HuasoFoundries\Histogram\Printer\Common')) { |
249
|
|
|
$printer->setHistogram($this); |
250
|
|
|
return $printer->generateOutput(); |
251
|
|
|
} else { |
252
|
|
|
throw new \PEAR_Exception('Invalid object, expecting a \HuasoFoundries\Histogram\Printer\* instance'); |
253
|
|
|
} |
254
|
|
|
} |
255
|
|
|
|
256
|
|
|
/** |
257
|
|
|
* Prints a simple ASCII representation of the histogram |
258
|
|
|
* |
259
|
|
|
* @deprecated |
260
|
|
|
* @access public |
261
|
|
|
* @param optional int $mode one of HISTOGRAM_LO_BINS, HISTOGRAM_MID_BINS, or HISTOGRAM_HI_BINS (default) |
262
|
|
|
* @return mixed a string on success, a \PEAR_Error object otherwise |
263
|
|
|
*/ |
264
|
|
|
public function printHistogram($mode = self::HISTOGRAM_HI_BINS) |
265
|
|
|
{ |
266
|
|
|
if (!$this->isCalculated()) { |
267
|
|
|
throw new \PEAR_Exception("histogram has not been calculated"); |
268
|
|
|
} |
269
|
|
|
$out = ($this->_type == self::HISTOGRAM_CUMMULATIVE) ? "Cummulative Frequency" : "Histogram"; |
270
|
|
|
$out .= "\n\tNumber of bins: " . $this->_nbins . "\n"; |
271
|
|
|
$out .= "\tPlot range: [" . $this->_rangeLow . ", " . $this->_rangeHigh . "]\n"; |
272
|
|
|
$hdata = $this->_histogramData(); |
273
|
|
|
$out .= "\tData range: [" . min($hdata) . ", " . max($hdata) . "]\n"; |
274
|
|
|
$out .= "\tOriginal data range: [" . min($this->_data) . ", " . max($this->_data) . "]\n"; |
275
|
|
|
$out .= "BIN (FREQUENCY) ASCII_BAR (%)\n"; |
276
|
|
|
$fmt = "%-4.3f (%-4d) |%s\n"; |
277
|
|
|
$bins = $this->_filterBins($mode); |
278
|
|
|
$maxfreq = max(array_values($bins)); |
279
|
|
|
$total = count($this->_data); |
280
|
|
|
foreach ($bins as $bin => $freq) { |
281
|
|
|
$out .= sprintf($fmt, $bin, $freq, $this->_bar($freq, $maxfreq, $total)); |
282
|
|
|
} |
283
|
|
|
|
284
|
|
|
return $out; |
285
|
|
|
} |
286
|
|
|
|
287
|
|
|
/** |
288
|
|
|
* Prints a simple ASCII bar |
289
|
|
|
* |
290
|
|
|
* @access private |
291
|
|
|
* @param int $freq the frequency |
292
|
|
|
* @param int $maxfreq the maximum frequency |
293
|
|
|
* @param int $total the total count |
294
|
|
|
* @return string |
295
|
|
|
*/ |
296
|
|
|
public function _bar($freq, $maxfreq, $total) |
297
|
|
|
{ |
298
|
|
|
$fact = floatval(($maxfreq > 40) ? 40 / $maxfreq : 1); |
299
|
|
|
$niter = round($freq * $fact); |
300
|
|
|
$out = ""; |
301
|
|
|
for ($i = 0; $i < $niter; $i++) { |
302
|
|
|
$out .= "*"; |
303
|
|
|
} |
304
|
|
|
|
305
|
|
|
return $out . sprintf(" (%.1f%%)", $freq / $total * 100); |
306
|
|
|
} |
307
|
|
|
|
308
|
|
|
/** |
309
|
|
|
* Returns a subset of the bins array by bin value type |
310
|
|
|
* |
311
|
|
|
* @access private |
312
|
|
|
* @param int $mode one of HISTOGRAM_MID_BINS, HISTOGRAM_LO_BINS or HISTOGRAM_HI_BINS |
313
|
|
|
* @return array |
314
|
|
|
*/ |
315
|
|
|
public function _filterBins($mode) |
316
|
|
|
{ |
317
|
|
|
$map = array( |
318
|
|
|
self::HISTOGRAM_MID_BINS => "mid", |
319
|
|
|
self::HISTOGRAM_LO_BINS => "low", |
320
|
|
|
self::HISTOGRAM_HI_BINS => "high", |
321
|
|
|
); |
322
|
|
|
$filtered = array(); |
323
|
|
|
foreach ($this->_bins as $bin) { |
324
|
|
|
$filtered["{$bin[$map[$mode]]}"] = $bin["count"]; |
325
|
|
|
} |
326
|
|
|
|
327
|
|
|
return $filtered; |
328
|
|
|
} |
329
|
|
|
|
330
|
|
|
/** |
331
|
|
|
* Returns an array of data contained within the range for the |
332
|
|
|
* histogram calculation. Overrides the empty implementation in |
333
|
|
|
* Math_AbstractHistogram::_histogramData() |
334
|
|
|
* |
335
|
|
|
* @access private |
336
|
|
|
* @return array |
337
|
|
|
*/ |
338
|
|
|
public function _histogramData() |
339
|
|
|
{ |
340
|
|
|
$data = array(); |
341
|
|
|
foreach ($this->_data as $val) { |
342
|
|
|
if ($val < $this->_rangeLow || $val > $this->_rangeHigh) { |
343
|
|
|
continue; |
344
|
|
|
} else { |
345
|
|
|
$data[] = $val; |
346
|
|
|
} |
347
|
|
|
} |
348
|
|
|
|
349
|
|
|
return $data; |
350
|
|
|
} |
351
|
|
|
} |
352
|
|
|
|
353
|
|
|
// vim: ts=4:sw=4:et: |
354
|
|
|
// vim6: fdl=1: |
355
|
|
|
|