跳到内容

基础核函数

smac.model.gaussian_process.kernels.base_kernels #

AbstractKernel #

AbstractKernel(
    *,
    operate_on: ndarray | None = None,
    has_conditions: bool = False,
    prior: AbstractPrior | None = None,
    **kwargs: Any
)

这是一个用于覆盖核函数功能的 mixin。由于它覆盖了核函数的功能，因此需要在继承层次结构中排在首位。正因为如此，不可能从核函数类中继承这个 Mixin，因为这将阻止其被实例化。因此，mypy 将无法识别与超类相关的任何内容，并且在访问超类中声明的成员（如 self.has_conditions、self._call、super().get_params 等）时，需要添加一些 type:ignore 语句。

参数#

operate_on : np.ndarray, 默认为 None 在哪个 numpy 数组上进行操作。 has_conditions : bool, 默认为 False 核函数是否有条件。 prior : AbstractPrior, 默认为 None 核函数正在使用哪个先验。

属性#

operate_on : np.ndarray, 默认为 None 在哪个 numpy 数组上进行操作。 has_conditions : bool, 默认为 False 核函数是否有条件。可能由高斯过程改变。 prior : AbstractPrior, 默认为 None 核函数正在使用哪个先验。主要由 sklearn 使用。

源代码位于 smac/model/gaussian_process/kernels/base_kernels.py 
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
def __init__(
    self,
    *,
    operate_on: np.ndarray | None = None,
    has_conditions: bool = False,
    prior: AbstractPrior | None = None,
    **kwargs: Any,
) -> None:
    self.operate_on = operate_on
    self.has_conditions = has_conditions
    self.prior = prior
    self._set_active_dims(operate_on)

    # Since this class is a mixin, we just pass all the other parameters to the next class.
    super().__init__(**kwargs)

    # Get variables from next class:
    # We make it explicit here to make sure the next class really has this attributes.
    self._hyperparameters: list[kernels.Hyperparameter] = super().hyperparameters  # type: ignore
    self._n_dims: int = super().n_dims  # type: ignore
    self._len_active: int | None

hyperparameters property #

hyperparameters: list[Hyperparameter]

返回所有超参数规范的列表。

meta property #

meta: dict[str, Any]

返回创建对象的元数据。此方法调用 get_params 方法来收集核函数的参数。

n_dims property #

n_dims: int

返回核函数的非固定超参数数量。

__call__ #

__call__(
    X: ndarray,
    Y: ndarray | None = None,
    eval_gradient: bool = False,
    active: ndarray | None = None,
) -> ndarray | tuple[ndarray, ndarray]

调用核函数。内部调用 self._call，该方法必须由子类指定。

源代码位于 smac/model/gaussian_process/kernels/base_kernels.py 
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
def __call__(
    self,
    X: np.ndarray,
    Y: np.ndarray | None = None,
    eval_gradient: bool = False,
    active: np.ndarray | None = None,
) -> np.ndarray | tuple[np.ndarray, np.ndarray]:
    """Call the kernel function. Internally, `self._call` is called, which must be specified by a subclass."""
    if active is None and self.has_conditions:
        if self.operate_on is None:
            active = get_conditional_hyperparameters(X, Y)
        else:
            if Y is None:
                active = get_conditional_hyperparameters(X[:, self.operate_on], None)
            else:
                active = get_conditional_hyperparameters(X[:, self.operate_on], Y[:, self.operate_on])

    if self.operate_on is None:
        rval = self._call(X, Y, eval_gradient, active)
    else:
        if self._len_active is None:
            raise RuntimeError("The internal variable `_len_active` is not set.")

        if Y is None:
            rval = self._call(
                X=X[:, self.operate_on].reshape([-1, self._len_active]),
                Y=None,
                eval_gradient=eval_gradient,
                active=active,
            )
            X = X[:, self.operate_on].reshape((-1, self._len_active))
        else:
            rval = self._call(
                X=X[:, self.operate_on].reshape([-1, self._len_active]),
                Y=Y[:, self.operate_on].reshape([-1, self._len_active]),
                eval_gradient=eval_gradient,
                active=active,
            )
            X = X[:, self.operate_on].reshape((-1, self._len_active))
            Y = Y[:, self.operate_on].reshape((-1, self._len_active))

    return rval

get_params #

get_params(deep: bool = True) -> dict[str, Any]

获取此核函数的参数。

参数#

deep : bool, 默认为 True 如果为 True，将返回此估计器及其包含的子对象（也是估计器）的参数。

返回值#

params : dict[str, Any] 参数名映射到其值。

源代码位于 smac/model/gaussian_process/kernels/base_kernels.py 
 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
def get_params(self, deep: bool = True) -> dict[str, Any]:
    """Get parameters of this kernel.

    Parameters
    ----------
    deep : bool, defaults to True
        If True, will return the parameters for this estimator and
        contained subobjects that are estimators.

    Returns
    -------
    params : dict[str, Any]
        Parameter names mapped to their values.
    """
    params = {}

    # ignore[misc] looks like it catches all kinds of errors, but misc is actually a category from mypy:
    # https://mypy.readthedocs.io/en/latest/error_code_list.html#miscellaneous-checks-misc
    tmp = super().get_params(deep)  # type: ignore[misc] # noqa F821
    args = list(tmp.keys())

    # Sum and Product do not clone the 'has_conditions' attribute by default. Instead of changing their
    # get_params() method, we simply add the attribute here!
    if "has_conditions" not in args:
        args.append("has_conditions")

    for arg in args:
        params[arg] = getattr(self, arg, None)

    return params

ConstantKernel #

ConstantKernel(
    constant_value: float = 1.0,
    constant_value_bounds: tuple[float, float] = (
        1e-05,
        100000.0,
    ),
    operate_on: ndarray | None = None,
    has_conditions: bool = False,
    prior: AbstractPrior | None = None,
)

基类: AbstractKernel, ConstantKernel

源代码位于 smac/model/gaussian_process/kernels/base_kernels.py 
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
def __init__(
    self,
    constant_value: float = 1.0,
    constant_value_bounds: tuple[float, float] = (1e-5, 1e5),
    operate_on: np.ndarray | None = None,
    has_conditions: bool = False,
    prior: AbstractPrior | None = None,
) -> None:
    super().__init__(
        operate_on=operate_on,
        has_conditions=has_conditions,
        prior=prior,
        constant_value=constant_value,
        constant_value_bounds=constant_value_bounds,
    )

hyperparameters property #

hyperparameters: list[Hyperparameter]

返回所有超参数规范的列表。

meta property #

meta: dict[str, Any]

返回创建对象的元数据。此方法调用 get_params 方法来收集核函数的参数。

n_dims property #

n_dims: int

返回核函数的非固定超参数数量。

__call__ #

__call__(
    X: ndarray,
    Y: ndarray | None = None,
    eval_gradient: bool = False,
    active: ndarray | None = None,
) -> ndarray | tuple[ndarray, ndarray]

返回核函数 k(X, Y) 以及可选的其梯度。

参数#

X : np.ndarray, 形状 (n_samples_X, n_features) 返回的核函数 k(X, Y) 的左参数。

np.ndarray, 形状 (n_samples_Y, n_features), (可选, 默认为 None)

返回的核函数 k(X, Y) 的右参数。如果为 None，则计算 k(X, X)。

bool (可选, 默认为 False)

确定是否计算相对于核函数超参数的梯度。仅在 Y 为 None 时支持。

np.ndarray (n_samples_X, n_features) (可选)

布尔数组，指定哪些超参数是活动的。

返回值#

K : np.ndarray, 形状 (n_samples_X, n_samples_Y) 核函数 k(X, Y)。

np.ndarray (可选), 形状 (n_samples_X, n_samples_X, n_dims)

核函数 k(X, X) 相对于核函数超参数的梯度。仅当 eval_gradient 为 True 时返回。

源代码位于 smac/model/gaussian_process/kernels/base_kernels.py 
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
def __call__(
    self,
    X: np.ndarray,
    Y: np.ndarray | None = None,
    eval_gradient: bool = False,
    active: np.ndarray | None = None,
) -> np.ndarray | tuple[np.ndarray, np.ndarray]:
    """Return the kernel k(X, Y) and optionally its gradient.

    Parameters
    ----------
    X : np.ndarray, shape (n_samples_X, n_features)
        Left argument of the returned kernel k(X, Y).

    Y : np.ndarray, shape (n_samples_Y, n_features), (optional, default=None)
        Right argument of the returned kernel k(X, Y). If None, k(X, X)
        is evaluated instead.

    eval_gradient : bool (optional, default=False)
        Determines whether the gradient with respect to the kernel
        hyperparameter is determined. Only supported when Y is None.

    active : np.ndarray (n_samples_X, n_features) (optional)
        Boolean array specifying which hyperparameters are active.

    Returns
    -------
    K : np.ndarray, shape (n_samples_X, n_samples_Y)
        Kernel k(X, Y).

    K_gradient : np.ndarray (opt.), shape (n_samples_X, n_samples_X, n_dims)
        The gradient of the kernel k(X, X) with respect to the
        hyperparameter of the kernel. Only returned when eval_gradient
        is True.
    """
    X = np.atleast_2d(X)
    if Y is None:
        Y = X
    elif eval_gradient:
        raise ValueError("Gradient can only be evaluated when Y is None.")

    K = np.full(
        (X.shape[0], Y.shape[0]),
        self.constant_value,
        dtype=np.array(self.constant_value).dtype,
    )
    if eval_gradient:
        if not self.hyperparameter_constant_value.fixed:
            return (
                K,
                np.full(
                    (X.shape[0], X.shape[0], 1),
                    self.constant_value,
                    dtype=np.array(self.constant_value).dtype,
                ),
            )
        else:
            return K, np.empty((X.shape[0], X.shape[0], 0))
    else:
        return K

get_params #

get_params(deep: bool = True) -> dict[str, Any]

获取此核函数的参数。

参数#

deep : bool, 默认为 True 如果为 True，将返回此估计器及其包含的子对象（也是估计器）的参数。

返回值#

params : dict[str, Any] 参数名映射到其值。

源代码位于 smac/model/gaussian_process/kernels/base_kernels.py 
 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
def get_params(self, deep: bool = True) -> dict[str, Any]:
    """Get parameters of this kernel.

    Parameters
    ----------
    deep : bool, defaults to True
        If True, will return the parameters for this estimator and
        contained subobjects that are estimators.

    Returns
    -------
    params : dict[str, Any]
        Parameter names mapped to their values.
    """
    params = {}

    # ignore[misc] looks like it catches all kinds of errors, but misc is actually a category from mypy:
    # https://mypy.readthedocs.io/en/latest/error_code_list.html#miscellaneous-checks-misc
    tmp = super().get_params(deep)  # type: ignore[misc] # noqa F821
    args = list(tmp.keys())

    # Sum and Product do not clone the 'has_conditions' attribute by default. Instead of changing their
    # get_params() method, we simply add the attribute here!
    if "has_conditions" not in args:
        args.append("has_conditions")

    for arg in args:
        params[arg] = getattr(self, arg, None)

    return params

ProductKernel #

ProductKernel(
    k1: Kernel,
    k2: Kernel,
    operate_on: ndarray | None = None,
    has_conditions: bool = False,
)

基类: AbstractKernel, Product

乘积核实现。

源代码位于 smac/model/gaussian_process/kernels/base_kernels.py 
331
332
333
334
335
336
337
338
339
340
341
342
343
def __init__(
    self,
    k1: kernels.Kernel,
    k2: kernels.Kernel,
    operate_on: np.ndarray | None = None,
    has_conditions: bool = False,
) -> None:
    super().__init__(
        operate_on=operate_on,
        has_conditions=has_conditions,
        k1=k1,
        k2=k2,
    )

hyperparameters property #

hyperparameters: list[Hyperparameter]

返回所有超参数规范的列表。

meta property #

meta: dict[str, Any]

返回创建对象的元数据。此方法调用 get_params 方法来收集核函数的参数。

n_dims property #

n_dims: int

返回核函数的非固定超参数数量。

__call__ #

__call__(
    X: ndarray,
    Y: ndarray | None = None,
    eval_gradient: bool = False,
    active: ndarray | None = None,
) -> ndarray | tuple[ndarray, ndarray]

返回核函数 k(X, Y) 以及可选的其梯度。

参数#

X : np.ndarray, 形状 (n_samples_X, n_features) 返回的核函数 k(X, Y) 的左参数。

np.ndarray, 形状 (n_samples_Y, n_features), (可选, 默认为 None)

返回的核函数 k(X, Y) 的右参数。如果为 None，则计算 k(X, X)。

bool (可选, 默认为 False)

确定是否计算相对于核函数超参数的梯度。

np.ndarray (n_samples_X, n_features) (可选)

布尔数组，指定哪些超参数是活动的。

返回值#

K : np.ndarray, 形状 (n_samples_X, n_samples_Y) 核函数 k(X, Y)。

np.ndarray (可选), 形状 (n_samples_X, n_samples_X, n_dims)

核函数 k(X, X) 相对于核函数超参数的梯度。仅当 eval_gradient 为 True 时返回。

源代码位于 smac/model/gaussian_process/kernels/base_kernels.py 
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
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
def __call__(
    self,
    X: np.ndarray,
    Y: np.ndarray | None = None,
    eval_gradient: bool = False,
    active: np.ndarray | None = None,
) -> np.ndarray | tuple[np.ndarray, np.ndarray]:
    """Return the kernel k(X, Y) and optionally its gradient.

    Parameters
    ----------
    X : np.ndarray, shape (n_samples_X, n_features)
        Left argument of the returned kernel k(X, Y).

    Y : np.ndarray, shape (n_samples_Y, n_features), (optional, default=None)
        Right argument of the returned kernel k(X, Y). If None, k(X, X)
        is evaluated instead.

    eval_gradient : bool (optional, default=False)
        Determines whether the gradient with respect to the kernel
        hyperparameter is determined.

    active : np.ndarray (n_samples_X, n_features) (optional)
        Boolean array specifying which hyperparameters are active.

    Returns
    -------
    K : np.ndarray, shape (n_samples_X, n_samples_Y)
        Kernel k(X, Y).

    K_gradient : np.ndarray (opt.), shape (n_samples_X, n_samples_X, n_dims)
        The gradient of the kernel k(X, X) with respect to the
        hyperparameter of the kernel. Only returned when eval_gradient
        is True.
    """
    if eval_gradient:
        K1, K1_gradient = self.k1(X, Y, eval_gradient=True, active=active)
        K2, K2_gradient = self.k2(X, Y, eval_gradient=True, active=active)

        return K1 * K2, np.dstack((K1_gradient * K2[:, :, np.newaxis], K2_gradient * K1[:, :, np.newaxis]))
    else:
        return self.k1(X, Y, active=active) * self.k2(X, Y, active=active)

get_params #

get_params(deep: bool = True) -> dict[str, Any]

获取此核函数的参数。

参数#

deep : bool, 默认为 True 如果为 True，将返回此估计器及其包含的子对象（也是估计器）的参数。

返回值#

params : dict[str, Any] 参数名映射到其值。

源代码位于 smac/model/gaussian_process/kernels/base_kernels.py 
 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
def get_params(self, deep: bool = True) -> dict[str, Any]:
    """Get parameters of this kernel.

    Parameters
    ----------
    deep : bool, defaults to True
        If True, will return the parameters for this estimator and
        contained subobjects that are estimators.

    Returns
    -------
    params : dict[str, Any]
        Parameter names mapped to their values.
    """
    params = {}

    # ignore[misc] looks like it catches all kinds of errors, but misc is actually a category from mypy:
    # https://mypy.readthedocs.io/en/latest/error_code_list.html#miscellaneous-checks-misc
    tmp = super().get_params(deep)  # type: ignore[misc] # noqa F821
    args = list(tmp.keys())

    # Sum and Product do not clone the 'has_conditions' attribute by default. Instead of changing their
    # get_params() method, we simply add the attribute here!
    if "has_conditions" not in args:
        args.append("has_conditions")

    for arg in args:
        params[arg] = getattr(self, arg, None)

    return params

SumKernel #

SumKernel(
    k1: Kernel,
    k2: Kernel,
    operate_on: ndarray | None = None,
    has_conditions: bool = False,
)

基类: AbstractKernel, Sum

求和核实现。

源代码位于 smac/model/gaussian_process/kernels/base_kernels.py 
270
271
272
273
274
275
276
277
278
279
280
281
282
def __init__(
    self,
    k1: kernels.Kernel,
    k2: kernels.Kernel,
    operate_on: np.ndarray | None = None,
    has_conditions: bool = False,
) -> None:
    super().__init__(
        operate_on=operate_on,
        has_conditions=has_conditions,
        k1=k1,
        k2=k2,
    )

hyperparameters property #

hyperparameters: list[Hyperparameter]

返回所有超参数规范的列表。

meta property #

meta: dict[str, Any]

返回创建对象的元数据。此方法调用 get_params 方法来收集核函数的参数。

n_dims property #

n_dims: int

返回核函数的非固定超参数数量。

__call__ #

__call__(
    X: ndarray,
    Y: ndarray | None = None,
    eval_gradient: bool = False,
    active: ndarray | None = None,
) -> ndarray | tuple[ndarray, ndarray]

返回核函数 k(X, Y) 以及可选的其梯度。

参数#

X : np.ndarray, 形状 (n_samples_X, n_features) 返回的核函数 k(X, Y) 的左参数。

np.ndarray, 形状 (n_samples_Y, n_features), (可选, 默认为 None)

返回的核函数 k(X, Y) 的右参数。如果为 None，则计算 k(X, X)。

bool (可选, 默认为 False)

确定是否计算相对于核函数超参数的梯度。

np.ndarray (n_samples_X, n_features) (可选)

布尔数组，指定哪些超参数是活动的。

返回值#

K : np.ndarray, 形状 (n_samples_X, n_samples_Y) 核函数 k(X, Y)。

np.ndarray (可选), 形状 (n_samples_X, n_samples_X, n_dims)

核函数 k(X, X) 相对于核函数超参数的梯度。仅当 eval_gradient 为 True 时返回。

源代码位于 smac/model/gaussian_process/kernels/base_kernels.py 
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
def __call__(
    self,
    X: np.ndarray,
    Y: np.ndarray | None = None,
    eval_gradient: bool = False,
    active: np.ndarray | None = None,
) -> np.ndarray | tuple[np.ndarray, np.ndarray]:
    """Return the kernel k(X, Y) and optionally its gradient.

    Parameters
    ----------
    X : np.ndarray, shape (n_samples_X, n_features)
        Left argument of the returned kernel k(X, Y).

    Y : np.ndarray, shape (n_samples_Y, n_features), (optional, default=None)
        Right argument of the returned kernel k(X, Y). If None, k(X, X)
        is evaluated instead.

    eval_gradient : bool (optional, default=False)
        Determines whether the gradient with respect to the kernel
        hyperparameter is determined.

    active : np.ndarray (n_samples_X, n_features) (optional)
        Boolean array specifying which hyperparameters are active.

    Returns
    -------
    K : np.ndarray, shape (n_samples_X, n_samples_Y)
        Kernel k(X, Y).

    K_gradient : np.ndarray (opt.), shape (n_samples_X, n_samples_X, n_dims)
        The gradient of the kernel k(X, X) with respect to the
        hyperparameter of the kernel. Only returned when eval_gradient
        is True.
    """
    if eval_gradient:
        K1, K1_gradient = self.k1(X, Y, eval_gradient=True, active=active)
        K2, K2_gradient = self.k2(X, Y, eval_gradient=True, active=active)

        return K1 + K2, np.dstack((K1_gradient, K2_gradient))
    else:
        return self.k1(X, Y, active=active) + self.k2(X, Y, active=active)

get_params #

get_params(deep: bool = True) -> dict[str, Any]

获取此核函数的参数。

参数#

deep : bool, 默认为 True 如果为 True，将返回此估计器及其包含的子对象（也是估计器）的参数。

返回值#

params : dict[str, Any] 参数名映射到其值。

源代码位于 smac/model/gaussian_process/kernels/base_kernels.py 
 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
def get_params(self, deep: bool = True) -> dict[str, Any]:
    """Get parameters of this kernel.

    Parameters
    ----------
    deep : bool, defaults to True
        If True, will return the parameters for this estimator and
        contained subobjects that are estimators.

    Returns
    -------
    params : dict[str, Any]
        Parameter names mapped to their values.
    """
    params = {}

    # ignore[misc] looks like it catches all kinds of errors, but misc is actually a category from mypy:
    # https://mypy.readthedocs.io/en/latest/error_code_list.html#miscellaneous-checks-misc
    tmp = super().get_params(deep)  # type: ignore[misc] # noqa F821
    args = list(tmp.keys())

    # Sum and Product do not clone the 'has_conditions' attribute by default. Instead of changing their
    # get_params() method, we simply add the attribute here!
    if "has_conditions" not in args:
        args.append("has_conditions")

    for arg in args:
        params[arg] = getattr(self, arg, None)

    return params