pyfvcom2.cmems_reader module

Read CMEMS data

class pyfvcom2.cmems_reader.CMEMSReader(file_path: str | List[str], reference_var_name: str, dimension_var_names: dict | None = None)

Bases: object

Class to read CMEMS Data files

close()

Close all datasets to free up resources.

contains_date(date_time: datetime) → bool

Check if the dataset contains the given date_time.

Args:

date_time (datetime): Date time to check.

Returns:

bool: True if date_time is within the dataset time range, False otherwise.

property dates

property depth_levels

get_bottom_indices() → ndarray

Get indices of the deepest unmasked level for each horizontal point.

Returns:

np.ndarray: 2D array of bottom indices (lat, lon)

get_closest_date_index(date_time) → int

Get the index of the datetime variable closest to the given date_time.

Args:

date_time (datetime): Date time to find closest index for.

Returns:

int: Index of the closest time variable.

get_filled_2D_var(var_name: str, target_datetime: datetime, tolerance=None) → ndarray

Fill masked values in a 2D variable by interpolation.

Uses a Delaunay triangulation of unmasked ocean points to linearly interpolate over land-masked cells on the regular CMEMS grid.

Args:

var_name (str): Variable name. target_datetime (datetime): Target datetime to retrieve data for. tolerance (timedelta, optional): Maximum allowed time difference. Defaults to None.

Returns:

np.ndarray: Filled variable values on the full (lat, lon) grid.

get_filled_3D_var(var_name: str, target_datetime: datetime, tolerance=None) → ndarray

Fill masked values in a 3D variable by interpolation

First, use griddata to interpolate over all masked surface values. Then, for each horizontal point, fill masked depth levels by downward extrapolation from the nearest unmasked depth level.

Args:

var_name (str): Variable name. target_datetime (datetime): Target datetime to retrieve data for. tolerance (timedelta, optional): Maximum allowed time difference. Defaults to None.

Returns:

np.ndarray: Filled variable values.

get_filled_3D_var_by_index(var_name: str, time_index: int = 0) → ndarray

Backward compatibility method using time index

get_mask(var) → ndarray

Get the mask for a variable. Args:

var (xarray.DataArray): Variable to get the mask for.

Returns:

np.ndarray: Boolean mask array, where True indicates a masked value.

get_unmasked_var_by_index(var_name: str, time_index: int = 0, depth_index: int = None) → ndarray

Backward compatibility method using time index

get_unmasked_variable(var_name: str, target_datetime: datetime, depth_index: int = None, tolerance=None) → ndarray

Get the unmasked values of a variable at a given datetime and depth index.

Args:

var_name (str): Variable name. target_datetime (datetime): Target datetime to retrieve data for. depth_index (int, optional): Depth index for 3D variables. Defaults to None. tolerance (timedelta, optional): Maximum allowed time difference. Defaults to None.

Returns:

np.ndarray: Unmasked variable values.

get_var(var_name: str, target_datetime: datetime, depth_index: int = None, tolerance=None) → ndarray

Get the values of a variable at a given datetime and depth index.

Args:

var_name (str): Variable name. target_datetime (datetime): Target datetime to retrieve data for. depth_index (int, optional): Depth index for 3D variables. Defaults to None. tolerance (timedelta, optional): Maximum allowed time difference. Defaults to None.

Returns:

np.ndarray: Variable values.

get_var_by_index(var_name: str, time_index: int = 0, depth_index: int = None) → ndarray

Backward compatibility method using time index

get_var_ndims(var_name: str) → int

Get the number of dimensions of a variable.

Args:

var_name (str): Variable name.

Returns:

int: Number of dimensions.

property lats

property lats_2D

property lons

property lons_2D

property n_depths

property n_files

Get the number of files being used.

property time_span

Get the time span covered by the dataset.

property unmasked_lats

property unmasked_lons