1 | /*************************************************************************
|
---|
2 | Copyright (c) 2007, Sergey Bochkanov (ALGLIB project).
|
---|
3 |
|
---|
4 | >>> SOURCE LICENSE >>>
|
---|
5 | This program is free software; you can redistribute it and/or modify
|
---|
6 | it under the terms of the GNU General Public License as published by
|
---|
7 | the Free Software Foundation (www.fsf.org); either version 2 of the
|
---|
8 | License, or (at your option) any later version.
|
---|
9 |
|
---|
10 | This program is distributed in the hope that it will be useful,
|
---|
11 | but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
12 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
13 | GNU General Public License for more details.
|
---|
14 |
|
---|
15 | A copy of the GNU General Public License is available at
|
---|
16 | http://www.fsf.org/licensing/licenses
|
---|
17 |
|
---|
18 | >>> END OF LICENSE >>>
|
---|
19 | *************************************************************************/
|
---|
20 |
|
---|
21 | using System;
|
---|
22 |
|
---|
23 | namespace alglib
|
---|
24 | {
|
---|
25 | public class descriptivestatistics
|
---|
26 | {
|
---|
27 | /*************************************************************************
|
---|
28 | Calculation of the distribution moments: mean, variance, slewness, kurtosis.
|
---|
29 |
|
---|
30 | Input parameters:
|
---|
31 | X - sample. Array with whose indexes range within [0..N-1]
|
---|
32 | N - sample size.
|
---|
33 |
|
---|
34 | Output parameters:
|
---|
35 | Mean - mean.
|
---|
36 | Variance- variance.
|
---|
37 | Skewness- skewness (if variance<>0; zero otherwise).
|
---|
38 | Kurtosis- kurtosis (if variance<>0; zero otherwise).
|
---|
39 |
|
---|
40 | -- ALGLIB --
|
---|
41 | Copyright 06.09.2006 by Bochkanov Sergey
|
---|
42 | *************************************************************************/
|
---|
43 | public static void calculatemoments(ref double[] x,
|
---|
44 | int n,
|
---|
45 | ref double mean,
|
---|
46 | ref double variance,
|
---|
47 | ref double skewness,
|
---|
48 | ref double kurtosis)
|
---|
49 | {
|
---|
50 | int i = 0;
|
---|
51 | double v = 0;
|
---|
52 | double v1 = 0;
|
---|
53 | double v2 = 0;
|
---|
54 | double stddev = 0;
|
---|
55 |
|
---|
56 | mean = 0;
|
---|
57 | variance = 0;
|
---|
58 | skewness = 0;
|
---|
59 | kurtosis = 0;
|
---|
60 | stddev = 0;
|
---|
61 | if( n<=0 )
|
---|
62 | {
|
---|
63 | return;
|
---|
64 | }
|
---|
65 |
|
---|
66 | //
|
---|
67 | // Mean
|
---|
68 | //
|
---|
69 | for(i=0; i<=n-1; i++)
|
---|
70 | {
|
---|
71 | mean = mean+x[i];
|
---|
72 | }
|
---|
73 | mean = mean/n;
|
---|
74 |
|
---|
75 | //
|
---|
76 | // Variance (using corrected two-pass algorithm)
|
---|
77 | //
|
---|
78 | if( n!=1 )
|
---|
79 | {
|
---|
80 | v1 = 0;
|
---|
81 | for(i=0; i<=n-1; i++)
|
---|
82 | {
|
---|
83 | v1 = v1+AP.Math.Sqr(x[i]-mean);
|
---|
84 | }
|
---|
85 | v2 = 0;
|
---|
86 | for(i=0; i<=n-1; i++)
|
---|
87 | {
|
---|
88 | v2 = v2+(x[i]-mean);
|
---|
89 | }
|
---|
90 | v2 = AP.Math.Sqr(v2)/n;
|
---|
91 | variance = (v1-v2)/(n-1);
|
---|
92 | if( (double)(variance)<(double)(0) )
|
---|
93 | {
|
---|
94 | variance = 0;
|
---|
95 | }
|
---|
96 | stddev = Math.Sqrt(variance);
|
---|
97 | }
|
---|
98 |
|
---|
99 | //
|
---|
100 | // Skewness and kurtosis
|
---|
101 | //
|
---|
102 | if( (double)(stddev)!=(double)(0) )
|
---|
103 | {
|
---|
104 | for(i=0; i<=n-1; i++)
|
---|
105 | {
|
---|
106 | v = (x[i]-mean)/stddev;
|
---|
107 | v2 = AP.Math.Sqr(v);
|
---|
108 | skewness = skewness+v2*v;
|
---|
109 | kurtosis = kurtosis+AP.Math.Sqr(v2);
|
---|
110 | }
|
---|
111 | skewness = skewness/n;
|
---|
112 | kurtosis = kurtosis/n-3;
|
---|
113 | }
|
---|
114 | }
|
---|
115 |
|
---|
116 |
|
---|
117 | /*************************************************************************
|
---|
118 | ADev
|
---|
119 |
|
---|
120 | Input parameters:
|
---|
121 | X - sample (array indexes: [0..N-1])
|
---|
122 | N - sample size
|
---|
123 |
|
---|
124 | Output parameters:
|
---|
125 | ADev- ADev
|
---|
126 |
|
---|
127 | -- ALGLIB --
|
---|
128 | Copyright 06.09.2006 by Bochkanov Sergey
|
---|
129 | *************************************************************************/
|
---|
130 | public static void calculateadev(ref double[] x,
|
---|
131 | int n,
|
---|
132 | ref double adev)
|
---|
133 | {
|
---|
134 | int i = 0;
|
---|
135 | double mean = 0;
|
---|
136 |
|
---|
137 | mean = 0;
|
---|
138 | adev = 0;
|
---|
139 | if( n<=0 )
|
---|
140 | {
|
---|
141 | return;
|
---|
142 | }
|
---|
143 |
|
---|
144 | //
|
---|
145 | // Mean
|
---|
146 | //
|
---|
147 | for(i=0; i<=n-1; i++)
|
---|
148 | {
|
---|
149 | mean = mean+x[i];
|
---|
150 | }
|
---|
151 | mean = mean/n;
|
---|
152 |
|
---|
153 | //
|
---|
154 | // ADev
|
---|
155 | //
|
---|
156 | for(i=0; i<=n-1; i++)
|
---|
157 | {
|
---|
158 | adev = adev+Math.Abs(x[i]-mean);
|
---|
159 | }
|
---|
160 | adev = adev/n;
|
---|
161 | }
|
---|
162 |
|
---|
163 |
|
---|
164 | /*************************************************************************
|
---|
165 | Median calculation.
|
---|
166 |
|
---|
167 | Input parameters:
|
---|
168 | X - sample (array indexes: [0..N-1])
|
---|
169 | N - sample size
|
---|
170 |
|
---|
171 | Output parameters:
|
---|
172 | Median
|
---|
173 |
|
---|
174 | -- ALGLIB --
|
---|
175 | Copyright 06.09.2006 by Bochkanov Sergey
|
---|
176 | *************************************************************************/
|
---|
177 | public static void calculatemedian(double[] x,
|
---|
178 | int n,
|
---|
179 | ref double median)
|
---|
180 | {
|
---|
181 | int i = 0;
|
---|
182 | int ir = 0;
|
---|
183 | int j = 0;
|
---|
184 | int l = 0;
|
---|
185 | int midp = 0;
|
---|
186 | int k = 0;
|
---|
187 | double a = 0;
|
---|
188 | double tval = 0;
|
---|
189 |
|
---|
190 | x = (double[])x.Clone();
|
---|
191 |
|
---|
192 |
|
---|
193 | //
|
---|
194 | // Some degenerate cases
|
---|
195 | //
|
---|
196 | median = 0;
|
---|
197 | if( n<=0 )
|
---|
198 | {
|
---|
199 | return;
|
---|
200 | }
|
---|
201 | if( n==1 )
|
---|
202 | {
|
---|
203 | median = x[0];
|
---|
204 | return;
|
---|
205 | }
|
---|
206 | if( n==2 )
|
---|
207 | {
|
---|
208 | median = 0.5*(x[0]+x[1]);
|
---|
209 | return;
|
---|
210 | }
|
---|
211 |
|
---|
212 | //
|
---|
213 | // Common case, N>=3.
|
---|
214 | // Choose X[(N-1)/2]
|
---|
215 | //
|
---|
216 | l = 0;
|
---|
217 | ir = n-1;
|
---|
218 | k = (n-1)/2;
|
---|
219 | while( true )
|
---|
220 | {
|
---|
221 | if( ir<=l+1 )
|
---|
222 | {
|
---|
223 |
|
---|
224 | //
|
---|
225 | // 1 or 2 elements in partition
|
---|
226 | //
|
---|
227 | if( ir==l+1 & (double)(x[ir])<(double)(x[l]) )
|
---|
228 | {
|
---|
229 | tval = x[l];
|
---|
230 | x[l] = x[ir];
|
---|
231 | x[ir] = tval;
|
---|
232 | }
|
---|
233 | break;
|
---|
234 | }
|
---|
235 | else
|
---|
236 | {
|
---|
237 | midp = (l+ir)/2;
|
---|
238 | tval = x[midp];
|
---|
239 | x[midp] = x[l+1];
|
---|
240 | x[l+1] = tval;
|
---|
241 | if( (double)(x[l])>(double)(x[ir]) )
|
---|
242 | {
|
---|
243 | tval = x[l];
|
---|
244 | x[l] = x[ir];
|
---|
245 | x[ir] = tval;
|
---|
246 | }
|
---|
247 | if( (double)(x[l+1])>(double)(x[ir]) )
|
---|
248 | {
|
---|
249 | tval = x[l+1];
|
---|
250 | x[l+1] = x[ir];
|
---|
251 | x[ir] = tval;
|
---|
252 | }
|
---|
253 | if( (double)(x[l])>(double)(x[l+1]) )
|
---|
254 | {
|
---|
255 | tval = x[l];
|
---|
256 | x[l] = x[l+1];
|
---|
257 | x[l+1] = tval;
|
---|
258 | }
|
---|
259 | i = l+1;
|
---|
260 | j = ir;
|
---|
261 | a = x[l+1];
|
---|
262 | while( true )
|
---|
263 | {
|
---|
264 | do
|
---|
265 | {
|
---|
266 | i = i+1;
|
---|
267 | }
|
---|
268 | while( (double)(x[i])<(double)(a) );
|
---|
269 | do
|
---|
270 | {
|
---|
271 | j = j-1;
|
---|
272 | }
|
---|
273 | while( (double)(x[j])>(double)(a) );
|
---|
274 | if( j<i )
|
---|
275 | {
|
---|
276 | break;
|
---|
277 | }
|
---|
278 | tval = x[i];
|
---|
279 | x[i] = x[j];
|
---|
280 | x[j] = tval;
|
---|
281 | }
|
---|
282 | x[l+1] = x[j];
|
---|
283 | x[j] = a;
|
---|
284 | if( j>=k )
|
---|
285 | {
|
---|
286 | ir = j-1;
|
---|
287 | }
|
---|
288 | if( j<=k )
|
---|
289 | {
|
---|
290 | l = i;
|
---|
291 | }
|
---|
292 | }
|
---|
293 | }
|
---|
294 |
|
---|
295 | //
|
---|
296 | // If N is odd, return result
|
---|
297 | //
|
---|
298 | if( n%2==1 )
|
---|
299 | {
|
---|
300 | median = x[k];
|
---|
301 | return;
|
---|
302 | }
|
---|
303 | a = x[n-1];
|
---|
304 | for(i=k+1; i<=n-1; i++)
|
---|
305 | {
|
---|
306 | if( (double)(x[i])<(double)(a) )
|
---|
307 | {
|
---|
308 | a = x[i];
|
---|
309 | }
|
---|
310 | }
|
---|
311 | median = 0.5*(x[k]+a);
|
---|
312 | }
|
---|
313 |
|
---|
314 |
|
---|
315 | /*************************************************************************
|
---|
316 | Percentile calculation.
|
---|
317 |
|
---|
318 | Input parameters:
|
---|
319 | X - sample (array indexes: [0..N-1])
|
---|
320 | N - sample size, N>1
|
---|
321 | P - percentile (0<=P<=1)
|
---|
322 |
|
---|
323 | Output parameters:
|
---|
324 | V - percentile
|
---|
325 |
|
---|
326 | -- ALGLIB --
|
---|
327 | Copyright 01.03.2008 by Bochkanov Sergey
|
---|
328 | *************************************************************************/
|
---|
329 | public static void calculatepercentile(double[] x,
|
---|
330 | int n,
|
---|
331 | double p,
|
---|
332 | ref double v)
|
---|
333 | {
|
---|
334 | int i1 = 0;
|
---|
335 | double t = 0;
|
---|
336 |
|
---|
337 | x = (double[])x.Clone();
|
---|
338 |
|
---|
339 | System.Diagnostics.Debug.Assert(n>1, "CalculatePercentile: N<=1!");
|
---|
340 | System.Diagnostics.Debug.Assert((double)(p)>=(double)(0) & (double)(p)<=(double)(1), "CalculatePercentile: incorrect P!");
|
---|
341 | internalstatheapsort(ref x, n);
|
---|
342 | if( (double)(p)==(double)(0) )
|
---|
343 | {
|
---|
344 | v = x[0];
|
---|
345 | return;
|
---|
346 | }
|
---|
347 | if( (double)(p)==(double)(1) )
|
---|
348 | {
|
---|
349 | v = x[n-1];
|
---|
350 | return;
|
---|
351 | }
|
---|
352 | t = p*(n-1);
|
---|
353 | i1 = (int)Math.Floor(t);
|
---|
354 | t = t-(int)Math.Floor(t);
|
---|
355 | v = x[i1]*(1-t)+x[i1+1]*t;
|
---|
356 | }
|
---|
357 |
|
---|
358 |
|
---|
359 | private static void internalstatheapsort(ref double[] arr,
|
---|
360 | int n)
|
---|
361 | {
|
---|
362 | int i = 0;
|
---|
363 | int k = 0;
|
---|
364 | int t = 0;
|
---|
365 | double tmp = 0;
|
---|
366 |
|
---|
367 | if( n==1 )
|
---|
368 | {
|
---|
369 | return;
|
---|
370 | }
|
---|
371 | i = 2;
|
---|
372 | do
|
---|
373 | {
|
---|
374 | t = i;
|
---|
375 | while( t!=1 )
|
---|
376 | {
|
---|
377 | k = t/2;
|
---|
378 | if( (double)(arr[k-1])>=(double)(arr[t-1]) )
|
---|
379 | {
|
---|
380 | t = 1;
|
---|
381 | }
|
---|
382 | else
|
---|
383 | {
|
---|
384 | tmp = arr[k-1];
|
---|
385 | arr[k-1] = arr[t-1];
|
---|
386 | arr[t-1] = tmp;
|
---|
387 | t = k;
|
---|
388 | }
|
---|
389 | }
|
---|
390 | i = i+1;
|
---|
391 | }
|
---|
392 | while( i<=n );
|
---|
393 | i = n-1;
|
---|
394 | do
|
---|
395 | {
|
---|
396 | tmp = arr[i];
|
---|
397 | arr[i] = arr[0];
|
---|
398 | arr[0] = tmp;
|
---|
399 | t = 1;
|
---|
400 | while( t!=0 )
|
---|
401 | {
|
---|
402 | k = 2*t;
|
---|
403 | if( k>i )
|
---|
404 | {
|
---|
405 | t = 0;
|
---|
406 | }
|
---|
407 | else
|
---|
408 | {
|
---|
409 | if( k<i )
|
---|
410 | {
|
---|
411 | if( (double)(arr[k])>(double)(arr[k-1]) )
|
---|
412 | {
|
---|
413 | k = k+1;
|
---|
414 | }
|
---|
415 | }
|
---|
416 | if( (double)(arr[t-1])>=(double)(arr[k-1]) )
|
---|
417 | {
|
---|
418 | t = 0;
|
---|
419 | }
|
---|
420 | else
|
---|
421 | {
|
---|
422 | tmp = arr[k-1];
|
---|
423 | arr[k-1] = arr[t-1];
|
---|
424 | arr[t-1] = tmp;
|
---|
425 | t = k;
|
---|
426 | }
|
---|
427 | }
|
---|
428 | }
|
---|
429 | i = i-1;
|
---|
430 | }
|
---|
431 | while( i>=1 );
|
---|
432 | }
|
---|
433 | }
|
---|
434 | }
|
---|