StatisticalTestHinkley¶

class StatisticalTestHinkley(*args, **kwargs)¶

Bases: StatisticalTestAbstract

This class implements the Hinkley’s cumulative sum test.

The Hinkley’s cumulative sum test is designed to detect drift in the mean of an observed signal \(s(t)\) . It is known to be robust (by taking into account all the past of the observed quantity), efficient, and inducing a very low computational load. The other attractive features of this test are two-fold. First, it can straightforwardly and accurately provide the drift instant. Secondly, due to its formulation (cumulative sum test), it can simultaneously handle both very abrupt and important changes, and gradual smaller ones without adapting the involved thresholds.

Two tests are performed in parallel to look for downwards or upwards drifts in \(s(t)\) , respectively defined by:

\[S_k = \sum_{t=0}^{k} (s(t) - m_0 + \frac{\delta}{2}) \]

\[M_k = \max_{0 \leq i \leq k} S_i\]

\[T_k = \sum_{t=0}^{k} (s(t) - m_0 - \frac{\delta}{2}) \]

\[N_k = \min_{0 \leq i \leq k} T_i\]

In which \(m_o\) is computed on-line and corresponds to the mean of the signal \(s(t)\) we want to detect a drift. \(m_o\) is re-initialized at zero after each drift detection. \(\delta\) denotes the drift minimal magnitude that we want to detect and \(\alpha\) is a predefined threshold. These values are set by default to 0.2 in the default constructor vpHinkley() . To modify the default values use setAlpha() and setDelta() or the vpHinkley(double alpha, double delta) constructor.

A downward drift is detected if \(M_k - S_k > \alpha\) . A upward drift is detected if \(T_k - N_k > \alpha\) .

To detect only downward drifts in \(s(t)\) use testDownwardMeanDrift() .To detect only upward drifts in \(s(t)\) use testUpwardMeanDrift() . To detect both, downward and upward drifts use testDownUpwardMeanDrift() .

If a drift is detected, the drift location is given by the last instant \(k^{'}\) when \(M_{k^{'}} - S_{k^{'}} = 0\) , or \(T_{k^{'}} - N_{k^{'}} = 0\) .

Overloaded function.

__init__(self: visp._visp.core.StatisticalTestHinkley) -> None

Construct a new vpStatisticalTestHinkley object. Call init() to initialise the Hinkley’s test and set \(\alpha\) and \(\delta\) to default values.

By default \(\delta = 0.2\) and \(\alpha = 0.2\) . Use setDelta() and setAlpha() to modify these values.

__init__(self: visp._visp.core.StatisticalTestHinkley, alpha: float, delta: float, nbSamplesForInit: int = 30) -> None

Call init() to initialise the Hinkley’s test and set \(\alpha\) and \(\delta\) thresholds.

Note

See setAlpha() , setDelta()

Parameters:

alpha: math:alpha threshold indicating that a mean drift occurs.
delta: math:delta denotes the drift minimal magnitude that we want to detect.
nbSamplesForInit: number of signal samples to initialize the mean of the signal.

__init__(self: visp._visp.core.StatisticalTestHinkley, h: float, k: float, computeAlphaDeltaFromStdev: bool, nbSamplesForInit: int = 30) -> None

Construct a new vpStatisticalTestHinkley object. \(\alpha\) and \(\delta\) will be computed from the standard deviation of the signal.

Parameters:

h: the alarm factor that permits to compute \(\alpha\) from the standard deviation.
k: the detection factor that permits to compute \(\delta\) from the standard deviation.
computeAlphaDeltaFromStdev: must be equal to true, otherwise throw a vpException .
nbSamplesForInit: number of signal samples to initialize the mean of the signal.

__init__(self: visp._visp.core.StatisticalTestHinkley, h: float, k: float, mean: float, stdev: float) -> None

Construct a new vpStatisticalTestHinkley object. \(\alpha\) and \(\delta\) will be computed from the standard deviation of the signal.

Parameters:

h: the alarm factor that permits to compute \(\alpha\) from the standard deviation.
k: the detection factor that permits to compute \(\delta\) from the standard deviation.
mean: the expected mean of the signal.
stdev: the expected standard deviation of the signal.

Methods

`__init__`	Overloaded function.
`getAlpha`	Get the \(\alpha\) threshold indicating that a mean drift occurs.
`getMk`	Get the maximum of the test signal for downward mean drift \(S_k\) .
`getNk`	Get the minimum of the test signal for upward mean drift \(T_k\) .
`getSk`	Get the test signal for downward mean drift.
`getTk`	Get the test signal for upward mean drift..
`init`	Overloaded function.
`setAlpha`	The threshold indicating that a mean drift occurs.
`setDelta`	Set the drift minimal magnitude that we want to detect.

Inherited Methods

`getAvailableMeanDriftType`	Get the list of available vpMeanDriftType objects that are handled.
`setMinStdev`	Set the minimum value of the standard deviation that is expected.
`testUpwardMeanDrift`	Test if an upward mean drift occurred according to the new value of the signal.
`MEAN_DRIFT_DOWNWARD`
`MEAN_DRIFT_NONE`
`setNbSamplesForStat`	Set the number of samples required to compute the mean and standard deviation of the signal and allocate the memory accordingly.
`getStdev`	Get the standard deviation used as reference.
`testDownUpwardMeanDrift`	Test if a downward or an upward mean drift occurred according to the new value of the signal.
`MeanDriftType`	Enum that indicates if a drift of the mean occurred.
`vpMeanDriftTypeToString`
`MEAN_DRIFT_BOTH`
`MEAN_DRIFT_UNKNOWN`
`vpMeanDriftTypeFromString`	Cast a string into a vpMeanDriftType .
`getLimits`	Get the upper and lower limits of the test signal.
`MEAN_DRIFT_UPWARD`
`MEAN_DRIFT_COUNT`
`testDownwardMeanDrift`	Test if a downward mean drift occurred according to the new value of the signal.
`print`
`getMean`	Get the mean used as reference.

Operators

`__doc__`
`__init__`	Overloaded function.
`__module__`

Attributes

`MEAN_DRIFT_BOTH`
`MEAN_DRIFT_COUNT`
`MEAN_DRIFT_DOWNWARD`
`MEAN_DRIFT_NONE`
`MEAN_DRIFT_UNKNOWN`
`MEAN_DRIFT_UPWARD`
`__annotations__`

class MeanDriftType(self, value: int)¶

Bases: pybind11_object

Enum that indicates if a drift of the mean occurred.

Values:

MEAN_DRIFT_NONE: No mean drift occurred
MEAN_DRIFT_DOWNWARD: A downward drift of the mean occurred.
MEAN_DRIFT_UPWARD: An upward drift of the mean occurred.
MEAN_DRIFT_BOTH: Both an aupward and a downward drifts occurred.
MEAN_DRIFT_COUNT
MEAN_DRIFT_UNKNOWN

__and__(self, other: object) → object¶

__eq__(self, other: object) → bool¶

__ge__(self, other: object) → bool¶

__getstate__(self) → int¶

__gt__(self, other: object) → bool¶

__hash__(self) → int¶

__index__(self) → int¶

__init__(self, value: int)¶

__int__(self) → int¶

__invert__(self) → object¶

__le__(self, other: object) → bool¶

__lt__(self, other: object) → bool¶

__ne__(self, other: object) → bool¶

__or__(self, other: object) → object¶

__rand__(self, other: object) → object¶

__ror__(self, other: object) → object¶

__rxor__(self, other: object) → object¶

__setstate__(self, state: int) → None¶

__xor__(self, other: object) → object¶

property name : str¶

__init__(*args, **kwargs)¶

Overloaded function.

__init__(self: visp._visp.core.StatisticalTestHinkley) -> None

Construct a new vpStatisticalTestHinkley object. Call init() to initialise the Hinkley’s test and set \(\alpha\) and \(\delta\) to default values.

By default \(\delta = 0.2\) and \(\alpha = 0.2\) . Use setDelta() and setAlpha() to modify these values.

__init__(self: visp._visp.core.StatisticalTestHinkley, alpha: float, delta: float, nbSamplesForInit: int = 30) -> None

Call init() to initialise the Hinkley’s test and set \(\alpha\) and \(\delta\) thresholds.

Note

See setAlpha() , setDelta()

Parameters:

alpha: math:alpha threshold indicating that a mean drift occurs.
delta: math:delta denotes the drift minimal magnitude that we want to detect.
nbSamplesForInit: number of signal samples to initialize the mean of the signal.

__init__(self: visp._visp.core.StatisticalTestHinkley, h: float, k: float, computeAlphaDeltaFromStdev: bool, nbSamplesForInit: int = 30) -> None

Construct a new vpStatisticalTestHinkley object. \(\alpha\) and \(\delta\) will be computed from the standard deviation of the signal.

Parameters:

h: the alarm factor that permits to compute \(\alpha\) from the standard deviation.
k: the detection factor that permits to compute \(\delta\) from the standard deviation.
computeAlphaDeltaFromStdev: must be equal to true, otherwise throw a vpException .
nbSamplesForInit: number of signal samples to initialize the mean of the signal.

__init__(self: visp._visp.core.StatisticalTestHinkley, h: float, k: float, mean: float, stdev: float) -> None

Construct a new vpStatisticalTestHinkley object. \(\alpha\) and \(\delta\) will be computed from the standard deviation of the signal.

Parameters:

h: the alarm factor that permits to compute \(\alpha\) from the standard deviation.
k: the detection factor that permits to compute \(\delta\) from the standard deviation.
mean: the expected mean of the signal.
stdev: the expected standard deviation of the signal.

getAlpha(self) → float¶

Get the \(\alpha\) threshold indicating that a mean drift occurs.

Returns:: The \(\alpha\) threshold.

static getAvailableMeanDriftType(prefix: str = <, sep: str =, suffix: str = >) → str¶

Get the list of available vpMeanDriftType objects that are handled.

Parameters:

prefix: The prefix that should be placed before the list.
sep: The separator between each element of the list.
suffix: The suffix that should terminate the list.

Returns:

std::string The list of handled type of process tests, presented as a string.

getLimits(self, limitDown: float, limitUp: float) → tuple[float, float]¶

Get the upper and lower limits of the test signal.

Parameters:

limitDown: float¶: The lower limit.
limitUp: float¶: The upper limit.

Returns:

A tuple containing:

limitDown: The lower limit.
limitUp: The upper limit.

getMean(self) → float¶

Get the mean used as reference.

Returns:: float The mean.

getMk(self) → float¶

Get the maximum of the test signal for downward mean drift \(S_k\) .

Returns:: The value of \(M_k\) , the maximum value of \(S_k\) .

getNk(self) → float¶

Get the minimum of the test signal for upward mean drift \(T_k\) .

Returns:: The value of \(N_k\) , the minimum value of \(T_k\) .

getSk(self) → float¶

Get the test signal for downward mean drift.

Returns:: The value of \(S_k = \sum_{t=0}^{k} (s(t) - m_0 + \frac{\delta}{2})\) .

getStdev(self) → float¶

Get the standard deviation used as reference.

Returns:: float The standard deviation.

getTk(self) → float¶

Get the test signal for upward mean drift..

Returns:: The value of \(T_k = \sum_{t=0}^{k} (s(t) - m_0 - \frac{\delta}{2})\) .

init(*args, **kwargs)¶

Overloaded function.

init(self: visp._visp.core.StatisticalTestHinkley) -> None

Initialise the Hinkley’s test by setting the mean signal value \(m_0\) to zero as well as \(S_k, M_k, T_k, N_k\) .

init(self: visp._visp.core.StatisticalTestHinkley, alpha: float, delta: float, nbSamplesForInit: int) -> None

Call init() to initialise the Hinkley’s test and set \(\alpha\) and \(\delta\) thresholds.

Parameters:

alpha: The threshold indicating that a mean drift occurs.
delta: The drift minimal magnitude that we want to detect.
nbSamplesForInit: number of signal samples to initialize the mean of the signal.

init(self: visp._visp.core.StatisticalTestHinkley, h: float, k: float, computeAlphaDeltaFromStdev: bool, nbSamplesForInit: int) -> None

(Re)Initialize a new vpStatisticalTestHinkley object. \(\alpha\) and \(\delta\) will be computed from the standard deviation of the signal.

Parameters:

h: the alarm factor that permits to compute \(\alpha\) from the standard deviation.
k: the detection factor that permits to compute \(\delta\) from the standard deviation.
computeAlphaDeltaFromStdev: must be equal to true, otherwise throw a vpException .
nbSamplesForInit: number of signal samples to initialize the mean of the signal.

init(self: visp._visp.core.StatisticalTestHinkley, alpha: float, delta: float, mean: float) -> None

Call init() to initialise the Hinkley’s test, set \(\alpha\) and \(\delta\) thresholds, and the mean of the signal \(m_0\) .

Parameters:

alpha: The threshold indicating that a mean drift occurs.
delta: The drift minimal magnitude that we want to detect.
mean: The expected value of the mean.

init(self: visp._visp.core.StatisticalTestHinkley, h: float, k: float, mean: float, stdev: float) -> None

(Re)Initialize a new vpStatisticalTestHinkley object. \(\alpha\) and \(\delta\) will be computed from the standard deviation of the signal.

Parameters:

h: the alarm factor that permits to compute \(\alpha\) from the standard deviation.
k: the detection factor that permits to compute \(\delta\) from the standard deviation.
mean: the expected mean of the signal.
stdev: the expected standard deviation of the signal.

init(self: visp._visp.core.StatisticalTestAbstract) -> None

(Re)Initialize the algorithm.

static print(type: visp._visp.core.StatisticalTestAbstract.MeanDriftType) → None¶

setAlpha(self, alpha: float) → None¶

The threshold indicating that a mean drift occurs.

Parameters:

alpha: float¶: The threshold.

setDelta(self, delta: float) → None¶

Set the drift minimal magnitude that we want to detect.

Parameters:

delta: float¶: The drift magnitude.

setMinStdev(self, stdevmin: float) → None¶

Set the minimum value of the standard deviation that is expected. The computed standard deviation cannot be lower this value if set.

Parameters:

stdevmin: float¶: The minimum value of the standard deviation that is expected.

setNbSamplesForStat(self, nbSamples: int) → None¶

Set the number of samples required to compute the mean and standard deviation of the signal and allocate the memory accordingly.

Parameters:

nbSamples: int¶: The number of samples we want to use.

testDownUpwardMeanDrift(self, signal: float) → visp._visp.core.StatisticalTestAbstract.MeanDriftType¶

Test if a downward or an upward mean drift occurred according to the new value of the signal.

Note

See testDownwardMeanDrift() testUpwardMeanDrift()

Parameters:

signal: float¶: The new value of the signal.

Returns:

vpMeanDriftType The type of mean drift that occurred.

testDownwardMeanDrift(self, signal: float) → visp._visp.core.StatisticalTestAbstract.MeanDriftType¶

Test if a downward mean drift occurred according to the new value of the signal.

Note

See testUpwardMeanDrift()

Parameters:

signal: float¶: The new value of the signal.

Returns:

vpMeanDriftType The type of mean drift that occurred.

testUpwardMeanDrift(self, signal: float) → visp._visp.core.StatisticalTestAbstract.MeanDriftType¶

Test if an upward mean drift occurred according to the new value of the signal.

Note

See testDownwardMeanDrift()

Parameters:

signal: float¶: The new value of the signal.

Returns:

vpMeanDriftType The type of mean drift that occurred.

static vpMeanDriftTypeFromString(name: str) → visp._visp.core.StatisticalTestAbstract.MeanDriftType¶

Cast a string into a vpMeanDriftType .

Parameters:

name: str¶: The name of the mean drift.

Returns:

vpMeanDriftType The corresponding vpMeanDriftType .

static vpMeanDriftTypeToString(type: visp._visp.core.StatisticalTestAbstract.MeanDriftType) → str¶