refactor: rename some internal functions, and rename module.py to `…

…modules.py`;

docs/pypots.cli.rst

@@ -0,0 +1,56 @@
+pypots.cli package
+==================
+
+pypots.cli.base module
+----------------------
+
+.. automodule:: pypots.cli.base
+ :members:
+ :undoc-members:
+ :show-inheritance:
+ :inherited-members:
+
+pypots.cli.dev module
+---------------------
+
+.. automodule:: pypots.cli.dev
+ :members:
+ :undoc-members:
+ :show-inheritance:
+ :inherited-members:
+
+pypots.cli.doc module
+---------------------
+
+.. automodule:: pypots.cli.doc
+ :members:
+ :undoc-members:
+ :show-inheritance:
+ :inherited-members:
+
+pypots.cli.env module
+---------------------
+
+.. automodule:: pypots.cli.env
+ :members:
+ :undoc-members:
+ :show-inheritance:
+ :inherited-members:
+
+pypots.cli.pypots\_cli module
+-----------------------------
+
+.. automodule:: pypots.cli.pypots_cli
+ :members:
+ :undoc-members:
+ :show-inheritance:
+ :inherited-members:
+
+Module contents
+---------------
+
+.. automodule:: pypots.cli
+ :members:
+ :undoc-members:
+ :show-inheritance:
+ :inherited-members:

pypots/base.py

@@ -54,22 +54,16 @@ class BaseModel(ABC):
 
  """
 
- # leverage typing to show type hints in IDEs
- # SAVING_STRATEGY = Literal["best", "better"]
- SAVING_STRATEGY = [None, "best", "better"]
-
  def __init__(
  self,
  device: Optional[Union[str, torch.device]] = None,
  saving_path: str = None,
  model_saving_strategy: Optional[str] = "best",
  ):
-
- assert model_saving_strategy in [
- None,
- "best",
- "better",
- ], f"saving_strategy must be one of {self.SAVING_STRATEGY}, but got f{model_saving_strategy}."
+ saving_strategies = [None, "best", "better"]
+ assert (
+ model_saving_strategy in saving_strategies
+ ), f"saving_strategy must be one of {saving_strategies}, but got f{model_saving_strategy}."
 
  self.device = None
  self.saving_path = saving_path
@@ -121,7 +115,7 @@ def __init__(
  f"the tensorboard file will be saved to {tb_saving_path}"
  )
 
- def save_log_into_tb_file(self, step: int, stage: str, loss_dict: dict) -> None:
+ def _save_log_into_tb_file(self, step: int, stage: str, loss_dict: dict) -> None:
  """Saving training logs into the tensorboard file specified by the given path `tb_file_saving_path`.
 
  Parameters
@@ -190,7 +184,7 @@ def save_model(
  f'Failed to save the model to "{saving_path}" because of the below error! \n{e}'
  )
 
- def auto_save_model_if_necessary(
+ def _auto_save_model_if_necessary(
  self,
  training_finished: bool = True,
  saving_name: str = None,

pypots/classification/base.py

@@ -211,7 +211,7 @@ def _train_model(
 
  # save training loss logs into the tensorboard file for every step if in need
  if self.summary_writer is not None:
- self.save_log_into_tb_file(training_step, "training", results)
+ self._save_log_into_tb_file(training_step, "training", results)
 
  # mean training loss of the current epoch
  mean_train_loss = np.mean(epoch_train_loss_collector)
@@ -232,7 +232,7 @@ def _train_model(
  val_loss_dict = {
  "classification_loss": mean_val_loss,
  }
- self.save_log_into_tb_file(epoch, "validating", val_loss_dict)
+ self._save_log_into_tb_file(epoch, "validating", val_loss_dict)
 
  logger.info(
  f"epoch {epoch}: "
@@ -249,7 +249,7 @@ def _train_model(
  self.best_model_dict = self.model.state_dict()
  self.patience = self.original_patience
  # save the model if necessary
- self.auto_save_model_if_necessary(
+ self._auto_save_model_if_necessary(
  training_finished=False,
  saving_name=f"{self.__class__.__name__}_epoch{epoch}_loss{mean_loss}",
  )

pypots/classification/brits/model.py

@@ -14,32 +14,12 @@
 
 from pypots.classification.base import BaseNNClassifier
 from pypots.classification.brits.dataset import DatasetForBRITS
+from pypots.classification.brits.modules import RITS
 from pypots.imputation.brits.model import (
- RITS as imputation_RITS,
  _BRITS as imputation_BRITS,
 )
 
 
-class RITS(imputation_RITS):
- def __init__(
- self,
- n_steps: int,
- n_features: int,
- rnn_hidden_size: int,
- n_classes: int,
- device: Union[str, torch.device],
- ):
- super().__init__(n_steps, n_features, rnn_hidden_size, device)
- self.dropout = nn.Dropout(p=0.25)
- self.classifier = nn.Linear(self.rnn_hidden_size, n_classes)
-
- def forward(self, inputs: dict, direction: str = "forward") -> dict:
- ret_dict = super().forward(inputs, direction)
- logits = self.classifier(ret_dict["final_hidden_state"])
- ret_dict["prediction"] = torch.softmax(logits, dim=1)
- return ret_dict
-
-
 class _BRITS(imputation_BRITS, nn.Module):
  def __init__(
  self,
@@ -361,7 +341,7 @@ def fit(
  self.model.eval() # set the model as eval status to freeze it.
 
  # Step 3: save the model if necessary
- self.auto_save_model_if_necessary(training_finished=True)
+ self._auto_save_model_if_necessary(training_finished=True)
 
  def classify(self, X: Union[dict, str], file_type: str = "h5py"):
  """Classify the input data with the trained model.

pypots/classification/brits/modules.py

@@ -0,0 +1,35 @@
+"""
+PyTorch BRITS model for both the time-series imputation task and the classification task.
+"""
+
+# Created by Wenjie Du <[email protected]>
+# License: GPL-v3
+
+from typing import Union
+
+import torch
+import torch.nn as nn
+
+from pypots.imputation.brits.model import (
+ RITS as imputation_RITS,
+)
+
+
+class RITS(imputation_RITS):
+ def __init__(
+ self,
+ n_steps: int,
+ n_features: int,
+ rnn_hidden_size: int,
+ n_classes: int,
+ device: Union[str, torch.device],
+ ):
+ super().__init__(n_steps, n_features, rnn_hidden_size, device)
+ self.dropout = nn.Dropout(p=0.25)
+ self.classifier = nn.Linear(self.rnn_hidden_size, n_classes)
+
+ def forward(self, inputs: dict, direction: str = "forward") -> dict:
+ ret_dict = super().forward(inputs, direction)
+ logits = self.classifier(ret_dict["final_hidden_state"])
+ ret_dict["prediction"] = torch.softmax(logits, dim=1)
+ return ret_dict

pypots/classification/grud/model.py

@@ -16,7 +16,7 @@
 
 from pypots.classification.base import BaseNNClassifier
 from pypots.classification.grud.dataset import DatasetForGRUD
-from pypots.imputation.brits.module import TemporalDecay
+from pypots.imputation.brits.modules import TemporalDecay
 
 
 class _GRUD(nn.Module):
@@ -311,7 +311,7 @@ def fit(
  self.model.eval() # set the model as eval status to freeze it.
 
  # Step 3: save the model if necessary
- self.auto_save_model_if_necessary(training_finished=True)
+ self._auto_save_model_if_necessary(training_finished=True)
 
  def classify(self, X: Union[dict, str], file_type: str = "h5py") -> np.ndarray:
  """Classify the input data with the trained model.

pypots/classification/raindrop/model.py

@@ -25,7 +25,7 @@
 
 from pypots.classification.base import BaseNNClassifier
 from pypots.classification.grud.dataset import DatasetForGRUD
-from pypots.classification.raindrop.module import (
+from pypots.classification.raindrop.modules import (
  PositionalEncoding,
  ObservationPropagation,
 )
@@ -517,7 +517,7 @@ def fit(
  self.model.eval() # set the model as eval status to freeze it.
 
  # Step 3: save the model if necessary
- self.auto_save_model_if_necessary(training_finished=True)
+ self._auto_save_model_if_necessary(training_finished=True)
 
  def classify(self, X: Union[dict, str], file_type: str = "h5py") -> np.ndarray:
  """Classify the input data with the trained model.

pypots/clustering/crli/model.py

@@ -1,7 +1,7 @@
 """
 Torch implementation of CRLI (Clustering Representation Learning on Incomplete time-series data).
 
-Please refer to :cite:``ma2021CRLI``.
+Please refer to :cite:`ma2021CRLI`.
 """
 
 # Created by Wenjie Du <[email protected]>
@@ -18,7 +18,7 @@
 
 from pypots.clustering.base import BaseNNClusterer
 from pypots.clustering.crli.dataset import DatasetForCRLI
-from pypots.clustering.crli.module import Generator, Decoder, Discriminator
+from pypots.clustering.crli.modules import Generator, Decoder, Discriminator
 from pypots.utils.logging import logger
 from pypots.utils.metrics import cal_mse
 
@@ -292,7 +292,7 @@ def _train_model(
  "generation_loss": mean_step_train_G_loss,
  "discrimination_loss": mean_step_train_D_loss,
  }
- self.save_log_into_tb_file(
+ self._save_log_into_tb_file(
  training_step, "training", loss_results
  )
  mean_epoch_train_D_loss = np.mean(epoch_train_loss_D_collector)
@@ -309,7 +309,7 @@ def _train_model(
  self.best_model_dict = self.model.state_dict()
  self.patience = self.original_patience
  # save the model if necessary
- self.auto_save_model_if_necessary(
+ self._auto_save_model_if_necessary(
  training_finished=False,
  saving_name=f"{self.__class__.__name__}_epoch{epoch}_loss{mean_loss}",
  )
@@ -376,7 +376,7 @@ def fit(
  self.model.eval() # set the model as eval status to freeze it.
 
  # Step 3: save the model if necessary
- self.auto_save_model_if_necessary(training_finished=True)
+ self._auto_save_model_if_necessary(training_finished=True)
 
  def cluster(
  self,

pypots/clustering/crli/module.py → pypots/clustering/crli/modules.py

@@ -1,7 +1,7 @@
 """
 Torch implementation of CRLI (Clustering Representation Learning on Incomplete time-series data).
 
-Please refer to :cite:``ma2021CRLI``.
+Please refer to :cite:`ma2021CRLI` for details.
 """
 
 # Created by Wenjie Du <[email protected]>

pypots/clustering/vader/model.py

@@ -19,7 +19,7 @@
 from torch.utils.data import DataLoader
 
 from pypots.clustering.base import BaseNNClusterer
-from pypots.clustering.vader.module import (
+from pypots.clustering.vader.modules import (
  GMMLayer,
  PeepholeLSTMCell,
  ImplicitImputation,
@@ -412,7 +412,9 @@ def _train_model(
 
  # save pre-training loss logs into the tensorboard file for every step if in need
  if self.summary_writer is not None:
- self.save_log_into_tb_file(pretraining_step, "pretraining", results)
+ self._save_log_into_tb_file(
+ pretraining_step, "pretraining", results
+ )
 
  with torch.no_grad():
  sample_collector = []
@@ -485,7 +487,7 @@ def _train_model(
 
  # save training loss logs into the tensorboard file for every step if in need
  if self.summary_writer is not None:
- self.save_log_into_tb_file(training_step, "training", results)
+ self._save_log_into_tb_file(training_step, "training", results)
 
  # mean training loss of the current epoch
  mean_train_loss = np.mean(epoch_train_loss_collector)
@@ -506,7 +508,7 @@ def _train_model(
  val_loss_dict = {
  "loss": mean_val_loss,
  }
- self.save_log_into_tb_file(epoch, "validating", val_loss_dict)
+ self._save_log_into_tb_file(epoch, "validating", val_loss_dict)
 
  logger.info(
  f"epoch {epoch}: "
@@ -523,7 +525,7 @@ def _train_model(
  self.best_model_dict = self.model.state_dict()
  self.patience = self.original_patience
  # save the model if necessary
- self.auto_save_model_if_necessary(
+ self._auto_save_model_if_necessary(
  training_finished=False,
  saving_name=f"{self.__class__.__name__}_epoch{epoch}_loss{mean_loss}",
  )
@@ -594,7 +596,7 @@ def fit(
  self.model.eval() # set the model as eval status to freeze it.
 
  # Step 3: save the model if necessary
- self.auto_save_model_if_necessary(training_finished=True)
+ self._auto_save_model_if_necessary(training_finished=True)
 
  def cluster(self, X: Union[dict, str], file_type: str = "h5py") -> np.ndarray:
  """Cluster the input with the trained model.

pypots/data/base.py

@@ -72,7 +72,7 @@ def __init__(
  else: # data from array
  X = data["X"]
  y = None if "y" not in data.keys() else data["y"]
- self.X, self.y = self.check_input(X, y)
+ self.X, self.y = self._check_input(X, y)
 
  self.sample_num = self._get_sample_num()
 
@@ -103,7 +103,7 @@ def __len__(self) -> int:
  return self.sample_num
 
  @staticmethod
- def check_input(
+ def _check_input(
  X: Union[np.ndarray, torch.Tensor, list],
  y: Optional[Union[np.ndarray, torch.Tensor, list]] = None,
  out_dtype: str = "tensor",

pypots/forecasting/base.py

@@ -198,7 +198,7 @@ def _train_model(
 
  # save training loss logs into the tensorboard file for every step if in need
  if self.summary_writer is not None:
- self.save_log_into_tb_file(training_step, "training", results)
+ self._save_log_into_tb_file(training_step, "training", results)
 
  # mean training loss of the current epoch
  mean_train_loss = np.mean(epoch_train_loss_collector)
@@ -219,7 +219,7 @@ def _train_model(
  val_loss_dict = {
  "imputation_loss": mean_val_loss,
  }
- self.save_log_into_tb_file(epoch, "validating", val_loss_dict)
+ self._save_log_into_tb_file(epoch, "validating", val_loss_dict)
 
  logger.info(
  f"epoch {epoch}: "

pypots/forecasting/bttf/__init__.py

@@ -1,5 +1,5 @@
 """
-
+The package for BTTF model .
 """
 
 # Created by Wenjie Du <[email protected]>