SMOF-MOF 0.3

No project description provided

SMOF-MOF

Installation

You can install SMOF-MOF using pip

pip install SMOF-MOF

Mass ratio variance-based outlier factor (MOF)

---

the outlier score of each data point is called MOF. It measures the global deviation of density given sample with respect to other data points. it is global in the outlier score depend on how isolated. data point is with respect to all data points in the data set. the variance of mass ratio can identify data points that have a substantially. lower density compared to other data points. These are considered outliers.

MOF()

Initial a model of MOF

Parameters :
Return :
        self : object
                object of MOF model

MOF.fit(Data)

Fit data to MOF model

Parameters :
        Data  : numpy array of shape (n_points, d_dimensions)
                The input samples.
Return :
        self  : object
                fitted estimator

Process_MOF(X =[])

Process data with MOF

Parameters :
        X  : numpy array of shape (n_points, d_dimensions)
                The input samples.
Return :
        decision_scores_ : numpy array of shape (n_samples)
                decision score for each point

MOF attributes

Attributes	Type	Details
MOF.Data	numpy array of shape (n_points, d_dimensions)	input data for model
MOF.MassRatio	numpy array of shape (n_samples, n_points)	MassRatio for each a pair of points
MOF.decision_scores_	numpy array of shape (n_samples)	decision score for each point

Examples

# This example demonstrates  the usage of MOF
from SMOF_MOF import Process_MOF
import numpy as np
X = [[ 0.74193734,  6.84385655,  5.90835464],
        [ 0.74193734,  6.69344758,  5.76863345],
        [ 0.74193734,  9.45125311,  7.21457799],
        [ 1.62924054,  7.58482389,  5.79026593],
        [ 1.80828877,  7.46054784,  5.79939564],
        [ 1.62924054 , 7.03623644,  5.78720408],
        [ 2.6461748 ,  9.44786798,  7.21604848],
        [ 0.09531018,  7.91465424,  6.00166237],
        [ 0.09531018,  6.61217543,  5.77796187]]
X = np.array(X)
score = Process_MOF(X)
print(scores)

Output

[0.05043327 0.03993228 0.34079475 0.07714844 0.10691165 0.063680160.4016755  0.81238359 0.04089301]

Note The data size should not exceed 10000 points because MOF uses high memory.

Streaming Mass ratio variance-based outlier factor (SMOF)

---

the outlier score of each data point is called SMOF. It measures the global deviation of density of data points in current window with respect to data points in the refernce window. this algoihtm process in non-overlapping window. the mutiply between variance of mass ratio of each data point and average of them in present window can identify data points that have a substantially lower density compared to other data points. These are considered outliers.

SMOF(win=256, seed=42, Train=[])

Initial a model of SMOF

Parameters :
        win : int (default = 256)
                window size of sum of reference and current window
        seed : int (default = 42)
                random seed
        Train : numpy array of shape (n_points, d_dimensions)
                data points for intilized model
Return :
        self : object
                object of SMOF model

SMOF.fit(X_current)

Move current data point to current window and assign SMOF scores

Parameters :
        X_current  : numpy array of shape (n_points, d_dimensions)
                current data point from streaming data
Return :
        self  : object
                fitted estimator

Process_SMOF(Win=500, X =[], Seed=42)

Process data with SMOF

Parameters :
        win : int (default = 500)
                window size of sum of reference and current window
        X  : numpy array of shape (n_points, d_dimensions)
                The input samples.
        seed : int (default = 42)
                random seed
Return :
        SMOFs : numpy array of shape (n_samples)
                decision score for each point

SMOF attributes

Attributes	Type	Details
MOF.win	int	window size of sum of reference and current window
MOF.window_current	numpy array of shape (n_samples, n_points)	current data point from streaming data
MOF.SMOFs	numpy array of shape (n_samples)	decision score for each point

Examples

# This example demonstrates  the usage of SMOF
from SMOF_MOF import Process_SMOF
import numpy as np
X = [[-2.30258509,  7.71114603,  5.81740873]
    [-2.30258509,  6.93449458,  5.81144035]
    [ 0.09531018,  7.26410017,  5.99670081]
    [ 1.13140211,  7.17019646,  5.80242044]
    [ 0.74193734,  7.06227732,  5.90835464]
    [-2.30258509,  6.64391993,  5.80844275]
    [ 1.13140211,  7.69807454,  5.91106761]
    [ 0.09531018,  6.79805201,  5.81442899]
    [ 0.09531018,  7.42362814,  5.92719266]
    [ 0.09531018,  6.9118469 ,  5.81144035]
    [-2.30258509,  7.01040212,  5.80242044]
    [ 0.09531018,  7.13894636,  5.91106761]
    [ 0.09531018,  7.61928251,  5.80242044]
    [ 0.09531018,  7.29580291,  6.01640103]
    [-2.30258509, 12.43197678,  5.79331844]
    [ 1.13140211,  9.53156118,  7.22336862]
    [-2.30258509,  7.09431783,  5.79939564]
    [ 0.09531018,  7.50444662,  5.82037962]
    [ 0.09531018,  7.8184705 ,  5.82334171]
    [ 0.09531018,  7.25212482,  5.91106761]]
X = np.array(X)
score = Process_SMOF(Win=10,X=X)
print(score)

Output

<runtime>  10it [00:00, 509.68it/s]
[0.02466581 0.00734274 0.00502668 0.00333289 0.00544248 0.0046538 0.00933626 0.00316922 0.00294779 0.00122089 0.01376553 0.00126867 0.00730856 0.00496102 0.03370523 0.13718442 0.03043219 0.02105821 0.02646965 0.00197488]