Skip to content

create_empty_dataset

Functions:

Name Description
empty_dataset_like

Creates a new xr.Dataset containing the same variables and metadata as the input

Functions#

empty_dataset_like #

empty_dataset_like(
    input_dataset: Dataset,
    coord_name: str,
    coord_values: ndarray,
    coord_bounds: ndarray,
    add_transform_qc: bool = True,
    add_metric_vars: bool = False,
) -> xr.Dataset

Creates a new xr.Dataset containing the same variables and metadata as the input dataset, but with new coordinate values and bounds for the provided coordinate variable.

Other coordinate variables and coordinate bounds variables will remain untouched. All data variables that are at least partially dimensioned by the provided coordinate will have their metadata copied but their values will be set to NaN. Variables that are dimensioned solely by other coordinate variables will be copied directly over.

Parameters:

Name Type Description Default
input_dataset Dataset

The input xarray Dataset.

 required
coord_name str

The name of the coordinate variable to modify.

 required
coord_values ndarray

The new values for the coordinate variable.

 required
coord_bounds ndarray

The new bounds for the coordinate variable.

 required

Returns:

Type Description
Dataset

xr.Dataset: The new xarray Dataset with modified coordinate values and bounds.
Source code in tsdat/transform_v2/utils/create_empty_dataset.py 
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
def empty_dataset_like(
    input_dataset: xr.Dataset,
    coord_name: str,
    coord_values: np.ndarray,
    coord_bounds: np.ndarray,
    add_transform_qc: bool = True,
    add_metric_vars: bool = False,
) -> xr.Dataset:
    """Creates a new xr.Dataset containing the same variables and metadata as the input
    dataset, but with new coordinate values and bounds for the provided coordinate
    variable.

    Other coordinate variables and coordinate bounds variables will remain untouched.
    All data variables that are at least partially dimensioned by the provided
    coordinate will have their metadata copied but their values will be set to NaN.
    Variables that are dimensioned solely by other coordinate variables will be copied
    directly over.

    Args:
        input_dataset (xr.Dataset): The input xarray Dataset.
        coord_name (str): The name of the coordinate variable to modify.
        coord_values (np.ndarray): The new values for the coordinate variable.
        coord_bounds (np.ndarray): The new bounds for the coordinate variable.

    Returns:
        xr.Dataset: The new xarray Dataset with modified coordinate values and bounds.
    """

    # Create a new empty dataset
    new_dataset = xr.Dataset(attrs=input_dataset.attrs)

    # Add coordinates
    for in_coord_name, in_coord_data_array in input_dataset.coords.items():
        if in_coord_name == coord_name:
            new_dataset[in_coord_name] = xr.DataArray(
                coord_values,
                dims=in_coord_data_array.dims,
                attrs=in_coord_data_array.attrs,
            )
        else:
            new_dataset = new_dataset.assign_coords({coord_name: in_coord_data_array})

    # Add bounds variables
    for new_coord_name in new_dataset.coords:
        bounds_name = f"{new_coord_name}_bounds"
        bounds_data_array = input_dataset.get(bounds_name, None)
        if new_coord_name == coord_name:
            new_dataset[bounds_name] = xr.DataArray(
                coord_bounds,
                dims=(new_coord_name, "bound"),
                attrs=bounds_data_array.attrs if bounds_data_array is not None else {},
            )
        elif bounds_data_array is not None:
            new_dataset[bounds_name] = bounds_data_array

    # Add data variables
    for var_name, data_array in input_dataset.data_vars.items():
        if var_name in new_dataset:  # skip bounds vars
            continue

        if coord_name in data_array.dims:
            if add_transform_qc and is_qc_var(data_array):
                continue
            if add_metric_vars and is_metric_var(data_array):
                continue

            new_shape = tuple(new_dataset.sizes[d] for d in data_array.dims)
            new_dataset[var_name] = xr.DataArray(
                np.full(new_shape, fill_value=np.nan),
                dims=data_array.dims,
                attrs=data_array.attrs,
            )
            if add_transform_qc:
                add_empty_transform_qc_var(new_dataset, var_name)

            if add_metric_vars:
                add_empty_std_dev_var(new_dataset, var_name)
                add_empty_goodfrac_var(new_dataset, var_name)
        else:
            new_dataset[var_name] = data_array

    return new_dataset