File: quantile_confidence_estimation.rst

package info (click to toggle)
openturns 1.26-4
  • links: PTS, VCS
  • area: main
  • in suites: forky, sid
  • size: 67,708 kB
  • sloc: cpp: 261,605; python: 67,030; ansic: 4,378; javascript: 406; sh: 185; xml: 164; makefile: 101
file content (371 lines) | stat: -rw-r--r-- 16,934 bytes parent folder | download
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
.. _quantile_confidence_estimation:

Exact quantile confidence interval based on order statistics
------------------------------------------------------------

We consider a random variable :math:`X` of dimension 1 and its quantile :math:`x_{\alpha}`
of level :math:`\alpha` (:math:`\alpha \in [0, 1]`).
We seek to evaluate an upper bound of :math:`x_{\alpha}` with a confidence greater or equal to
:math:`\beta`, using order statistics.

Let :math:`(X_1, \dots, X_\sampleSize)` be some independent copies of :math:`X`.
Let :math:`X_{(k)}` be the :math:`k` -th order statistics of :math:`(X_1, \dots, X_\sampleSize)` which means that
:math:`X_{(k)}` is the :math:`k` -th minimum of :math:`(X_1, \dots, X_\sampleSize)` for :math:`1 \leq k \leq \sampleSize`. For
example, :math:`X_{(1)} = \min (X_1, \dots, X_\sampleSize)` is the minimum
and :math:`X_{(\sampleSize)} = \max (X_1, \dots, X_\sampleSize)` is the maximum. We have:

.. math::

    X_{(1)} \leq X_{(2)} \leq \dots \leq X_{(\sampleSize)}

The probability density and cumulative distribution functions of the order
statistics :math:`X_{(k)}` are:

.. math::
    :label: DistOrderStat

    F_{X_{(k)}}(x) & = \sum_{i=k}^{\sampleSize} \binom{\sampleSize}{i}\left(F(x)
    \right)^i \left(1-F(x)
    \right)^{\sampleSize-i} \\
    p_{X_{(k)}}(x) & = (\sampleSize-k+1)\binom{\sampleSize}{k-1}\left(F(x)\right)^{k-1}
    \left(1-F(x)
    \right)^{\sampleSize-k} p(x)

We notice that :math:`F_{X_{(k)}}(x) = \overline{F}_{(\sampleSize,F(x))}(k-1)` where
:math:`F_{(\sampleSize,F(x))}` is the cumulative
distribution function of the Binomial distribution :math:`\cB(\sampleSize,F(x))` and
:math:`\overline{F}_{(\sampleSize,F(x))}(k) = 1 - F_{(\sampleSize,F(x))}(k)` is the
complementary cumulative distribution fonction (also named survival function in dimension
1).
Therefore:

.. math::

    F_{X_{(k)}}(x_{\alpha}) = \sum_{i=k}^{\sampleSize} \binom{\sampleSize}{i} \alpha^i (1-\alpha)^{\sampleSize-i}
    = \overline{F}_{(\sampleSize,\alpha)}(k-1)

Rank for an upper bound of the quantile
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Let :math:`(x_1, \dots, x_\sampleSize)` be an i.i.d. sample of size :math:`\sampleSize` of
the random variable :math:`X`.
Given a quantile level :math:`\alpha \in [0,1]`, a confidence level
:math:`\beta \in [0,1]`, and a sample size :math:`\sampleSize`, we seek the smallest
rank :math:`k \in \llbracket 1, \sampleSize \rrbracket` such that:

.. math::
    :label: EqOrderStatB

    \Prob{x_{\alpha} \leq X_{(k)}} \geq \beta

As equation :eq:`EqOrderStatB` can be written as:

.. math::
    :label: EqOrderStat2B

    1-F_{X_{(k)}}(x_{\alpha})\geq \beta

or even as:

.. math::

    F_{\sampleSize, \alpha}(k-1)\geq \beta

Then, the smallest rank :math:`k_{sol}` such that the previous equation is satisfied is:

.. math::

    k_{sol} & = \min \{ k \in \llbracket 1, \sampleSize \rrbracket \, | \, F_{\sampleSize, \alpha}(k-1)\geq \beta \}\\
            & = 1 + \min \{ k \in \llbracket 1, \sampleSize\rrbracket \, | \, F_{\sampleSize, \alpha}(k)\geq \beta \}

An upper bound of :math:`x_{\alpha}` is estimated by the value of :math:`X_{(k_{sol})}`
on the sample :math:`(x_1, \dots, x_\sampleSize)`.

Here is a recap of the existence of solutions for this case:

+------------------------+------------------+-------------------------------------+---------------------------------+
| :math:`k_{sol}`        | :math:`\beta=0`  | :math:`0 < \beta < 1`               | :math:`\beta=1`                 |
+========================+==================+=====================================+=================================+
| :math:`\alpha=0`       | 1                | 1                                   | 1                               |
+------------------------+------------------+-------------------------------------+---------------------------------+
| :math:`0 < \alpha < 1` | 1                | see :eq:`EqOrderStatBgen`           | :math:`\emptyset`               |
+------------------------+------------------+-------------------------------------+---------------------------------+
| :math:`\alpha=1`       | 1                | :math:`\emptyset`                   | :math:`\emptyset`               |
+------------------------+------------------+-------------------------------------+---------------------------------+

where:

.. math::
    :label: EqOrderStatBgen

    k_{sol} & = 1 + F_{\sampleSize,\alpha}^{-1}(\beta)   \quad \text{if} \quad  1-\alpha^\sampleSize \geq \beta \\
            & = \emptyset   \quad \text{else}


Rank for a lower bound of the quantile
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Given the same data as previoulsy, we seek the greatest rank :math:`k \in \llbracket 1, \sampleSize \rrbracket` such that:

.. math::
    :label: EqOrderStatA

    \Prob{X_{(k)} \leq x_{\alpha}} \geq \beta

which can be written as:

.. math::
    :label: EqOrderStat2A

    F_{X_{(k)}}(x_{\alpha})\geq \beta

and finally as:

.. math::

    F_{\sampleSize, \alpha}(k - 1)\leq 1 - \beta

Then, the greatest rank :math:`k_{sol}` such that the previous equation is satisfied is:


.. math::

    k_{sol} & = \max \{ k \in \llbracket 1, \sampleSize \rrbracket \, | \, F_{\sampleSize, \alpha}(k-1)\leq \beta \}\\
            & = 1+\max \{ k \in \llbracket 1, \sampleSize\rrbracket \, | \, F_{\sampleSize, \alpha}(k)\leq \beta \}

Here is a recap of the existence of solutions for this case:

+------------------------+--------------------------------------+-----------------------------------------+---------------------------------+
| :math:`k_{sol}`        | :math:`\beta=0`                      | :math:`0 < \beta < 1`                   | :math:`\beta=1`                 |
+========================+======================================+=========================================+=================================+
| :math:`\alpha=0`       | :math:`\sampleSize`                  | :math:`\emptyset`                       | :math:`\emptyset`               |
+------------------------+--------------------------------------+-----------------------------------------+---------------------------------+
| :math:`0 < \alpha < 1` | :math:`\sampleSize`                  | see :eq:`EqOrderStatAgen`               | :math:`\emptyset`               |
+------------------------+--------------------------------------+-----------------------------------------+---------------------------------+
| :math:`\alpha=1`       | :math:`\sampleSize`                  |  :math:`\sampleSize`                    | :math:`\sampleSize`             |
+------------------------+--------------------------------------+-----------------------------------------+---------------------------------+

where:

.. math::
    :label: EqOrderStatAgen

    k_{sol} & = \emptyset \quad \text{if} \quad  (1-\alpha)^\sampleSize > 1 - \beta \\
            & = 1 + F_{\sampleSize,\alpha}^{-1}(1-\beta)  \quad  \text{otherwise and if} \quad  \exists k_0 \, | \, 1-\beta = F_{(\sampleSize,\alpha}(k_0 - 1) \\
            & = F_{\sampleSize,\alpha}^{-1}(1-\beta)  \quad  \text{else}


Ranks for bilateral bounds of the quantile
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Given the same data as previoulsy, we can seek the ranks
:math:`k_1, k_2 \in \llbracket 1, \sampleSize \rrbracket^2` as solution of different problems.

The problem can be:

.. math::
    :label: EqOrderStatC3

    \begin{array}{ll}
    (k_1, k_2) = & \argmin \Prob{X_{(k_1)} \leq x_{\alpha} \leq X_{(k_2)}}\\
                 & \mbox{s.t.} \Prob{X_{(k_1)} \leq x_{\alpha} \leq X_{(k_2)}} \geq \beta
    \end{array}


or:

.. math::
    :label: EqOrderStatC4

    \begin{array}{ll}
    (k_1, k_2) = & \argmin (k_2-k_1)\\
                 & \mbox{s.t.} \Prob{X_{(k_1)} \leq x_{\alpha} \leq X_{(k_2)}} \geq \beta
    \end{array}


or:

.. math::
    :label: EqOrderStatC1

    \Prob{X_{(k_1)} \leq x_{\alpha} } \geq 1 - \dfrac{1-\beta}{2}\\
    \Prob{x_{\alpha} \leq X_{(k_2)}} \geq 1 - \dfrac{1-\beta}{2}


or with :math:`(k_1, k_2) = (k,\sampleSize-k+1)` and :math:`1 \leq k \leq \sampleSize` the greatest rank such that:

.. math::
    :label: EqOrderStatC2

    \Prob{X_{(k)} \leq x_{\alpha} \leq X_{(\sampleSize-k+1)}} \geq \beta.


The solutions of :eq:`EqOrderStatC3` and :eq:`EqOrderStatC4` are determined numerically, using an optimization algorithm.

The solutions of :eq:`EqOrderStatC1` are respectively defined by:

.. math::

   \overline{F}_{(\sampleSize,\alpha)}(k_1 - 1) \leq \dfrac{1 - \beta}{2} \\
   F_{(\sampleSize,\alpha)}(k_2 - 1) \geq 1-\dfrac{1 - \beta}{2}

which leads to the respective solutions:

.. math::

    k_{1, sol} & = \max \{ k \in \llbracket 1, \sampleSize \rrbracket \, | \, F_{\sampleSize, \alpha}(k-1)\leq \dfrac{1 - \beta}{2} \}\\
            & = 1 + \max \{ k \in \llbracket 1, \sampleSize\rrbracket \, | \, F_{\sampleSize, \alpha}(k)\leq \dfrac{1-\beta}{2} \}

and

.. math::

    k_{2, sol} & = \min \{ k \in \llbracket 1, \sampleSize \rrbracket \, | \, F_{\sampleSize, \alpha}(k - 1)\geq 1- \dfrac{1 - \beta}{2} \}\\
            & = 1 + \min \{ k \in \llbracket 1, \sampleSize\rrbracket \, | \, F_{\sampleSize, \alpha}(k)\geq 1 - \dfrac{1 - \beta}{2} \}

Then, the previous tables written for the lower and upper bounds can be used to find :math:`k_{1, sol}` and :math:`k_{2, sol}` respectively with
:math:`\beta \rightarrow \dfrac{1-\beta}{2}` or :math:`\beta \rightarrow 1-\dfrac{1-\beta}{2}`.

The solutions of :eq:`EqOrderStatC2` are gathered here:

+------------------------+-----------------------------------------------------------+---------------------------------+-------------------------+
| :math:`k_{sol}`        | :math:`\beta=0`                                           | :math:`0 < \beta < 1`           | :math:`\beta=1`         |
+========================+===========================================================+=================================+=========================+
| :math:`\alpha=0`       | :math:`\Bigl\lfloor \frac{n}{2} \Bigr\rfloor`             | :math:`\emptyset`               | :math:`\emptyset`       |
+------------------------+-----------------------------------------------------------+---------------------------------+-------------------------+
| :math:`0 < \alpha < 1` | 1                                                         | :math:`\emptyset` or 1          | :math:`\emptyset`       |
+------------------------+-----------------------------------------------------------+---------------------------------+-------------------------+
| :math:`\alpha=1`       | :math:`\Bigl\lfloor \frac{\sampleSize}{2} \Bigr\rfloor`   | :math:`\emptyset`               | :math:`\emptyset`       |
+------------------------+-------------------------------------------------+---------------------------------+-----------------------------------+

Minimum sample size for an upper bound of the quantile
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Given :math:`\alpha`, :math:`\beta`, and the rank :math:`1 \leq k \leq \sampleSize`, we seek the smallest sample size
:math:`\sampleSize` such that:

.. math::
    :label: EqOrderStatBbis

    \Prob{x_{\alpha} \leq X_{(\sampleSize-k+1)}} \geq \beta

As equation :eq:`EqOrderStatBbis` can be written as:

.. math::
    :label: EqOrderStat2Bbis

    1-F_{X_{(\sampleSize-k+1)}}(x_{\alpha})\geq \beta

or even as:

.. math::

    F_{\sampleSize, \alpha}(\sampleSize-k)\geq \beta

Note that the problem is defined differently than in equation :eq:`EqOrderStatB`. In order to do so, we solve
equation :eq:`EqOrderStat2Bbis` with respect to the sample size :math:`\sampleSize`. We use an optimization algorithm to determined
:math:`n_{sol}` in the interval :math:`\llbracket k, +\infty \llbracket`. We can reduce the research interval to the interval
:math:`\llbracket k, n_2 \rrbracket` where :math:`n_2` is a size
that verifies equation :eq:`EqOrderStat2Bbis`. It
can be determined using the approximation of the binomial distribution by the normal distribution with the same mean and variance.

Once the smallest size :math:`\sampleSize` has been estimated, a sample of size :math:`\sampleSize` can be
generated from :math:`X` and an upper bound of :math:`x_{\alpha}` is estimated using
:math:`x_{(n-k+1)}` i.e. the :math:`k`-th observation
in the decreasing ordered sample :math:`(x_{(\sampleSize)}, \dots, x_{(1)})`.

Minimum sample size for a lower bound of the quantile
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Given the same data as previoulsy, we seek the smallest sample size :math:`\sampleSize`
such that equation :eq:`EqOrderStatA` is satisfied.

Here is a recap of the existence of solutions for this case:

+------------------------+--------------------------------------+---------------------------------------------+---------------------------------+
|       :math:`n_{sol}`  | :math:`\beta=0`                      | :math:`0 < \beta < 1`                       | :math:`\beta=1`                 |
+========================+======================================+=============================================+=================================+
| :math:`\alpha=0`       | :math:`k`                            | :math:`\emptyset`                           | :math:`\emptyset`               |
+------------------------+--------------------------------------+---------------------------------------------+---------------------------------+
| :math:`0 < \alpha < 1` | :math:`\argmin \{\sampleSize \geq k | F_{\sampleSize,\alpha}(k-1) \leq 1-\beta \}` | :math:`\emptyset`               |
+------------------------+--------------------------------------+---------------------------------------------+---------------------------------+
| :math:`\alpha=1`       | :math:`k`                            | :math:`k`                                   | :math:`k`                       |
+------------------------+--------------------------------------+---------------------------------------------+---------------------------------+


Minimum sample size for bilateral bounds of the quantile
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Given two order statistics :math:`(k_1, k_2)` with :math:`1 \leq k_1 < k_2 \leq \sampleSize`, we seek the smallest sample size :math:`\sampleSize`
such that:

.. math::
    :label: EqOrderMinSizeBil

    \Prob{X_{(k_1)} \leq x_{\alpha} \leq X_{(n-k_2+1)}} \geq \beta

As equation :eq:`EqOrderMinSizeBil` can be written as:

.. math::
    :label: EqOrderMinSizeBilsol

    F_{X_{(\sampleSize-k_2+1)}}(x_{\alpha}) - F_{X_{(k_1)}}(x_{\alpha}) \geq \beta

or even as:

.. math::

    F_{\sampleSize, \alpha}(\sampleSize-k_2) - F_{\sampleSize, \alpha}(k_1-1)\geq \beta


Note that the problem is defined differently than in equation :eq:`EqOrderStatC4`. In order to do so, we solve
equation :eq:`EqOrderMinSizeBilsol` with respect to the sample size :math:`\sampleSize`. We use an optimization algorithm to determined
:math:`n_{sol}` in the interval :math:`\llbracket k, +\infty \llbracket`. We can reduce the research interval to the interval
:math:`\llbracket k, n_2 \rrbracket` where :math:`n_2` is a size
that verifies equation :eq:`EqOrderStat2Bbis`. It
can be determined using the approximation of the binomial distribution by the normal distribution with the same mean and variance.

Once the smallest size :math:`\sampleSize` has been estimated, a sample of size :math:`\sampleSize` can be
generated from :math:`X` and an lower and upper bound of :math:`x_{\alpha}` is estimated using
:math:`x_{(k_1)}` and :math:`x_{(n-k_2+1)}` i.e. the :math:`k_1`-th observation
in the ordered sample :math:`(x_{(1)}, \dots, x_{(\sampleSize)})` and the :math:`\sampleSize-k_2`-th observation
in the decreasing ordered sample :math:`(x_{(\sampleSize)}, \dots, x_{(1)})`.

In the particular case where :math:`(k_1, k_2) = (1,1)`, we seek the smallest sample size :math:`\sampleSize`
such that:

.. math::

    \Prob{ \min (X_1, \dots, X_\sampleSize) \leq x_{\alpha} \leq  \max (X_1, \dots, X_\sampleSize)} \geq \beta.

Then, equantion :eq:`EqOrderMinSizeBilsol` can be written as:

.. math::

    1-\alpha^\sampleSize - (1-\alpha)^\sampleSize \geq \beta.

The optimal :math:`\sampleSize` is determined using an optimization algorithm which research is reduced to the interval:

.. math::

    \left \lfloor \dfrac{\log (1-\beta)}{\log \gamma} \right \rfloor \leq n \leq \left \lfloor \dfrac{\log \left(\dfrac{1-\beta}{2}\right)}{\log \gamma} \right \rfloor

where :math:`\gamma = \max(\alpha, 1-\alpha)`.

.. topic:: API:

    - See :class:`~openturns.experimental.QuantileConfidence`
    - See :class:`~openturns.Sample`

.. topic:: Examples:

    - See :doc:`/auto_data_analysis/sample_analysis/plot_quantile_confidence_estimation`
    - See :doc:`/auto_data_analysis/sample_analysis/plot_quantile_confidence_chemical_process`
    - See :doc:`/auto_data_analysis/sample_analysis/plot_quantile_confidence_exact_empirical`

.. topic:: References:

    - [meeker2017]_
    - [wilks1948]_
    - [robert2004]_
    - [rubinstein2017]_