SNN2.src.actions.dataFrame module

SNN2.src.actions.dataFrame.MNIST_pack(mnist: List[ndarray]) → Tuple[Tuple[Any, ...], ...]

SNN2.src.actions.dataFrame.MNIST_unpack(mnist: Tuple[Tuple[ndarray, ...], ...]) → Tuple[ndarray, ...]

SNN2.src.actions.dataFrame.add_window_column(*args, df: DataFrame | None = None, window_size: int = 120, time_col: str = 'timestamp', groupby_col: str | None = 'op_id') → DataFrame

SNN2.src.actions.dataFrame.compute_dayofweek(*args, df: DataFrame | None = None, in_col: str = None, out_col: str = None) → DataFrame

SNN2.src.actions.dataFrame.compute_lag_features(*args, df: DataFrame | None = None, lag_col: List[str] = None, window_size: int | None = 120, time_col: str | None = 'timestamp', groupby_col: str | None = 'op_id') → DataFrame

SNN2.src.actions.dataFrame.compute_roll_features(*args, df: DataFrame | None = None, operations: Dict[str, List[str]] | None = None, window_size: int | None = 120, time_col: str | None = 'timestamp', groupby_col: str | None = 'op_id') → DataFrame

SNN2.src.actions.dataFrame.df_none(df: Any) → None

SNN2.src.actions.dataFrame.dropColumns(*args, df: DataFrame = None, axis=1, **kwargs) → DataFrame | None

SNN2.src.actions.dataFrame.dropNaN(*args, df: DataFrame = None, **kwargs) → DataFrame | None

SNN2.src.actions.dataFrame.dropOutliers(*args, df: DataFrame = None, threshold: float = 90.0, **kwargs) → DataFrame | None

SNN2.src.actions.dataFrame.drop_anomalous(*args, df: DataFrame | None = None, drop: Dict[int, List[Tuple[str, str]]] = None) → DataFrame

SNN2.src.actions.dataFrame.drop_not_anomalous(*args, df: DataFrame | None = None, keep: List[Dict[str, Any]] = None, delta: int = 0, rewrite_op_id: bool = False) → DataFrame

SNN2.src.actions.dataFrame.generateExpectationMatrix(labels) → ndarray

SNN2.src.actions.dataFrame.generateExpectationMatrixMNIST(*args, df: Tuple[Tuple[ndarray, ...], ...] = None, **kwargs) → Tuple[Tuple[ndarray, ...], ...]

SNN2.src.actions.dataFrame.generateImgTriplets(images, labels) → ndarray

SNN2.src.actions.dataFrame.generateImgUnrelatedGroups(images, labels) → ndarray

SNN2.src.actions.dataFrame.generateTripletsMNIST(*args, df: Tuple[Tuple[ndarray, ...], ...] = None, **kwargs) → Tuple[Tuple[ndarray, ...], ...]

SNN2.src.actions.dataFrame.generateUnrelatedGroupsMNIST(*args, df: Tuple[Tuple[ndarray, ...], ...] = None, **kwargs) → Tuple[Tuple[ndarray, ...], ...]

SNN2.src.actions.dataFrame.group_standardize(*args, df: DataFrame | None = None, standardize_columns: List[str] = ['n_ul'], groupby_col: str | None = 'op_id') → DataFrame

SNN2.src.actions.dataFrame.keepColumns(df: DataFrame = None, columns: List[str] = []) → DataFrame | None

SNN2.src.actions.dataFrame.load(*args, logger=None, write_msg=<function f_logger.<locals>.__dummy_log>, **kwargs) → DataFrame

SNN2.src.actions.dataFrame.loadMNIST(*args, **kwargs) → Tuple[Tuple[ndarray, ...], ...]

SNN2.src.actions.dataFrame.normalizeMNIST(*args, df: Tuple[Tuple[ndarray, ...], ...] = None, **kwargs) → Tuple[Tuple[ndarray, ...], ...]

SNN2.src.actions.dataFrame.removeNaN(*args, df: DataFrame = None, **kwargs) → DataFrame | None

SNN2.src.actions.dataFrame.remove_over_threshold(*args, df: DataFrame | None = None, column: str | None = None, threshold: float | None = None, **kwargs) → DataFrame

SNN2.src.actions.dataFrame.reshapeMNIST(*args, df: Tuple[Tuple[ndarray, ...], ...] = None, **kwargs) → Tuple[Tuple[ndarray, ...], ...]

SNN2.src.actions.dataFrame.rolling_feature(df: DataFrame, feature: str, frm_col: list, *args, function: Callable | None = None, group_id: str = 'id', **kwargs) → DataFrame

Compute rolling feature for each id in the DataFrame

Parameters:

• df (pd.DataFrame) – The DataFrame to compute the feature on

• feature (str) – The name of the feature to compute

• frm_col (list) – The list of columns to compute the feature on

• *args (list) – Additional arguments to pass to the rolling function

• **kwargs (dict) – Additional keyword arguments to pass to the rolling function

Returns:

The DataFrame with the new feature columns added

Return type:

pd.DataFrame

SNN2.src.actions.dataFrame.toTfDatasetsMNIST(*args, df: Tuple[Tuple[ndarray, ...], ...] = None, **kwargs) → Tuple[Tuple[DatasetV2, Tensor], ...]

SNN2.src.actions.dataFrame.trn_val_test_splitMNIST(*args, df: Tuple[Tuple[ndarray, ...], ...] = None, **kwargs) → Tuple[Tuple[ndarray, ...], ...]

SNN2.src.actions.dataFrame.write(*args, df: DataFrame = None, **kwargs) → None