fix error-raising docstring

Author: GiovanniCanali

pina/_src/callback/optim/switch_scheduler.py

@@ -22,8 +22,8 @@ def __init__(self, new_schedulers, epoch_switch):
         :type new_schedulers: pina.optim.TorchScheduler |
             list[pina.optim.TorchScheduler]
         :param int epoch_switch: The epoch at which the scheduler switch occurs.
-        :raise AssertionError: If epoch_switch is less than 1.
-        :raise ValueError: If each scheduler in ``new_schedulers`` is not an
+        :raises AssertionError: If epoch_switch is less than 1.
+        :raises ValueError: If each scheduler in ``new_schedulers`` is not an
             instance of :class:`pina.optim.TorchScheduler`.
 
         Example:

pina/_src/model/spline.py

@@ -202,7 +202,7 @@ def basis(self, x, collection=False):
         :param torch.Tensor x: The points to be evaluated.
         :param bool collection: If True, returns a list of basis functions for
             all orders up to the spline order. Default is False.
-        :raise ValueError: If ``collection`` is not a boolean.
+        :raises ValueError: If ``collection`` is not a boolean.
         :return: The basis functions evaluated at x.
         :rtype: torch.Tensor | list[torch.Tensor]
         """
@@ -290,7 +290,7 @@ def derivative(self, x, degree):
         :param x: The input tensor.
         :type x: torch.Tensor | LabelTensor
         :param int degree: The derivative degree to compute.
-        :raise ValueError: If ``degree`` is not an integer.
+        :raises ValueError: If ``degree`` is not an integer.
         :return: The derivative tensor.
         :rtype: torch.Tensor
         """

pina/_src/model/spline_surface.py

@@ -134,8 +134,8 @@ def derivative(self, x, degree_u, degree_v):
             parameter direction.
         :param int degree_v: The degree of the derivative along the second
             parameter direction.
-        :raise ValueError: If ``degree_u`` is not an integer.
-        :raise ValueError: If ``degree_v`` is not an integer.
+        :raises ValueError: If ``degree_u`` is not an integer.
+        :raises ValueError: If ``degree_v`` is not an integer.
         :return: The derivative tensor.
         :rtype: torch.Tensor
         """

pina/_src/model/vectorized_spline.py

@@ -293,7 +293,7 @@ def basis(self, x, collection=False):
         :param torch.Tensor x: The points to be evaluated.
         :param bool collection: If True, returns a list of basis functions for
             all orders up to the spline order. Default is False.
-        :raise ValueError: If ``collection`` is not a boolean.
+        :raises ValueError: If ``collection`` is not a boolean.
         :raises ValueError: If ``x`` is not two-dimensional.
         :raises ValueError: If the number of input features does not match
             the number of univariate splines.

pina/_src/solver/autoregressive_solver/autoregressive_solver.py

@@ -66,8 +66,9 @@ def __init__(
             averages used for adaptive weighting are reset at the start of each
             epoch. Setting this parameter to ``False`` can improve training
             stability, especially when data are scarce. Default is ``True``.
-        :raise ValueError: If the provided loss function is not compatible.
-        :raise ValueError: If ``reset_weights_at_epoch_start`` is not a boolean.
+        :raises ValueError: If the provided loss function is not compatible.
+        :raises ValueError: If ``reset_weights_at_epoch_start`` is not a
+            boolean.
         """
         super().__init__(
             problem=problem,
@@ -147,7 +148,7 @@ def loss_autoregressive(
         :param input: The input tensor containing unroll windows.
         :type input: torch.Tensor | LabelTensor
         :param dict kwargs: Additional keyword arguments for loss computation.
-        :raise ValueError: If ``input`` has less than 4 dimensions.
+        :raises ValueError: If ``input`` has less than 4 dimensions.
         :return: The scalar loss value for the given batch.
         :rtype: torch.Tensor | LabelTensor
         """
@@ -270,7 +271,7 @@ def predict(self, initial_state, n_steps, **kwargs):
         :type initial_state: torch.Tensor | LabelTensor
         :param int n_steps: The number of autoregressive steps to predict.
         :param dict kwargs: Additional keyword arguments.
-        :raise ValueError: If the provided initial_state tensor has less than 3
+        :raises ValueError: If the provided initial_state tensor has less than 3
             dimensions.
         :return: The predicted trajectory, including the initial state. It has
             shape ``[trajectories, n_steps + 1, *features]``, where the first
@@ -320,8 +321,8 @@ def unroll(data, unroll_length, n_unrolls=None, randomize=True):
             If ``None``, all valid windows are returned. Default is ``None``.
         :param bool randomize: If ``True``, starting indices are randomly
             permuted before applying ``n_unrolls``. Default is ``True``.
-        :raise ValueError: If the input ``data`` has less than 3 dimensions.
-        :raise ValueError: If ``unroll_length`` is greater or equal to the
+        :raises ValueError: If the input ``data`` has less than 3 dimensions.
+        :raises ValueError: If ``unroll_length`` is greater or equal to the
             number of time steps in ``data``.
         :return: A tensor of unrolled windows.
         :rtype: torch.Tensor | LabelTensor
@@ -358,7 +359,7 @@ def _get_start_idx(n_steps, unroll_length, n_unrolls=None, randomize=True):
             If ``None``, all valid windows are returned. Default is ``None``.
         :param bool randomize: If ``True``, starting indices are randomly
             permuted before applying ``n_unrolls``. Default is ``True``.
-        :raise ValueError: If ``unroll_length`` is greater or equal to the
+        :raises ValueError: If ``unroll_length`` is greater or equal to the
             number of time steps in ``data``.
         :return: A tensor of starting indices for unroll windows.
         :rtype: torch.Tensor