tensor_conversion

deque2rolling_tensor(window, device='cpu', dtype=torch.float32)

Convert a dictionary to a rolling tensor.

PARAMETER	DESCRIPTION
window	Rolling window. TYPE: Deque
device	Device. DEFAULT: 'cpu'
dtype	Dtype. DEFAULT: float32

RETURNS	DESCRIPTION
torch.Tensor

df2tensor(X, features, default_value=0, device='cpu', dtype=torch.float32)

Convert a dataframe to a tensor.

PARAMETER	DESCRIPTION
X	Dataframe. TYPE: DataFrame
features	Set of possible features. TYPE: OrderedSet
default_value	Value to use for features not present in x. TYPE: float DEFAULT: 0
device	Device. DEFAULT: 'cpu'
dtype	Dtype. DEFAULT: float32

RETURNS	DESCRIPTION
torch.Tensor

dict2tensor(x, features, default_value=0, device='cpu', dtype=torch.float32)

Convert a dictionary to a tensor.

PARAMETER	DESCRIPTION
x	Dictionary. TYPE: dict
features	Set of possible features. TYPE: OrderedSet
default_value	Value to use for features not present in x. TYPE: float DEFAULT: 0
device	Device. TYPE: str DEFAULT: 'cpu'
dtype	Dtype. TYPE: dtype DEFAULT: float32

RETURNS	DESCRIPTION
torch.Tensor

float2tensor(y, device='cpu', dtype=torch.float32)

Convert a float to a tensor.

PARAMETER	DESCRIPTION
y	Float. TYPE: Union[float, int, RegTarget, dict]
device	Device. DEFAULT: 'cpu'
dtype	Dtype. DEFAULT: float32

RETURNS	DESCRIPTION
torch.Tensor

labels2onehot(y, classes, n_classes=None, device='cpu', dtype=torch.float32)

Convert a label or a list of labels to a one-hot encoded tensor.

PARAMETER	DESCRIPTION
y	Label or list of labels. TYPE: Union[ClfTarget, Series]
classes	Classes. TYPE: OrderedSet[ClfTarget]
n_classes	Number of classes. TYPE: Optional[int] DEFAULT: None
device	Device. DEFAULT: 'cpu'
dtype	Dtype. DEFAULT: float32

RETURNS	DESCRIPTION
torch.Tensor