Skip to content

CFComplianceMixin

Provides CF (Climate and Forecast) convention support.

Overview

The CFComplianceMixin helps create CF-compliant datasets by:

  • Detecting axis types - Automatically infer X, Y, Z, T axes
  • Validating metadata - Check for CF compliance issues
  • Setting attributes - Apply CF-standard axis attributes
  • Querying coordinates - Find coordinates by axis type

Key Methods

  • infer_axis() - Detect axis types for all coordinates
  • set_axis_attributes() - Set axis attributes on coordinates
  • get_axis_coordinates(axis) - Get coordinates for specific axis
  • validate_cf(strict=False) - Validate CF compliance

Axis Detection

Axes are detected based on:

  • Coordinate names - time, lat, lon, lev, etc.
  • Units - degrees_north, days since, hPa, etc.
  • Standard names - latitude, longitude, time, etc.

Usage

See the CF Compliance Guide for detailed examples.

API Reference

Mixin providing CF compliance and axis detection capabilities.

Source code in src/dummyxarray/cf_compliance.py 
  9
 10
 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
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
class CFComplianceMixin:
    """Mixin providing CF compliance and axis detection capabilities."""

    def infer_axis(self, coord_name=None):
        """
        Infer axis attribute (X/Y/Z/T) for coordinates based on CF conventions.

        Uses coordinate names, standard_name attributes, units, and dimension
        patterns to automatically detect axis types.

        Parameters
        ----------
        coord_name : str, optional
            Specific coordinate to infer axis for. If None, infers for all coordinates.

        Returns
        -------
        dict
            Dictionary mapping coordinate names to inferred axis values ('X', 'Y', 'Z', 'T')

        Examples
        --------
        >>> ds = DummyDataset()
        >>> ds.add_dim("time", 10)
        >>> ds.add_dim("lat", 64)
        >>> ds.add_dim("lon", 128)
        >>> ds.add_coord("time", dims=["time"], attrs={"units": "days since 2000-01-01"})
        >>> ds.add_coord("lat", dims=["lat"], attrs={"units": "degrees_north"})
        >>> ds.add_coord("lon", dims=["lon"], attrs={"units": "degrees_east"})
        >>> axes = ds.infer_axis()
        >>> # Returns: {'time': 'T', 'lat': 'Y', 'lon': 'X'}
        """
        axes = {}
        coords_to_check = [coord_name] if coord_name else list(self.coords.keys())

        for name in coords_to_check:
            if name not in self.coords:
                continue

            coord = self.coords[name]
            axis = self._detect_axis_type(name, coord)
            if axis:
                axes[name] = axis

        return axes

    def _detect_axis_type(self, name, coord):
        """
        Detect axis type for a coordinate based on CF conventions.

        Parameters
        ----------
        name : str
            Coordinate name
        coord : DummyArray
            Coordinate object

        Returns
        -------
        str or None
            Axis type ('X', 'Y', 'Z', 'T') or None if cannot be determined
        """
        # Check if axis already set
        if coord.attrs.get("axis"):
            return coord.attrs["axis"]

        # Check standard_name (CF convention)
        standard_name = coord.attrs.get("standard_name", "").lower()
        standard_name_map = {
            "longitude": "X",
            "projection_x_coordinate": "X",
            "grid_longitude": "X",
            "latitude": "Y",
            "projection_y_coordinate": "Y",
            "grid_latitude": "Y",
            "altitude": "Z",
            "height": "Z",
            "depth": "Z",
            "air_pressure": "Z",
            "model_level_number": "Z",
            "time": "T",
        }
        if standard_name in standard_name_map:
            return standard_name_map[standard_name]

        # Check units (CF convention)
        units = coord.attrs.get("units", "").lower()

        # Time axis patterns
        time_patterns = ["since", "days", "hours", "minutes", "seconds"]
        if any(pattern in units for pattern in time_patterns):
            return "T"

        # Longitude patterns
        if units in ["degrees_east", "degree_east", "degreee", "degreese"]:
            return "X"

        # Latitude patterns
        if units in ["degrees_north", "degree_north", "degreen", "degreesn"]:
            return "Y"

        # Vertical coordinate patterns
        vertical_units = ["pa", "hpa", "mbar", "bar", "m", "km", "level", "sigma", "eta"]
        if any(units.startswith(u) for u in vertical_units):
            return "Z"

        # Check coordinate name patterns (common conventions)
        name_lower = name.lower()

        # X-axis patterns
        x_patterns = ["lon", "longitude", "x", "i", "ni", "xc"]
        if any(name_lower.startswith(p) or name_lower == p for p in x_patterns):
            return "X"

        # Y-axis patterns
        y_patterns = ["lat", "latitude", "y", "j", "nj", "yc"]
        if any(name_lower.startswith(p) or name_lower == p for p in y_patterns):
            return "Y"

        # Z-axis patterns
        z_patterns = ["lev", "level", "plev", "height", "depth", "alt", "z", "k", "nk"]
        if any(name_lower.startswith(p) or name_lower == p for p in z_patterns):
            return "Z"

        # T-axis patterns
        t_patterns = ["time", "t", "date"]
        if any(name_lower.startswith(p) or name_lower == p for p in t_patterns):
            return "T"

        return None

    def set_axis_attributes(self, inferred_only=False):
        """
        Set axis attributes on coordinates based on inferred axis types.

        This modifies coordinate attributes in-place to add 'axis' attributes
        following CF conventions.

        Parameters
        ----------
        inferred_only : bool, default False
            If True, only set axis for coordinates that don't already have one.
            If False, overwrite existing axis attributes with inferred values.

        Returns
        -------
        dict
            Dictionary of coordinate names and their assigned axis values

        Examples
        --------
        >>> ds = DummyDataset()
        >>> ds.add_dim("time", 10)
        >>> ds.add_coord("time", dims=["time"], attrs={"units": "days since 2000-01-01"})
        >>> ds.set_axis_attributes()
        >>> print(ds.coords["time"].attrs["axis"])
        T
        """
        assigned = {}

        for coord_name, coord in self.coords.items():
            # Check if we should skip this coordinate
            if inferred_only and "axis" in coord.attrs:
                continue

            # Temporarily remove axis attribute to force re-inference
            existing_axis = coord.attrs.pop("axis", None)

            # Infer axis for this coordinate
            axis = self._detect_axis_type(coord_name, coord)

            if axis:
                # Set the inferred axis attribute
                coord.attrs["axis"] = axis
                assigned[coord_name] = axis
            elif existing_axis:
                # Restore existing axis if we couldn't infer a new one
                coord.attrs["axis"] = existing_axis

        return assigned

    def get_axis_coordinates(self, axis):
        """
        Get all coordinates with a specific axis attribute.

        Parameters
        ----------
        axis : str
            Axis type to search for ('X', 'Y', 'Z', 'T')

        Returns
        -------
        list
            List of coordinate names with the specified axis

        Examples
        --------
        >>> ds = DummyDataset()
        >>> ds.add_coord("lon", dims=["lon"], attrs={"axis": "X"})
        >>> ds.add_coord("lat", dims=["lat"], attrs={"axis": "Y"})
        >>> x_coords = ds.get_axis_coordinates("X")
        >>> # Returns: ['lon']
        """
        coords = []
        for name, coord in self.coords.items():
            if coord.attrs.get("axis") == axis:
                coords.append(name)
        return coords

    def validate_cf(self, strict=False):
        """
        Validate dataset against CF conventions.

        Checks for common CF compliance issues like missing axis attributes,
        invalid units, missing standard_name, etc.

        Parameters
        ----------
        strict : bool, default False
            If True, raise ValueError on any CF violation.
            If False, return a list of warnings/errors.

        Returns
        -------
        dict
            Dictionary with 'errors' and 'warnings' lists

        Examples
        --------
        >>> ds = DummyDataset()
        >>> ds.add_dim("time", 10)
        >>> ds.add_coord("time", dims=["time"])
        >>> result = ds.validate_cf()
        >>> print(result['warnings'])
        ['time: Missing axis attribute', 'time: Missing units attribute']
        """
        errors = []
        warnings = []

        # Check coordinates for axis attributes
        for name, coord in self.coords.items():
            # Check for axis attribute
            if "axis" not in coord.attrs:
                inferred = self.infer_axis(name)
                if name in inferred:
                    warnings.append(
                        f"{name}: Missing 'axis' attribute (can be inferred as '{inferred[name]}')"
                    )
                else:
                    warnings.append(f"{name}: Missing 'axis' attribute (cannot infer)")

            # Check for units
            if "units" not in coord.attrs:
                warnings.append(f"{name}: Missing 'units' attribute")

            # Check for standard_name on coordinates
            if "standard_name" not in coord.attrs:
                warnings.append(f"{name}: Missing 'standard_name' attribute")

        # Check variables for required attributes
        for name, var in self.variables.items():
            # Check for units
            if "units" not in var.attrs:
                warnings.append(f"{name}: Variable missing 'units' attribute")

            # Check for long_name or standard_name
            if "long_name" not in var.attrs and "standard_name" not in var.attrs:
                warnings.append(f"{name}: Variable missing both 'long_name' and 'standard_name'")

        # Check global attributes
        required_global = ["Conventions"]
        for attr in required_global:
            if attr not in self.attrs:
                warnings.append(f"Missing required global attribute: '{attr}'")

        # Check for CF Conventions version
        if "Conventions" in self.attrs:
            conventions = self.attrs["Conventions"]
            if not any(cf in conventions for cf in ["CF-", "cf-"]):
                warnings.append(f"Conventions attribute '{conventions}' does not reference CF")

        # Check dimension ordering (CF recommends T, Z, Y, X)
        for name, var in self.variables.items():
            if var.dims and len(var.dims) > 1:
                # Get axis types for dimensions
                dim_axes = []
                for dim in var.dims:
                    if dim in self.coords:
                        axis = self.coords[dim].attrs.get("axis")
                        if axis:
                            dim_axes.append(axis)

                # Check if order is T, Z, Y, X
                expected_order = ["T", "Z", "Y", "X"]
                actual_order = [a for a in dim_axes if a in expected_order]
                sorted_order = sorted(actual_order, key=lambda x: expected_order.index(x))

                if actual_order != sorted_order:
                    warnings.append(
                        f"{name}: Dimension order {actual_order} does not follow "
                        f"CF recommendation (T, Z, Y, X)"
                    )

        result = {"errors": errors, "warnings": warnings}

        if strict and (errors or warnings):
            all_issues = errors + warnings
            raise ValueError("CF validation failed:\n" + "\n".join(all_issues))

        return result

infer_axis 

infer_axis(coord_name=None)

Infer axis attribute (X/Y/Z/T) for coordinates based on CF conventions.

Uses coordinate names, standard_name attributes, units, and dimension patterns to automatically detect axis types.

Parameters:

Name Type Description Default
coord_name str

Specific coordinate to infer axis for. If None, infers for all coordinates.

 None

Returns:

Type Description
dict

Dictionary mapping coordinate names to inferred axis values ('X', 'Y', 'Z', 'T')

Examples:

>>> ds = DummyDataset()
>>> ds.add_dim("time", 10)
>>> ds.add_dim("lat", 64)
>>> ds.add_dim("lon", 128)
>>> ds.add_coord("time", dims=["time"], attrs={"units": "days since 2000-01-01"})
>>> ds.add_coord("lat", dims=["lat"], attrs={"units": "degrees_north"})
>>> ds.add_coord("lon", dims=["lon"], attrs={"units": "degrees_east"})
>>> axes = ds.infer_axis()
>>> # Returns: {'time': 'T', 'lat': 'Y', 'lon': 'X'}
Source code in src/dummyxarray/cf_compliance.py 
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
def infer_axis(self, coord_name=None):
    """
    Infer axis attribute (X/Y/Z/T) for coordinates based on CF conventions.

    Uses coordinate names, standard_name attributes, units, and dimension
    patterns to automatically detect axis types.

    Parameters
    ----------
    coord_name : str, optional
        Specific coordinate to infer axis for. If None, infers for all coordinates.

    Returns
    -------
    dict
        Dictionary mapping coordinate names to inferred axis values ('X', 'Y', 'Z', 'T')

    Examples
    --------
    >>> ds = DummyDataset()
    >>> ds.add_dim("time", 10)
    >>> ds.add_dim("lat", 64)
    >>> ds.add_dim("lon", 128)
    >>> ds.add_coord("time", dims=["time"], attrs={"units": "days since 2000-01-01"})
    >>> ds.add_coord("lat", dims=["lat"], attrs={"units": "degrees_north"})
    >>> ds.add_coord("lon", dims=["lon"], attrs={"units": "degrees_east"})
    >>> axes = ds.infer_axis()
    >>> # Returns: {'time': 'T', 'lat': 'Y', 'lon': 'X'}
    """
    axes = {}
    coords_to_check = [coord_name] if coord_name else list(self.coords.keys())

    for name in coords_to_check:
        if name not in self.coords:
            continue

        coord = self.coords[name]
        axis = self._detect_axis_type(name, coord)
        if axis:
            axes[name] = axis

    return axes

set_axis_attributes 

set_axis_attributes(inferred_only=False)

Set axis attributes on coordinates based on inferred axis types.

This modifies coordinate attributes in-place to add 'axis' attributes following CF conventions.

Parameters:

Name Type Description Default
inferred_only bool

If True, only set axis for coordinates that don't already have one. If False, overwrite existing axis attributes with inferred values.

 False

Returns:

Type Description
dict

Dictionary of coordinate names and their assigned axis values

Examples:

>>> ds = DummyDataset()
>>> ds.add_dim("time", 10)
>>> ds.add_coord("time", dims=["time"], attrs={"units": "days since 2000-01-01"})
>>> ds.set_axis_attributes()
>>> print(ds.coords["time"].attrs["axis"])
T
Source code in src/dummyxarray/cf_compliance.py 
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
def set_axis_attributes(self, inferred_only=False):
    """
    Set axis attributes on coordinates based on inferred axis types.

    This modifies coordinate attributes in-place to add 'axis' attributes
    following CF conventions.

    Parameters
    ----------
    inferred_only : bool, default False
        If True, only set axis for coordinates that don't already have one.
        If False, overwrite existing axis attributes with inferred values.

    Returns
    -------
    dict
        Dictionary of coordinate names and their assigned axis values

    Examples
    --------
    >>> ds = DummyDataset()
    >>> ds.add_dim("time", 10)
    >>> ds.add_coord("time", dims=["time"], attrs={"units": "days since 2000-01-01"})
    >>> ds.set_axis_attributes()
    >>> print(ds.coords["time"].attrs["axis"])
    T
    """
    assigned = {}

    for coord_name, coord in self.coords.items():
        # Check if we should skip this coordinate
        if inferred_only and "axis" in coord.attrs:
            continue

        # Temporarily remove axis attribute to force re-inference
        existing_axis = coord.attrs.pop("axis", None)

        # Infer axis for this coordinate
        axis = self._detect_axis_type(coord_name, coord)

        if axis:
            # Set the inferred axis attribute
            coord.attrs["axis"] = axis
            assigned[coord_name] = axis
        elif existing_axis:
            # Restore existing axis if we couldn't infer a new one
            coord.attrs["axis"] = existing_axis

    return assigned

get_axis_coordinates 

get_axis_coordinates(axis)

Get all coordinates with a specific axis attribute.

Parameters:

Name Type Description Default
axis str

Axis type to search for ('X', 'Y', 'Z', 'T')

 required

Returns:

Type Description
list

List of coordinate names with the specified axis

Examples:

>>> ds = DummyDataset()
>>> ds.add_coord("lon", dims=["lon"], attrs={"axis": "X"})
>>> ds.add_coord("lat", dims=["lat"], attrs={"axis": "Y"})
>>> x_coords = ds.get_axis_coordinates("X")
>>> # Returns: ['lon']
Source code in src/dummyxarray/cf_compliance.py 
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
def get_axis_coordinates(self, axis):
    """
    Get all coordinates with a specific axis attribute.

    Parameters
    ----------
    axis : str
        Axis type to search for ('X', 'Y', 'Z', 'T')

    Returns
    -------
    list
        List of coordinate names with the specified axis

    Examples
    --------
    >>> ds = DummyDataset()
    >>> ds.add_coord("lon", dims=["lon"], attrs={"axis": "X"})
    >>> ds.add_coord("lat", dims=["lat"], attrs={"axis": "Y"})
    >>> x_coords = ds.get_axis_coordinates("X")
    >>> # Returns: ['lon']
    """
    coords = []
    for name, coord in self.coords.items():
        if coord.attrs.get("axis") == axis:
            coords.append(name)
    return coords

validate_cf 

validate_cf(strict=False)

Validate dataset against CF conventions.

Checks for common CF compliance issues like missing axis attributes, invalid units, missing standard_name, etc.

Parameters:

Name Type Description Default
strict bool

If True, raise ValueError on any CF violation. If False, return a list of warnings/errors.

 False

Returns:

Type Description
dict

Dictionary with 'errors' and 'warnings' lists

Examples:

>>> ds = DummyDataset()
>>> ds.add_dim("time", 10)
>>> ds.add_coord("time", dims=["time"])
>>> result = ds.validate_cf()
>>> print(result['warnings'])
['time: Missing axis attribute', 'time: Missing units attribute']
Source code in src/dummyxarray/cf_compliance.py 
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
def validate_cf(self, strict=False):
    """
    Validate dataset against CF conventions.

    Checks for common CF compliance issues like missing axis attributes,
    invalid units, missing standard_name, etc.

    Parameters
    ----------
    strict : bool, default False
        If True, raise ValueError on any CF violation.
        If False, return a list of warnings/errors.

    Returns
    -------
    dict
        Dictionary with 'errors' and 'warnings' lists

    Examples
    --------
    >>> ds = DummyDataset()
    >>> ds.add_dim("time", 10)
    >>> ds.add_coord("time", dims=["time"])
    >>> result = ds.validate_cf()
    >>> print(result['warnings'])
    ['time: Missing axis attribute', 'time: Missing units attribute']
    """
    errors = []
    warnings = []

    # Check coordinates for axis attributes
    for name, coord in self.coords.items():
        # Check for axis attribute
        if "axis" not in coord.attrs:
            inferred = self.infer_axis(name)
            if name in inferred:
                warnings.append(
                    f"{name}: Missing 'axis' attribute (can be inferred as '{inferred[name]}')"
                )
            else:
                warnings.append(f"{name}: Missing 'axis' attribute (cannot infer)")

        # Check for units
        if "units" not in coord.attrs:
            warnings.append(f"{name}: Missing 'units' attribute")

        # Check for standard_name on coordinates
        if "standard_name" not in coord.attrs:
            warnings.append(f"{name}: Missing 'standard_name' attribute")

    # Check variables for required attributes
    for name, var in self.variables.items():
        # Check for units
        if "units" not in var.attrs:
            warnings.append(f"{name}: Variable missing 'units' attribute")

        # Check for long_name or standard_name
        if "long_name" not in var.attrs and "standard_name" not in var.attrs:
            warnings.append(f"{name}: Variable missing both 'long_name' and 'standard_name'")

    # Check global attributes
    required_global = ["Conventions"]
    for attr in required_global:
        if attr not in self.attrs:
            warnings.append(f"Missing required global attribute: '{attr}'")

    # Check for CF Conventions version
    if "Conventions" in self.attrs:
        conventions = self.attrs["Conventions"]
        if not any(cf in conventions for cf in ["CF-", "cf-"]):
            warnings.append(f"Conventions attribute '{conventions}' does not reference CF")

    # Check dimension ordering (CF recommends T, Z, Y, X)
    for name, var in self.variables.items():
        if var.dims and len(var.dims) > 1:
            # Get axis types for dimensions
            dim_axes = []
            for dim in var.dims:
                if dim in self.coords:
                    axis = self.coords[dim].attrs.get("axis")
                    if axis:
                        dim_axes.append(axis)

            # Check if order is T, Z, Y, X
            expected_order = ["T", "Z", "Y", "X"]
            actual_order = [a for a in dim_axes if a in expected_order]
            sorted_order = sorted(actual_order, key=lambda x: expected_order.index(x))

            if actual_order != sorted_order:
                warnings.append(
                    f"{name}: Dimension order {actual_order} does not follow "
                    f"CF recommendation (T, Z, Y, X)"
                )

    result = {"errors": errors, "warnings": warnings}

    if strict and (errors or warnings):
        all_issues = errors + warnings
        raise ValueError("CF validation failed:\n" + "\n".join(all_issues))

    return result