Skip to content

manifold

Bases: object

Base class for manifold learning.

This class provides a generic interface for manifold learning algorithms, with methods to set the number of neighbors and components and to perform dimensionality reduction using a specified manifold learning technique.

Attributes:

Name Type Description
name str

The name of the manifold learning method.
n_neighbors int

The number of neighbors to consider for the manifold learning algorithm.
n_components int

The number of components for dimensionality reduction.
model object

The manifold learning model to be used.

Methods:

Name Description
get_n_neighbors

Get the number of neighbors used in the manifold learning algorithm.
set_n_neighbors

Set the number of neighbors and reinitialize the model.
get_n_components

Get the number of components for dimensionality reduction.
set_n_components

Set the number of components and reinitialize the model.
forward

Apply the manifold learning algorithm to reduce the dimensionality of the input data.
fit_transform

Perform dimensionality reduction on the input data.
init_model

Abstract method to initialize the manifold learning model.
Source code in tinybig/koala/manifold/manifold.py 
 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
class manifold(object):
    """
        Base class for manifold learning.

        This class provides a generic interface for manifold learning algorithms, with methods to set the number of
        neighbors and components and to perform dimensionality reduction using a specified manifold learning technique.

        Attributes
        ----------
        name : str
            The name of the manifold learning method.
        n_neighbors : int
            The number of neighbors to consider for the manifold learning algorithm.
        n_components : int
            The number of components for dimensionality reduction.
        model : object
            The manifold learning model to be used.

        Methods
        -------
        get_n_neighbors()
            Get the number of neighbors used in the manifold learning algorithm.
        set_n_neighbors(n_neighbors)
            Set the number of neighbors and reinitialize the model.
        get_n_components()
            Get the number of components for dimensionality reduction.
        set_n_components(n_components)
            Set the number of components and reinitialize the model.
        forward(X, device='cup', *args, **kwargs)
            Apply the manifold learning algorithm to reduce the dimensionality of the input data.
        fit_transform(X, device='cup', *args, **kwargs)
            Perform dimensionality reduction on the input data.
        init_model()
            Abstract method to initialize the manifold learning model.
    """
    def __init__(self, name: str = 'base_manifold', n_neighbors: int = 5, n_components: int = 2, *args, **kwargs):
        """
            Initialize the manifold learning class.

            Parameters
            ----------
            name : str, optional
                The name of the manifold learning method. Default is 'base_manifold'.
            n_neighbors : int, optional
                The number of neighbors to consider. Default is 5.
            n_components : int, optional
                The number of components for dimensionality reduction. Default is 2.
            *args, **kwargs
                Additional arguments for initialization.
        """
        self.name = name
        self.n_neighbors = n_neighbors
        self.n_components = n_components

        self.model = None
        self.init_model()

    def get_n_neighbors(self):
        """
            Get the number of neighbors used in the manifold learning algorithm.

            Returns
            -------
            int
                The number of neighbors.
        """
        return self.n_neighbors

    def set_n_neighbors(self, n_neighbors):
        """
            Set the number of neighbors and reinitialize the model.

            Parameters
            ----------
            n_neighbors : int
                The new number of neighbors.
        """
        self.n_neighbors = n_neighbors
        self.init_model()

    def get_n_components(self):
        """
        Get the number of components for dimensionality reduction.

        Returns
        -------
        int
            The number of components.
        """
        return self.n_components

    def set_n_components(self, n_components):
        """
        Set the number of components and reinitialize the model.

        Parameters
        ----------
        n_components : int
            The new number of components.
        """
        self.n_components = n_components
        self.init_model()

    def __call__(self, X: Union[np.ndarray, torch.Tensor], device: str = 'cup', *args, **kwargs):
        """
        Apply the manifold learning algorithm to the input data.

        Parameters
        ----------
        X : Union[np.ndarray, torch.Tensor]
            The input data for dimensionality reduction.
        device : str, optional
            The device to use ('cup' or 'cpu'). Default is 'cup'.
        *args, **kwargs
            Additional arguments.

        Returns
        -------
        Union[np.ndarray, torch.Tensor]
            The transformed data.
        """
        return self.forward(X=X, device=device, *args, **kwargs)

    def forward(self, X: Union[np.ndarray, torch.Tensor], device: str = 'cup', *args, **kwargs):
        """
        Perform dimensionality reduction on the input data.

        Parameters
        ----------
        X : Union[np.ndarray, torch.Tensor]
            The input data for dimensionality reduction.
        device : str, optional
            The device to use ('cup' or 'cpu'). Default is 'cup'.
        *args, **kwargs
            Additional arguments.

        Returns
        -------
        Union[np.ndarray, torch.Tensor]
            The transformed data.
        """
        return self.fit_transform(X=X, device=device, *args, **kwargs)

    def fit_transform(self, X: Union[np.ndarray, torch.Tensor], device: str = 'cup', *args, **kwargs):
        """
        Perform dimensionality reduction on the input data.

        Parameters
        ----------
        X : Union[np.ndarray, torch.Tensor]
            The input data for dimensionality reduction.
        device : str, optional
            The device to use ('cup' or 'cpu'). Default is 'cup'.
        *args, **kwargs
            Additional arguments.

        Returns
        -------
        Union[np.ndarray, torch.Tensor]
            The transformed data.
        """
        if isinstance(X, torch.Tensor):
            input_X = X.detach().cpu().numpy()  # Convert torch.Tensor to numpy
        else:
            input_X = X

        X_manifold = self.model.fit_transform(X=input_X)

        return torch.tensor(X_manifold) if isinstance(X, torch.Tensor) and not isinstance(X_manifold, torch.Tensor) else X_manifold

    @abstractmethod
    def init_model(self):
        """
        Abstract method to initialize the manifold learning model.
        """
        pass

__call__(X, device='cup', *args, **kwargs)

Apply the manifold learning algorithm to the input data.

Parameters:

Name Type Description Default
X Union[ndarray, Tensor]

The input data for dimensionality reduction.

 required
device str

The device to use ('cup' or 'cpu'). Default is 'cup'.

 'cup'
*args

Additional arguments.

 ()
**kwargs

Additional arguments.

 ()

Returns:

Type Description
Union[ndarray, Tensor]

The transformed data.
Source code in tinybig/koala/manifold/manifold.py 
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
def __call__(self, X: Union[np.ndarray, torch.Tensor], device: str = 'cup', *args, **kwargs):
    """
    Apply the manifold learning algorithm to the input data.

    Parameters
    ----------
    X : Union[np.ndarray, torch.Tensor]
        The input data for dimensionality reduction.
    device : str, optional
        The device to use ('cup' or 'cpu'). Default is 'cup'.
    *args, **kwargs
        Additional arguments.

    Returns
    -------
    Union[np.ndarray, torch.Tensor]
        The transformed data.
    """
    return self.forward(X=X, device=device, *args, **kwargs)

__init__(name='base_manifold', n_neighbors=5, n_components=2, *args, **kwargs)

Initialize the manifold learning class.

Parameters:

Name Type Description Default
name str

The name of the manifold learning method. Default is 'base_manifold'.

 'base_manifold'
n_neighbors int

The number of neighbors to consider. Default is 5.

 5
n_components int

The number of components for dimensionality reduction. Default is 2.

 2
*args

Additional arguments for initialization.

 ()
**kwargs

Additional arguments for initialization.

 ()
Source code in tinybig/koala/manifold/manifold.py 
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
def __init__(self, name: str = 'base_manifold', n_neighbors: int = 5, n_components: int = 2, *args, **kwargs):
    """
        Initialize the manifold learning class.

        Parameters
        ----------
        name : str, optional
            The name of the manifold learning method. Default is 'base_manifold'.
        n_neighbors : int, optional
            The number of neighbors to consider. Default is 5.
        n_components : int, optional
            The number of components for dimensionality reduction. Default is 2.
        *args, **kwargs
            Additional arguments for initialization.
    """
    self.name = name
    self.n_neighbors = n_neighbors
    self.n_components = n_components

    self.model = None
    self.init_model()

fit_transform(X, device='cup', *args, **kwargs)

Perform dimensionality reduction on the input data.

Parameters:

Name Type Description Default
X Union[ndarray, Tensor]

The input data for dimensionality reduction.

 required
device str

The device to use ('cup' or 'cpu'). Default is 'cup'.

 'cup'
*args

Additional arguments.

 ()
**kwargs

Additional arguments.

 ()

Returns:

Type Description
Union[ndarray, Tensor]

The transformed data.
Source code in tinybig/koala/manifold/manifold.py 
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
def fit_transform(self, X: Union[np.ndarray, torch.Tensor], device: str = 'cup', *args, **kwargs):
    """
    Perform dimensionality reduction on the input data.

    Parameters
    ----------
    X : Union[np.ndarray, torch.Tensor]
        The input data for dimensionality reduction.
    device : str, optional
        The device to use ('cup' or 'cpu'). Default is 'cup'.
    *args, **kwargs
        Additional arguments.

    Returns
    -------
    Union[np.ndarray, torch.Tensor]
        The transformed data.
    """
    if isinstance(X, torch.Tensor):
        input_X = X.detach().cpu().numpy()  # Convert torch.Tensor to numpy
    else:
        input_X = X

    X_manifold = self.model.fit_transform(X=input_X)

    return torch.tensor(X_manifold) if isinstance(X, torch.Tensor) and not isinstance(X_manifold, torch.Tensor) else X_manifold

forward(X, device='cup', *args, **kwargs)

Perform dimensionality reduction on the input data.

Parameters:

Name Type Description Default
X Union[ndarray, Tensor]

The input data for dimensionality reduction.

 required
device str

The device to use ('cup' or 'cpu'). Default is 'cup'.

 'cup'
*args

Additional arguments.

 ()
**kwargs

Additional arguments.

 ()

Returns:

Type Description
Union[ndarray, Tensor]

The transformed data.
Source code in tinybig/koala/manifold/manifold.py 
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
def forward(self, X: Union[np.ndarray, torch.Tensor], device: str = 'cup', *args, **kwargs):
    """
    Perform dimensionality reduction on the input data.

    Parameters
    ----------
    X : Union[np.ndarray, torch.Tensor]
        The input data for dimensionality reduction.
    device : str, optional
        The device to use ('cup' or 'cpu'). Default is 'cup'.
    *args, **kwargs
        Additional arguments.

    Returns
    -------
    Union[np.ndarray, torch.Tensor]
        The transformed data.
    """
    return self.fit_transform(X=X, device=device, *args, **kwargs)

get_n_components()

Get the number of components for dimensionality reduction.

Returns:

Type Description
int

The number of components.
Source code in tinybig/koala/manifold/manifold.py 
 99
100
101
102
103
104
105
106
107
108
def get_n_components(self):
    """
    Get the number of components for dimensionality reduction.

    Returns
    -------
    int
        The number of components.
    """
    return self.n_components

get_n_neighbors()

Get the number of neighbors used in the manifold learning algorithm.

Returns:

Type Description
int

The number of neighbors.
Source code in tinybig/koala/manifold/manifold.py 
76
77
78
79
80
81
82
83
84
85
def get_n_neighbors(self):
    """
        Get the number of neighbors used in the manifold learning algorithm.

        Returns
        -------
        int
            The number of neighbors.
    """
    return self.n_neighbors

init_model() abstractmethod

Abstract method to initialize the manifold learning model.

Source code in tinybig/koala/manifold/manifold.py 
189
190
191
192
193
194
@abstractmethod
def init_model(self):
    """
    Abstract method to initialize the manifold learning model.
    """
    pass

set_n_components(n_components)

Set the number of components and reinitialize the model.

Parameters:

Name Type Description Default
n_components int

The new number of components.

 required
Source code in tinybig/koala/manifold/manifold.py 
110
111
112
113
114
115
116
117
118
119
120
def set_n_components(self, n_components):
    """
    Set the number of components and reinitialize the model.

    Parameters
    ----------
    n_components : int
        The new number of components.
    """
    self.n_components = n_components
    self.init_model()

set_n_neighbors(n_neighbors)

Set the number of neighbors and reinitialize the model.

Parameters:

Name Type Description Default
n_neighbors int

The new number of neighbors.

 required
Source code in tinybig/koala/manifold/manifold.py 
87
88
89
90
91
92
93
94
95
96
97
def set_n_neighbors(self, n_neighbors):
    """
        Set the number of neighbors and reinitialize the model.

        Parameters
        ----------
        n_neighbors : int
            The new number of neighbors.
    """
    self.n_neighbors = n_neighbors
    self.init_model()