disdrodb 0.0.20__py3-none-any.whl → 0.1.0__py3-none-any.whl

disdrodb/l0/l0b_nc_processing.py

@@ -18,12 +18,11 @@
 # -----------------------------------------------------------------------------.
 """Functions to process DISDRODB raw netCDF files into DISDRODB L0B netCDF files."""
 
-import copy
 import logging
 
 import numpy as np
-import xarray as xr
 
+from disdrodb.api.checks import check_sensor_name
 from disdrodb.l0.l0b_processing import finalize_dataset
 from disdrodb.l0.standards import (
     get_bin_coords_dict,
@@ -36,7 +35,6 @@ from disdrodb.l0.standards import (
 from disdrodb.utils.logger import (
     # log_warning,
     # log_debug,
-    log_error,
     log_info,
 )
 
@@ -46,7 +44,7 @@ logger = logging.getLogger(__name__)
 
 
 def _check_dict_names_validity(dict_names, sensor_name):
-    """Check dict_names dictionary values validity."""
+    """Check ``dict_names`` dictionary values validity."""
     valid_names = get_valid_names(sensor_name)
     keys = np.array(list(dict_names.keys()))
     values = np.array(list(dict_names.values()))
@@ -56,13 +54,11 @@ def _check_dict_names_validity(dict_names, sensor_name):
         # Report invalid keys and raise error
         invalid_dict = {k: dict_names[k] for k in invalid_keys}
         msg = f"The following dict_names values are not valid: {invalid_dict}"
-        log_error(logger=logger, msg=msg, verbose=False)
         raise ValueError(msg)
-    return None
 
 
 def _get_dict_names_variables(dict_names, sensor_name):
-    """Get DISDRODB variables specified in dict_names."""
+    """Get DISDRODB variables specified in ``dict_names``."""
     possible_variables = get_valid_variable_names(sensor_name)
     dictionary_names = list(dict_names.values())
     variables = [name for name in dictionary_names if name in possible_variables]
@@ -78,7 +74,7 @@ def _get_missing_variables(ds, dict_names, sensor_name):
 
 
 def rename_dataset(ds, dict_names):
-    """Rename Dataset variables, coordinates and dimensions."""
+    """Rename xr.Dataset variables, coordinates and dimensions."""
     # Get dataset variables, coordinates and dimensions of the dataset
     ds_vars = list(ds.data_vars)
     ds_dims = list(ds.dims)
@@ -101,24 +97,23 @@ def rename_dataset(ds, dict_names):
 
 
 def subset_dataset(ds, dict_names, sensor_name):
-    """Subset Dataset with expected variables."""
+    """Subset xr.Dataset with expected variables."""
     # Get valid variable names
     possible_variables = get_valid_variable_names(sensor_name)
     # Get variables availables in the dict_names and dataset
     dataset_variables = list(ds.data_vars)
     dictionary_names = list(dict_names.values())
     # Get subset variables
-    subset_variables = []
-    for var in dataset_variables:
-        if var in dictionary_names and var in possible_variables:
-            subset_variables.append(var)
+    subset_variables = [var for var in dataset_variables if var in dictionary_names and var in possible_variables]
     # Subset the dataset
     ds = ds[subset_variables]
     return ds
 
 
 def add_dataset_missing_variables(ds, missing_vars, sensor_name):
-    """Add missing Dataset variables as nan DataArrays."""
+    """Add missing xr.Dataset variables as ``np.nan`` xr.DataArrays."""
+    import xarray as xr
+
     from disdrodb.l0.standards import get_variables_dimension
 
     # Get dimension of each variables
@@ -128,7 +123,7 @@ def add_dataset_missing_variables(ds, missing_vars, sensor_name):
         # Get variable dimension
         dims = var_dims_dict[var]
         # Retrieve expected shape
-        expected_shape = [ds.dims[dim] for dim in dims]
+        expected_shape = [ds.sizes[dim] for dim in dims]
         # Create DataArray
         arr = np.zeros(expected_shape) * np.nan
         da = xr.DataArray(arr, dims=dims)
@@ -137,15 +132,15 @@ def add_dataset_missing_variables(ds, missing_vars, sensor_name):
     return ds
 
 
-def preprocess_raw_netcdf(ds, dict_names, sensor_name):
+def standardize_raw_dataset(ds, dict_names, sensor_name):
     """This function preprocess raw netCDF to improve compatibility with DISDRODB standards.
 
-    This function checks validity of the dict_names, rename and subset the data accordingly.
-    If some variables specified in the dict_names are missing, it adds a NaN DataArray !
+    This function checks validity of the ``dict_names``, rename and subset the data accordingly.
+    If some variables specified in the ``dict_names`` are missing, it adds a ``np.nan`` xr.DataArray !
 
     Parameters
     ----------
-    ds : xr.Dataset
+    ds  : xarray.Dataset
         Raw netCDF to be converted to DISDRODB standards.
     dict_names : dict
         Dictionary mapping raw netCDF variables/coordinates/dimension names
@@ -155,10 +150,13 @@ def preprocess_raw_netcdf(ds, dict_names, sensor_name):
 
     Returns
     -------
-    ds : xr.Dataset
-        xarray Dataset with DISDRODB-compliant variable naming conventions.
+    ds  : xarray.Dataset
+        xarray Dataset with variables compliant with DISDRODB conventions.
 
     """
+    # Check if the sensor name is valid
+    check_sensor_name(sensor_name)
+
     # Check variable_dict has valid values
     # - Check valid DISDRODB variables + dimensions + coords
     _check_dict_names_validity(dict_names=dict_names, sensor_name=sensor_name)
@@ -175,32 +173,35 @@ def preprocess_raw_netcdf(ds, dict_names, sensor_name):
         ds = add_dataset_missing_variables(ds=ds, missing_vars=missing_vars, sensor_name=sensor_name)
 
     # Update the coordinates for (diameter and velocity)
-    coords = get_bin_coords_dict(sensor_name)
-    ds = ds.assign_coords(coords)
+    ds = ds.assign_coords(get_bin_coords_dict(sensor_name))
 
     # Return dataset
     return ds
 
 
-def replace_custom_nan_flags(ds, dict_nan_flags, verbose=False):
-    """Set values corresponding to nan_flags to np.nan.
+def replace_custom_nan_flags(ds, dict_nan_flags, logger=None, verbose=False):
+    """Set values corresponding to ``nan_flags`` to ``np.nan``.
 
     This function must be used in a reader, if necessary.
 
     Parameters
     ----------
-    df : xr.Dataset
+    df  : xarray.Dataset
         Input xarray dataset
     dict_nan_flags : dict
-        Dictionary with nan flags value to set as np.nan
+        Dictionary with nan flags value to set as ``np.nan``.
+    verbose : bool
+        Whether to verbose the processing. The default value is ``False``.
 
     Returns
     -------
-    xr.Dataset
-        Dataset without nan_flags values.
+    xarray.Dataset
+        Dataset without ``nan_flags`` values.
     """
     # Loop over the needed variable, and replace nan_flags values with np.nan
     for var, nan_flags in dict_nan_flags.items():
+        # Ensure nan_flags is a list
+        nan_flags = [nan_flags] if not isinstance(nan_flags, list) else nan_flags
         # If the variable is in the dataframe
         if var in ds:
             # Get occurrence of nan_flags
@@ -215,12 +216,12 @@ def replace_custom_nan_flags(ds, dict_nan_flags, verbose=False):
     return ds
 
 
-def replace_nan_flags(ds, sensor_name, verbose):
-    """Set values corresponding to nan_flags to np.nan.
+def replace_nan_flags(ds, sensor_name, verbose, logger=None):
+    """Set values corresponding to ``nan_flags`` to ``np.nan``.
 
     Parameters
     ----------
-    ds : xr.Dataset
+    ds  : xarray.Dataset
         Input xarray dataset
     dict_nan_flags : dict
         Dictionary with nan flags value to set as np.nan
@@ -229,22 +230,22 @@ def replace_nan_flags(ds, sensor_name, verbose):
 
     Returns
     -------
-    xr.Dataset
-        Dataset without nan_flags values.
+    xarray.Dataset
+        Dataset without ``nan_flags`` values.
     """
     # Get dictionary of nan flags
     dict_nan_flags = get_nan_flags_dict(sensor_name)
     # Replace nan flags with nan
-    ds = replace_custom_nan_flags(ds, dict_nan_flags, verbose=verbose)
+    ds = replace_custom_nan_flags(ds, dict_nan_flags=dict_nan_flags, logger=logger, verbose=verbose)
     return ds
 
 
-def set_nan_outside_data_range(ds, sensor_name, verbose):
-    """Set values outside the data range as np.nan.
+def set_nan_outside_data_range(ds, sensor_name, verbose, logger=None):
+    """Set values outside the data range as ``np.nan``.
 
     Parameters
     ----------
-    ds : xr.Dataset
+    ds  : xarray.Dataset
         Input xarray dataset
     sensor_name : str
         Name of the sensor.
@@ -253,7 +254,7 @@ def set_nan_outside_data_range(ds, sensor_name, verbose):
 
     Returns
     -------
-    xr.Dataset
+    xarray.Dataset
         Dataset without values outside the expected data range.
     """
     # Get dictionary of data_range
@@ -277,12 +278,12 @@ def set_nan_outside_data_range(ds, sensor_name, verbose):
     return ds
 
 
-def set_nan_invalid_values(ds, sensor_name, verbose):
-    """Set invalid (class) values to np.nan.
+def set_nan_invalid_values(ds, sensor_name, verbose, logger=None):
+    """Set invalid (class) values to ``np.nan``.
 
     Parameters
     ----------
-    ds : xr.Dataset
+    ds  : xarray.Dataset
         Input xarray dataset
     sensor_name : str
         Name of the sensor.
@@ -291,7 +292,7 @@ def set_nan_invalid_values(ds, sensor_name, verbose):
 
     Returns
     -------
-    xr.Dataset
+    xarray.Dataset
         Dataset without invalid values.
     """
     # Get dictionary of valid values
@@ -313,25 +314,146 @@ def set_nan_invalid_values(ds, sensor_name, verbose):
     return ds
 
 
-def create_l0b_from_raw_nc(
+def drop_timesteps(ds, timesteps: list):
+    """
+    Drop specific time steps from a Dataset.
+
+    Parameters
+    ----------
+    ds : xarray.Dataset
+        Input dataset with a 'time' dimension.
+    timesteps : list
+        List of datetime-like values to remove.
+
+    Returns
+    -------
+    xarray.Dataset
+        Dataset with specified timesteps removed.
+
+    Raises
+    ------
+    ValueError
+        If no timesteps remain after removal.
+    """
+    # Create a boolean mask of valid timesteps
+    times = ds["time"].to_numpy()
+    mask = ~np.isin(times, np.array(timesteps, dtype=times.dtype))
+    ds_filtered = ds.isel(time=mask)
+
+    # Ensure there's at least one timestep left
+    if ds_filtered.sizes.get("time", 0) == 0:
+        raise ValueError(
+            "No timesteps left after removing problematic timesteps. " "Maybe you need to adjust the issue YAML file.",
+        )
+    return ds_filtered
+
+
+def drop_time_periods(ds, time_periods: list):
+    """
+    Drop all time steps within any of the specified time intervals.
+
+    Parameters
+    ----------
+    ds : xarray.Dataset
+        Input dataset with a 'time' dimension.
+    time_periods : list of tuple
+        Each tuple is (start_time, end_time), datetime-like, inclusive.
+
+    Returns
+    -------
+    xarray.Dataset
+        Dataset with all times within the given periods removed.
+
+    Raises
+    ------
+    ValueError
+        If no timesteps remain after removal.
+    """
+    times = ds["time"].to_numpy()
+    mask = np.ones_like(times, dtype=bool)
+
+    for start, end in time_periods:
+        start_np = np.datetime64(start)
+        end_np = np.datetime64(end)
+        # exclude times in the inclusive interval [start, end]
+        mask &= ~((times >= start_np) & (times <= end_np))
+
+    ds_filtered = ds.isel(time=mask)
+
+    if ds_filtered.sizes.get("time", 0) == 0:
+        raise ValueError(
+            "No timesteps left after removing problematic time_periods. "
+            "Maybe you need to adjust the issue YAML file.",
+        )
+    return ds_filtered
+
+
+def remove_issue_timesteps(
+    ds,
+    issue_dict: dict,
+    logger=None,
+    verbose: bool = False,
+):
+    """
+    Remove bad timesteps and time periods from an xarray Dataset according to issue definitions.
+
+    Parameters
+    ----------
+    ds : xarray.Dataset
+        Input dataset with a 'time' dimension.
+    issue_dict : dict
+        Dictionary with optional keys 'timesteps' (list of datetimes) and
+        'time_periods' (list of (start, end) tuples).
+    logger : any, optional
+        Logger instance to record dropped steps, by default None.
+    verbose : bool, optional
+        Whether to log informational messages, by default False.
+
+    Returns
+    -------
+    xarray.Dataset
+        Cleaned dataset.
+
+    Raises
+    ------
+    ValueError
+        If after removing specified timesteps/periods no data remains.
+    """
+    n_initial = ds.sizes.get("time", 0)
+    timesteps = issue_dict.get("timesteps", []) or []
+    time_periods = issue_dict.get("time_periods", []) or []
+
+    # Drop individual timesteps
+    if timesteps:
+        ds = drop_timesteps(ds, timesteps)
+
+    # Drop intervals of time
+    if time_periods:
+        ds = drop_time_periods(ds, time_periods)
+
+    # Report number dropped
+    n_remaining = ds.sizes.get("time", 0)
+    dropped = n_initial - n_remaining
+    if dropped > 0:
+        msg = f"{dropped} timesteps were dropped according to the issue YAML file content."
+        log_info(logger=logger, msg=msg, verbose=verbose)
+    return ds
+
+
+def sanitize_ds(
     ds,
-    dict_names,
-    ds_sanitizer_fun,
     sensor_name,
-    verbose,
-    attrs,
+    metadata,
+    issue_dict=None,
+    verbose=False,
+    logger=None,
 ):
-    """Convert a raw xr.Dataset into a DISDRODB L0B netCDF.
+    """Convert a raw ``xr.Dataset`` into a DISDRODB L0B netCDF.
 
     Parameters
     ----------
-    ds : xr.Dataset
-        xr.Dataset
-    dict_names : dict
-        Dictionary mapping raw netCDF variables/coordinates/dimension names
-        to DISDRODB standards.
-    ds_sanitizer_fun : function
-        Sanitizer function to do ad-hoc processing of the xr.Dataset.
+    ds : xarray.Dataset
+        Raw xarray dataset
     attrs: dict
         Global metadata to attach as global attributes to the xr.Dataset.
     sensor_name : str
@@ -342,39 +464,64 @@ def create_l0b_from_raw_nc(
 
     Returns
     -------
-    xr.Dataset
+    xarray.Dataset
         L0B xr.Dataset
     """
-    # Preprocess netcdf
-    ds = preprocess_raw_netcdf(ds=ds, dict_names=dict_names, sensor_name=sensor_name)
-
-    # Add CRS and geolocation information
-    attrs = copy.deepcopy(attrs)
-    coords = {}
-    geolocation_vars = ["latitude", "longitude", "altitude"]
-    for var in geolocation_vars:
-        if var not in ds:
-            coords[var] = attrs[var]
-        _ = attrs.pop(var)
-    ds = ds.assign_coords(coords)
-
-    # Add global attributes
-    ds.attrs = attrs
-
-    # Apply dataset sanitizer function
-    ds = ds_sanitizer_fun(ds)
-
     # Replace nan flags values with np.nans
-    ds = replace_nan_flags(ds, sensor_name=sensor_name, verbose=verbose)
+    ds = replace_nan_flags(ds, sensor_name=sensor_name, logger=logger, verbose=verbose)
+
+    # Filter out problematic tiemsteps reported in the issue YAML file
+    ds = remove_issue_timesteps(ds, issue_dict=issue_dict, logger=logger, verbose=verbose)
 
     # Set values outside the data range to np.nan
-    ds = set_nan_outside_data_range(ds, sensor_name=sensor_name, verbose=verbose)
+    ds = set_nan_outside_data_range(ds, sensor_name=sensor_name, logger=logger, verbose=verbose)
 
     # Replace invalid values with np.nan
-    ds = set_nan_invalid_values(ds, sensor_name=sensor_name, verbose=verbose)
+    ds = set_nan_invalid_values(ds, sensor_name=sensor_name, logger=logger, verbose=verbose)
 
     # Finalize dataset
-    ds = finalize_dataset(ds, sensor_name=sensor_name)
+    ds = finalize_dataset(ds, sensor_name=sensor_name, attrs=metadata)
 
     # Return dataset
     return ds
+
+
+def open_raw_netcdf_file(
+    filepath,
+    logger=None,
+    engine="netcdf4",
+    cache=False,
+    chunks=None,
+    decode_timedelta=False,
+    **kwargs,
+):
+    """Open a raw netCDF file.
+
+    Parameters
+    ----------
+    filepath : str
+        Path to the raw netCDF file.
+
+    Returns
+    -------
+    xarray.Dataset
+        Raw netCDF file as an xarray Dataset.
+    """
+    import xarray as xr
+
+    # Note: chunks=None avoid usage of Dask
+
+    # Open the raw netCDF
+    with xr.open_dataset(
+        filepath,
+        decode_timedelta=decode_timedelta,
+        cache=cache,
+        engine=engine,
+        chunks=chunks,
+        **kwargs,
+    ) as data:
+        ds = data.load()
+
+    # Log information
+    log_info(logger=logger, msg=f"netCDF file {filepath} has been loaded successively into xarray.", verbose=False)
+    return ds