def test_x_ticks_are_rotated_for_time(self):
time = pd.date_range('2000-01-01', '2000-01-10')
a = DataArray(np.arange(len(time)), [('t', time)])
a.plot.line()
rotation = plt.gca().get_xticklabels()[0].get_rotation()
self.assertFalse(rotation == 0)
def impedance(n,
branch_components=None,
snapshot=None,
pu_system=True,
linear=True,
skip_pre=False):
"""
Calculate the impedance of the network branches.
Naturally the impdance of controllable branches is not existent. However,
in https://www.preprints.org/manuscript/202001.0352/v1 a method was
presented how to calculate the impendance of controllable branches if they
were passive AC lines. If 'Link' is included in branch_components,
the flow-dependent pseudo-impedance is calculated based on the formulation
presented in the paper. Note that in this case the flow must be given
for all branches.
Parameters
----------
n : pypsa.Network
branch_components : list, optional
List of branch components. The default None results in
n.passive_branch_components.
snapshot : str/pd.Timestamp, optional
Only relevant if 'Link' in branch_components. The default None results
in the first snapshot of n.
pu_system : bool, optional
Whether the use the per uni system for the impendance.
The default is True.
linear : bool, optional
Whether to use the linear approximation. The default is True.
skip_pre : bool, optional
Whether to calcuate dependent quantities beforehand. The default is False.
Returns
-------
z : xr.DataArray
Impedance for each branch in branch_components.
"""
# standard impedance, note z must not be inf or nan
branch_components = check_passive_branch_comps(branch_components, n)
x = 'x_pu' if pu_system else 'x'
r = 'r_pu' if pu_system else 'r'
if not skip_pre:
if pu_system and (n.lines[x] == 0).all():
n.calculate_dependent_values()
comps = sorted(set(branch_components) & n.passive_branch_components)
if linear:
z = pd.concat({
c: n.df(c)[x].where(
n.df(c).bus0.map(n.buses.carrier) == 'AC',
n.df(c)[r])
for c in comps
})
else:
z = pd.concat({c: n.df(c).eval(f'{r} + 1.j * {x}') for c in comps})
if not n.lines.empty:
assert not np.isinf(z).any() | z.isna().any(), (
'There '
f'seems to be a problem with your {x} or {r} values. At least one of '
f'these is nan or inf. Please check the values in components {comps}.'
)
z = DataArray(z.rename_axis(['component', 'branch_i']), dims='branch')
if ('Link' not in branch_components) | n.links.empty:
return z
# add pseudo impedance for links, in dependence on the current flow:
if snapshot is None:
logger.warn('Link in argument "branch_components", but no '
'snapshot given. Falling back to first snapshot')
snapshot = n.snapshots[0]
f = network_flow(n, snapshot)
branches_i = f.get_index('branch')
C = Cycles(n, branches_i[abs(f).values > 1e-8])\
.reindex(branch=branches_i, fill_value=0)
# C_mix is all the active cycles where at least one link is included
C_mix = C[:, ((C != 0) &
(f != 0)).groupby('component').any().loc['Link'].values]
if not C_mix.size:
sub = f.loc['Link'][abs(f.loc['Link']).values > 1e-8]
omega = DataArray(1, sub.coords)
elif not z.size:
omega = null(C_mix.loc['Link'] * f.loc['Link'])[0]
else:
d = {'branch': 'Link'}
omega = -dot(pinv(dot(C_mix.loc['Link'].T, diag(f.loc['Link']))),
dot(C_mix.drop_sel(d).T, diag(z), f.drop_sel(d)))
omega = omega.round(10).assign_coords({'component': 'Link'})
omega[(omega == 0) & (f.loc['Link'] != 0)] = 1
Z = z.reindex_like(f).copy()
links_i = branches_i[branches_i.get_loc('Link')]
Z.loc[links_i] = omega.reindex_like(f.loc['Link'], fill_value=0)
return Z.assign_coords(snapshot=snapshot)
def array_with_type(general, specific):
return DataArray(0.0, attrs={"datatype": (general, specific)})
def value_check(self):
Ne = 0 * self.Ne
with self.assertRaises(ValueError):
utilities.input_check("Ne",
Ne,
np.ndarray,
greater_than_or_equal_zero=False)
Ne = -1 * self.Ne
with self.assertRaises(ValueError):
utilities.input_check("Ne",
Ne,
np.ndarray,
greater_than_or_equal_zero=True)
Ne = np.nan * self.Ne
with self.assertRaises(ValueError):
utilities.input_check("Ne", Ne, np.ndarray)
Ne = np.inf * self.Ne
with self.assertRaises(ValueError):
utilities.input_check("Ne", Ne, np.ndarray)
Ne = -np.inf * self.Ne
with self.assertRaises(ValueError):
utilities.input_check("Ne", Ne, np.ndarray)
Ne = self.Ne[:, np.newaxis]
with self.assertRaises(ValueError):
utilities.input_check("Ne", Ne, np.ndarray, ndim_to_check=1)
# Check dropped channel handling
t = np.array([78.5, 80.5, 82.5])
rho = np.linspace(0, 1, 11)
Ne = np.logspace(19.0, 16.0, 11)
Ne = np.tile(Ne, [3, 1])
Ne[1, :] /= 10.0
Ne[2, :] *= 10.0
dropped_t_coord = np.array([80.5])
dropped_rho_coord = np.array([rho[3], rho[7]])
Ne = DataArray(
data=Ne,
coords=[("t", t), ("rho_poloidal", rho)],
dims=["t", "rho_poloidal"],
)
dropped = Ne.sel({"t": dropped_t_coord})
dropped = dropped.sel({"rho_poloidal": dropped_rho_coord})
Ne.loc[{
"t": dropped_t_coord,
"rho_poloidal": dropped_rho_coord
}] = np.nan
Ne.attrs["dropped"] = dropped
try:
utilities.input_check("Ne", Ne, DataArray, ndim_to_check=2)
except Exception as e:
raise e
def test_contains_cftime_datetimes_dask_1d(data):
assert contains_cftime_datetimes(data.time.chunk())
@requires_cftime
def test_contains_cftime_datetimes_3d(times_3d):
assert contains_cftime_datetimes(times_3d)
@requires_cftime
@requires_dask
def test_contains_cftime_datetimes_dask_3d(times_3d):
assert contains_cftime_datetimes(times_3d.chunk())
@pytest.mark.parametrize("non_cftime_data", [DataArray([]), DataArray([1, 2])])
def test_contains_cftime_datetimes_non_cftimes(non_cftime_data):
assert not contains_cftime_datetimes(non_cftime_data)
@requires_dask
@pytest.mark.parametrize("non_cftime_data", [DataArray([]), DataArray([1, 2])])
def test_contains_cftime_datetimes_non_cftimes_dask(non_cftime_data):
assert not contains_cftime_datetimes(non_cftime_data.chunk())
@requires_cftime
@pytest.mark.parametrize("shape", [(24, ), (8, 3), (2, 4, 3)])
def test_encode_cf_datetime_overflow(shape):
# Test for fix to GH 2272
dates = pd.date_range("2100", periods=24).values.reshape(shape)
xi, yi, mski = d['lont'], d['latt'], np.bool8(d['volmsk'])
msk = np.logical_and(msko, ~mski)
TAREAmsk = TAREA[msk]
A = TAREAmsk.sum()
# fnames = fnames[-90:]
t = np.array([])
Cterms = dict()
_ = [Cterms.update({term: np.array([], ndmin=1)}) for term in terms]
n = 0
for fname in fnames:
date = fname.split('.')[0][-10:]
print(date)
t = np.append(t, Timestamp(date).to_pydatetime())
ds = open_dataset(fname)
for term in terms:
dsterm = ds[term]
dsterm[:, 0] = np.nan
dsterm[:, -1] = np.nan
aux = np.nansum(dsterm.values[msk] * TAREAmsk) / A # Area-averaging.
# NOTE: Excluding first and last columns because curl did not wrap around.
Cterms[term] = np.append(Cterms[term], aux)
for term in terms:
Cterms[term] = DataArray(Cterms[term], coords=dict(t=t), dims='t')
ds = Datasetx(data_vars=Cterms, coords=dict(t=t))
fout = head + 'data/circulation_terms_circumpolar.nc'
ds.to_netcdf(fout)
def apply(raster, kernel, func=_calc_mean, name='focal_apply'):
"""
Returns custom function applied array using a user-created window.
Parameters
----------
raster : xarray.DataArray
2D array of input values to be filtered. Can be a NumPy backed,
or Dask with NumPy backed DataArray.
kernel : numpy.ndarray
2D array where values of 1 indicate the kernel.
func : callable, default=xrspatial.focal._calc_mean
Function which takes an input array and returns an array.
Returns
-------
agg : xarray.DataArray of same type as `raster`
2D aggregate array of filtered values.
Examples
--------
Focal apply works with NumPy backed xarray DataArray
.. sourcecode:: python
>>> import numpy as np
>>> import xarray as xr
>>> from xrspatial.convolution import circle_kernel
>>> from xrspatial.focal import apply
>>> data = np.arange(20, dtype=np.float64).reshape(4, 5)
>>> raster = xr.DataArray(data, dims=['y', 'x'], name='raster')
>>> print(raster)
<xarray.DataArray 'raster' (y: 4, x: 5)>
array([[ 0., 1., 2., 3., 4.],
[ 5., 6., 7., 8., 9.],
[10., 11., 12., 13., 14.],
[15., 16., 17., 18., 19.]])
Dimensions without coordinates: y, x
>>> kernel = circle_kernel(2, 2, 3)
>>> kernel
array([[0., 1., 0.],
[1., 1., 1.],
[0., 1., 0.]])
>>> # apply kernel mean by default
>>> apply_mean_agg = apply(raster, kernel)
>>> apply_mean_agg
<xarray.DataArray 'focal_apply' (y: 4, x: 5)>
array([[ 2. , 2.25 , 3.25 , 4.25 , 5.33333333],
[ 5.25 , 6. , 7. , 8. , 8.75 ],
[10.25 , 11. , 12. , 13. , 13.75 ],
[13.66666667, 14.75 , 15.75 , 16.75 , 17. ]])
Dimensions without coordinates: y, x
Focal apply works with Dask with NumPy backed xarray DataArray.
Note that if input raster is a numpy or dask with numpy backed data array,
the applied function must be decorated with ``numba.jit``
xrspatial already provides ``ngjit`` decorator, where:
``ngjit = numba.jit(nopython=True, nogil=True)``
.. sourcecode:: python
>>> from xrspatial.utils import ngjit
>>> from xrspatial.convolution import custom_kernel
>>> kernel = custom_kernel(np.array([
[0, 1, 0],
[0, 1, 1],
[0, 1, 0],
]))
>>> weight = np.array([
[0, 0.5, 0],
[0, 1, 0.5],
[0, 0.5, 0],
])
>>> @ngjit
>>> def func(kernel_data):
... weight = np.array([
... [0, 0.5, 0],
... [0, 1, 0.5],
... [0, 0.5, 0],
... ])
... return np.nansum(kernel_data * weight)
>>> import dask.array as da
>>> data_da = da.from_array(np.ones((6, 4), dtype=np.float64), chunks=(3, 2))
>>> raster_da = xr.DataArray(data_da, dims=['y', 'x'], name='raster_da')
>>> print(raster_da)
<xarray.DataArray 'raster_da' (y: 6, x: 4)>
dask.array<array, shape=(6, 4), dtype=float64, chunksize=(3, 2), chunktype=numpy.ndarray> # noqa
Dimensions without coordinates: y, x
>>> apply_func_agg = apply(raster_da, kernel, func)
>>> print(apply_func_agg)
<xarray.DataArray 'focal_apply' (y: 6, x: 4)>
dask.array<_trim, shape=(6, 4), dtype=float64, chunksize=(3, 2), chunktype=numpy.ndarray> # noqa
Dimensions without coordinates: y, x
>>> print(apply_func_agg.compute())
<xarray.DataArray 'focal_apply' (y: 6, x: 4)>
array([[2. , 2. , 2. , 1.5],
[2.5, 2.5, 2.5, 2. ],
[2.5, 2.5, 2.5, 2. ],
[2.5, 2.5, 2.5, 2. ],
[2.5, 2.5, 2.5, 2. ],
[2. , 2. , 2. , 1.5]])
Dimensions without coordinates: y, x
"""
# validate raster
if not isinstance(raster, DataArray):
raise TypeError("`raster` must be instance of DataArray")
if raster.ndim != 2:
raise ValueError("`raster` must be 2D")
# Validate the kernel
kernel = custom_kernel(kernel)
# apply kernel to raster values
# if raster is a numpy or dask with numpy backed data array,
# the function func must be a @ngjit
mapper = ArrayTypeFunctionMapping(
numpy_func=_apply_numpy,
cupy_func=lambda *args: not_implemented_func(
*args, messages='apply() does not support cupy backed DataArray.'),
dask_func=_apply_dask_numpy,
dask_cupy_func=lambda *args: not_implemented_func(
*args,
messages=
'apply() does not support dask with cupy backed DataArray.'),
)
out = mapper(raster)(raster.data, kernel, func)
result = DataArray(out,
name=name,
coords=raster.coords,
dims=raster.dims,
attrs=raster.attrs)
return result
def quantile(agg: xr.DataArray,
k: int = 4,
name: Optional[str] = 'quantile') -> xr.DataArray:
"""
Groups data for array (agg) into quantiles by distributing
the values into groups that contain an equal number of values.
The number of quantiles produced is based on (k) with a default value
of 4. The result is an xarray.DataArray.
Parameters:
----------
agg: xarray.DataArray
2D array of values to bin:
NumPy, CuPy, NumPy-backed Dask, or Cupy-backed Dask array.
k: int
Number of quantiles to be produced, default = 4.
name: str, optional (default = "quantile")
Name of the output aggregate array.
Returns:
----------
xarray.DataArray, quantiled aggregate
2D array, of the same type as the input, of quantile allocations.
All other input attributes are preserved.
Notes:
----------
Adapted from PySAL:
- https://pysal.org/mapclassify/_modules/mapclassify/classifiers.html#Quantiles # noqa
Note that dask's percentile algorithm is approximate,
while numpy's is exact. This may cause some differences
between results of vanilla numpy and dask version of the input agg.
- https://github.com/dask/dask/issues/3099
Examples:
----------
Imports
>>> import numpy as np
>>> import xarray as xr
>>> from xrspatial.classify import quantile
Create DataArray
>>> np.random.seed(0)
>>> agg = xr.DataArray(np.random.rand(4,4),
dims = ["lat", "lon"])
>>> height, width = agg.shape
>>> _lat = np.linspace(0, height - 1, height)
>>> _lon = np.linspace(0, width - 1, width)
>>> agg["lat"] = _lat
>>> agg["lon"] = _lon
>>> print(agg)
<xarray.DataArray (lat: 4, lon: 4)>
array([[0.5488135 , 0.71518937, 0.60276338, 0.54488318],
[0.4236548 , 0.64589411, 0.43758721, 0.891773 ],
[0.96366276, 0.38344152, 0.79172504, 0.52889492],
[0.56804456, 0.92559664, 0.07103606, 0.0871293 ]])
Coordinates:
* lon (lon) float64 0.0 1.0 2.0 3.0
* lat (lat) float64 0.0 1.0 2.0 3.0
Create Quantile Aggregate
>>> quantile_agg = quantile(agg)
>>> print(quantile_agg)
<xarray.DataArray 'quantile' (lat: 4, lon: 4)>
array([[1., 2., 2., 1.],
[0., 2., 1., 3.],
[3., 0., 3., 1.],
[2., 3., 0., 0.]], dtype=float32)
Coordinates:
* lon (lon) float64 0.0 1.0 2.0 3.0
* lat (lat) float64 0.0 1.0 2.0 3.0
With k quantiles
>>> quantile_agg = quantile(agg, k = 6, name = "Six Quantiles")
>>> print(quantile_agg)
<xarray.DataArray 'Six Quantiles' (lat: 4, lon: 4)>
array([[2., 4., 3., 2.],
[1., 3., 1., 5.],
[5., 0., 4., 1.],
[3., 5., 0., 0.]], dtype=float32)
Coordinates:
* lon (lon) float64 0.0 1.0 2.0 3.0
* lat (lat) float64 0.0 1.0 2.0 3.0
"""
q = _quantile(agg, k)
k_q = q.shape[0]
if k_q < k:
print("Quantile Warning: Not enough unique values"
"for k classes (using {} bins)".format(k_q))
k = k_q
out = _bin(agg.data, bins=q, new_values=np.arange(k))
return DataArray(out,
name=name,
dims=agg.dims,
coords=agg.coords,
attrs=agg.attrs)
def natural_breaks(agg: xr.DataArray,
num_sample: Optional[int] = None,
name: Optional[str] = 'natural_breaks',
k: int = 5) -> xr.DataArray:
"""
Groups data for array (agg) by distributing
values using the Jenks Natural Breaks or k-means
clustering method. Values are grouped so that
similar values are placed in the same group and
space between groups is maximized.
The result is an xarray.DataArray.
Parameters:
----------
agg: xarray.DataArray
2D array of values to bin.
NumPy, CuPy, NumPy-backed Dask, or Cupy-backed Dask array
num_sample: int (optional)
Number of sample data points used to fit the model.
Natural Breaks (Jenks) classification is indeed O(n²) complexity,
where n is the total number of data points, i.e: agg.size
When n is large, we should fit the model on a small sub-sample
of the data instead of using the whole dataset.
k: int (default = 5)
Number of classes to be produced.
name: str, optional (default = "natural_breaks")
Name of output aggregate.
Returns:
----------
natural_breaks_agg: xarray.DataArray
2D array, of the same type as the input, of class allocations.
Algorithm References:
----------
Map Classify:
- https://pysal.org/mapclassify/_modules/mapclassify/classifiers.html#NaturalBreaks # noqa
perrygeo:
- https://github.com/perrygeo/jenks/blob/master/jenks.pyx
Examples:
----------
Imports
>>> import numpy as np
>>> import xarray as xr
>>> from xrspatial.classify import natural_breaks
Create DataArray
>>> np.random.seed(0)
>>> agg = xr.DataArray(np.random.rand(4,4),
dims = ["lat", "lon"])
>>> height, width = agg.shape
>>> _lat = np.linspace(0, height - 1, height)
>>> _lon = np.linspace(0, width - 1, width)
>>> agg["lat"] = _lat
>>> agg["lon"] = _lon
>>> print(agg)
<xarray.DataArray (lat: 4, lon: 4)>
array([[0.5488135 , 0.71518937, 0.60276338, 0.54488318],
[0.4236548 , 0.64589411, 0.43758721, 0.891773 ],
[0.96366276, 0.38344152, 0.79172504, 0.52889492],
[0.56804456, 0.92559664, 0.07103606, 0.0871293 ]])
Coordinates:
* lon (lon) float64 0.0 1.0 2.0 3.0
* lat (lat) float64 0.0 1.0 2.0 3.0
Create Natural Breaks Aggregate
>>> natural_breaks_agg = natural_breaks(agg, k = 5)
>>> print(natural_breaks_agg)
<xarray.DataArray 'natural_breaks' (lat: 4, lon: 4)>
array([[2., 3., 2., 2.],
[1., 2., 1., 4.],
[4., 1., 3., 2.],
[2., 4., 0., 0.]], dtype=float32)
Coordinates:
* lat (lat) float64 0.0 1.0 2.0 3.0
* lon (lon) float64 0.0 1.0 2.0 3.0
"""
# numpy case
if isinstance(agg.data, np.ndarray):
out = _run_numpy_natural_break(agg.data, num_sample, k)
# cupy case
elif has_cuda() and isinstance(agg.data, cupy.ndarray):
out = _run_cupy_natural_break(agg.data, num_sample, k)
else:
raise TypeError('Unsupported Array Type: {}'.format(type(agg.data)))
return DataArray(out,
name=name,
coords=agg.coords,
dims=agg.dims,
attrs=agg.attrs)
def test_2d_before_squeeze(self):
a = DataArray(easy_array((1, 5)))
a.plot()
def reclassify(agg: xr.DataArray,
bins: List[int],
new_values: List[int],
name: Optional[str] = 'reclassify') -> xr.DataArray:
"""
Reclassifies data for array (agg) into new values based on bins.
Parameters:
----------
agg: xarray.DataArray
2D array of values to be reclassified.
NumPy, CuPy, NumPy-backed Dask, or Cupy-backed Dask array.
bins: array-like object
Values or ranges of values to be changed.
new_values: array-like object
New values for each bin.
name: str, optional (default = "reclassify")
Name of output aggregate.
Returns:
----------
xarray.DataArray, reclassified aggregate.
2D array of new values. All input attributes are preserved.
Notes:
----------
Adapted from PySal:
- https://pysal.org/mapclassify/_modules/mapclassify/classifiers.html
Examples:
----------
Imports
>>> import numpy as np
>>> import xarray as xr
>>> from xrspatial.classify import reclassify
Create Initial DataArray
>>> np.random.seed(1)
>>> agg = xr.DataArray(np.random.randint(2, 8, (4, 4)),
>>> dims = ["lat", "lon"])
>>> height, width = agg.shape
>>> _lon = np.linspace(0, width - 1, width)
>>> _lat = np.linspace(0, height - 1, height)
>>> agg["lon"] = _lon
>>> agg["lat"] = _lat
>>> print(agg)
<xarray.DataArray (lat: 4, lon: 4)>
array([[7, 5, 6, 2],
[3, 5, 7, 2],
[2, 3, 6, 7],
[6, 3, 4, 6]])
Coordinates:
* lon (lon) float64 0.0 1.0 2.0 3.0
* lat (lat) float64 0.0 1.0 2.0 3.0
Reclassify
>>> bins = list(range(2, 8))
>>> new_val = list(range(20, 80, 10))
>>> reclassify_agg = reclassify(agg, bins, new_val)
>>> print(reclassify_agg)
<xarray.DataArray 'reclassify' (lat: 4, lon: 4)>
array([[70., 50., 60., 20.],
[30., 50., 70., 20.],
[20., 30., 60., 70.],
[60., 30., 40., 60.]], dtype=float32)
Coordinates:
* lon (lon) float64 0.0 1.0 2.0 3.0
* lat (lat) float64 0.0 1.0 2.0 3.0
"""
if len(bins) != len(new_values):
raise ValueError('bins and new_values mismatch.'
'Should have same length.')
out = _bin(agg.data, bins, new_values)
return DataArray(out,
name=name,
dims=agg.dims,
coords=agg.coords,
attrs=agg.attrs)
def setUp(self):
self.darray = DataArray(easy_array((2, 3, 4)))
def test_nonnumeric_index_raises_typeerror(self):
a = DataArray(easy_array((3, 2)), coords=[['a', 'b', 'c'], ['d', 'e']])
with raises_regex(TypeError, r'[Pp]lot'):
self.plotfunc(a)
def test_3d_raises_valueerror(self):
a = DataArray(easy_array((2, 3, 4)))
with raises_regex(ValueError, r'DataArray must be 2d'):
self.plotfunc(a)
def binary(agg, values):
return DataArray(_binary(agg.data, values),
name='binary',
dims=agg.dims,
coords=agg.coords,
attrs=agg.attrs)
def equal_interval(agg: xr.DataArray,
k: int = 5,
name: Optional[str] = 'equal_interval') -> xr.DataArray:
"""
Groups data for array (agg) by distributing values into at equal intervals.
The result is an xarray.DataArray.
Parameters:
----------
agg: xarray.DataArray
2D array of values to bin.
NumPy, CuPy, NumPy-backed Dask, or Cupy-backed Dask array
k: int
Number of classes to be produced.
name: str, optional (default = "equal_interval")
Name of output aggregate.
Returns:
----------
equal_interval_agg: xarray.DataArray
2D array, of the same type as the input, of class allocations.
Notes:
----------
Intervals defined to have equal width:
Algorithm References:
----------
PySal:
- https://pysal.org/mapclassify/_modules/mapclassify/classifiers.html#EqualInterval # noqa
SciKit:
- https://scikit-learn.org/stable/auto_examples/classification/plot_classifier_comparison.html#sphx-glr-auto-examples-classification-plot-classifier-comparison-py # noqa
Examples:
----------
Imports
>>> import numpy as np
>>> import xarray as xr
>>> from xrspatial.classify import equal_interval, natural_breaks
Create Initial DataArray
>>> np.random.seed(1)
>>> agg = xr.DataArray(np.random.randint(2, 8, (4, 4)),
>>> dims = ["lat", "lon"])
>>> height, width = agg.shape
>>> _lon = np.linspace(0, width - 1, width)
>>> _lat = np.linspace(0, height - 1, height)
>>> agg["lon"] = _lon
>>> agg["lat"] = _lat
>>> print(agg)
<xarray.DataArray (lat: 4, lon: 4)>
array([[7, 5, 6, 2],
[3, 5, 7, 2],
[2, 3, 6, 7],
[6, 3, 4, 6]])
Coordinates:
* lon (lon) float64 0.0 1.0 2.0 3.0
* lat (lat) float64 0.0 1.0 2.0 3.0
Create Equal Interval DataArray
>>> equal_interval_agg = equal_interval(agg, k = 5)
>>> print(equal_interval_agg)
<xarray.DataArray 'equal_interval' (lat: 4, lon: 4)>
array([[4., 2., 3., 0.],
[0., 2., 4., 0.],
[0., 0., 3., 4.],
[3., 0., 1., 3.]], dtype=float32)
Coordinates:
* lon (lon) float64 0.0 1.0 2.0 3.0
* lat (lat) float64 0.0 1.0 2.0 3.0
"""
# numpy case
if isinstance(agg.data, np.ndarray):
out = _run_numpy_equal_interval(agg.data, k)
# cupy case
elif has_cuda() and isinstance(agg.data, cupy.ndarray):
out = _run_cupy_equal_interval(agg.data, k)
# dask + cupy case
elif has_cuda() and \
isinstance(agg.data, cupy.ndarray) and \
is_cupy_backed(agg):
out = _run_dask_cupy_equal_interval(agg.data, k)
# dask + numpy case
elif isinstance(agg.data, da.Array):
out = _run_dask_numpy_equal_interval(agg.data, k)
else:
raise TypeError('Unsupported Array Type: {}'.format(type(agg.data)))
return DataArray(out,
name=name,
coords=agg.coords,
dims=agg.dims,
attrs=agg.attrs)
def generate_terrain(canvas, seed=10, zfactor=4000, full_extent=None):
"""
Generates a pseudo-random terrain which can be helpful for testing raster functions
Parameters
----------
canvas : ds.Canvas instance for passing output dimensions / ranges
seed : seed for random number generator
zfactor : used as multipler for z values
full_extent : optional string, bbox<xmin, ymin, xmax, ymax>
full extent of coordinate system.
Returns
-------
terrain: DataArray
Notes:
------
Algorithm References:
- This was inspired by Michael McHugh's 2016 PyCon Canada talk:
https://www.youtube.com/watch?v=O33YV4ooHSo
- https://www.redblobgames.com/maps/terrain-from-noise/
"""
def _gen_heights(bumps):
out = np.zeros(len(bumps))
for i, b in enumerate(bumps):
x = b[0]
y = b[1]
val = agg.data[y, x]
if val >= 0.33 and val <= 3:
out[i] = 0.1
return out
def _scale(value, old_range, new_range):
return ((value - old_range[0]) /
(old_range[1] - old_range[0])) * (new_range[1] -
new_range[0]) + new_range[0]
if not isinstance(canvas, Canvas):
raise TypeError('canvas must be instance type datashader.Canvas')
mercator_extent = (-np.pi * 6378137, -np.pi * 6378137, np.pi * 6378137,
np.pi * 6378137)
crs_extents = {'3857': mercator_extent}
if isinstance(full_extent, str):
full_extent = crs_extents[full_extent]
elif full_extent is None:
full_extent = (canvas.x_range[0], canvas.y_range[0], canvas.x_range[1],
canvas.y_range[1])
elif not isinstance(full_extent, (list, tuple)) and len(full_extent) != 4:
raise TypeError('full_extent must be tuple(4) or str wkid')
full_xrange = (full_extent[0], full_extent[2])
full_yrange = (full_extent[1], full_extent[3])
x_range_scaled = (_scale(canvas.x_range[0], full_xrange, (0.0, 1.0)),
_scale(canvas.x_range[1], full_xrange, (0.0, 1.0)))
y_range_scaled = (_scale(canvas.y_range[0], full_yrange, (0.0, 1.0)),
_scale(canvas.y_range[1], full_yrange, (0.0, 1.0)))
data = _gen_terrain(canvas.plot_width,
canvas.plot_height,
seed,
x_range=x_range_scaled,
y_range=y_range_scaled)
data = (data - np.min(data)) / np.ptp(data)
data[data < 0.3] = 0 # create water
data *= zfactor
# DataArray coords were coming back different from cvs.points...
hack_agg = canvas.points(pd.DataFrame({'x': [], 'y': []}), 'x', 'y')
agg = DataArray(data,
name='terrain',
coords=hack_agg.coords,
dims=hack_agg.dims,
attrs={'res': 1})
return agg
def test_download_from_github(self, tmp_path) -> None:
cache_dir = tmp_path / tutorial._default_cache_dir_name
ds = tutorial.open_dataset(self.testfile, cache_dir=cache_dir).load()
tiny = DataArray(range(5), name="tiny").to_dataset()
assert_identical(ds, tiny)
def mean(agg, passes=1, excludes=[np.nan], name='mean'):
"""
Returns Mean filtered array using a 3x3 window.
Default behaviour to 'mean' is to exclude NaNs from calculations.
Parameters
----------
agg : xarray.DataArray
2D array of input values to be filtered.
passes : int, default=1
Number of times to run mean.
name : str, default='mean'
Output xr.DataArray.name property.
Returns
-------
mean_agg : xarray.DataArray of same type as `agg`
2D aggregate array of filtered values.
Examples
--------
Focal mean works with NumPy backed xarray DataArray
.. sourcecode:: python
>>> import numpy as np
>>> import xarray as xr
>>> from xrspatial.focal import mean
>>> data = np.array([
[0., 0., 0., 0., 0.],
[0., 0., 0., 0., 0.],
[0., 0., 9., 0., 0.],
[0., 0., 0., 0., 0.],
[0., 0., 0., 0., 0.]])
>>> raster = xr.DataArray(data)
>>> mean_agg = mean(raster)
>>> print(mean_agg)
<xarray.DataArray 'mean' (dim_0: 5, dim_1: 5)>
array([[0., 0., 0., 0., 0.],
[0., 1., 1., 1., 0.],
[0., 1., 1., 1., 0.],
[0., 1., 1., 1., 0.],
[0., 0., 0., 0., 0.]])
Dimensions without coordinates: dim_0, dim_1
Focal mean works with Dask with NumPy backed xarray DataArray.
Increase number of runs by setting a specific value for parameter `passes`
.. sourcecode:: python
>>> import dask.array as da
>>> data_da = da.from_array(data, chunks=(3, 3))
>>> raster_da = xr.DataArray(data_da, dims=['y', 'x'], name='raster_da') # noqa
>>> print(raster_da)
<xarray.DataArray 'raster_da' (y: 5, x: 5)>
dask.array<array, shape=(5, 5), dtype=int64, chunksize=(3, 3), chunktype=numpy.ndarray> # noqa
Dimensions without coordinates: y, x
>>> mean_da = mean(raster_da, passes=2)
>>> print(mean_da)
<xarray.DataArray 'mean' (y: 5, x: 5)>
dask.array<_trim, shape=(5, 5), dtype=float64, chunksize=(3, 3), chunktype=numpy.ndarray> # noqa
Dimensions without coordinates: y, x
>>> print(mean_da.compute())
<xarray.DataArray 'mean' (y: 5, x: 5)>
array([[0.25 , 0.33333333, 0.5 , 0.33333333, 0.25 ],
[0.33333333, 0.44444444, 0.66666667, 0.44444444, 0.33333333],
[0.5 , 0.66666667, 1. , 0.66666667, 0.5 ],
[0.33333333, 0.44444444, 0.66666667, 0.44444444, 0.33333333],
[0.25 , 0.33333333, 0.5 , 0.33333333, 0.25 ]])
Dimensions without coordinates: y, x
Focal mean works with CuPy backed xarray DataArray.
In this example, we set `passes` to the number of elements of the array,
we'll get a mean array where every element has the same value.
.. sourcecode:: python
>>> import cupy
>>> raster_cupy = xr.DataArray(cupy.asarray(data), name='raster_cupy')
>>> mean_cupy = mean(raster_cupy, passes=25)
>>> print(type(mean_cupy.data))
<class 'cupy.core.core.ndarray'>
>>> print(mean_cupy)
<xarray.DataArray 'mean' (dim_0: 5, dim_1: 5)>
array([[0.47928995, 0.47928995, 0.47928995, 0.47928995, 0.47928995],
[0.47928995, 0.47928995, 0.47928995, 0.47928995, 0.47928995],
[0.47928995, 0.47928995, 0.47928995, 0.47928995, 0.47928995],
[0.47928995, 0.47928995, 0.47928995, 0.47928995, 0.47928995],
[0.47928995, 0.47928995, 0.47928995, 0.47928995, 0.47928995]])
Dimensions without coordinates: dim_0, dim_1
"""
out = agg.data.astype(float)
for i in range(passes):
out = _mean(out, tuple(excludes))
return DataArray(out,
name=name,
dims=agg.dims,
coords=agg.coords,
attrs=agg.attrs)
def get_test_content(self, filename, filename_info, filetype_info):
"""Mimic reader input file content"""
start_time = filename_info['start_time']
end_time = filename_info['end_time'].replace(year=start_time.year,
month=start_time.month,
day=start_time.day)
prefix1 = 'Data_Products/{file_group}'.format(**filetype_info)
prefix2 = '{prefix}/{file_group}_Aggr'.format(prefix=prefix1,
**filetype_info)
prefix3 = 'All_Data/{file_group}_All'.format(**filetype_info)
begin_date = start_time.strftime('%Y%m%d')
begin_time = start_time.strftime('%H%M%S.%fZ')
ending_date = end_time.strftime('%Y%m%d')
ending_time = end_time.strftime('%H%M%S.%fZ')
if filename[:3] == 'SVI':
geo_prefix = 'GIMGO'
elif filename[:3] == 'SVM':
geo_prefix = 'GMODO'
else:
geo_prefix = None
file_content = {
"{prefix2}/attr/AggregateBeginningDate":
begin_date,
"{prefix2}/attr/AggregateBeginningTime":
begin_time,
"{prefix2}/attr/AggregateEndingDate":
ending_date,
"{prefix2}/attr/AggregateEndingTime":
ending_time,
"{prefix2}/attr/G-Ring_Longitude":
np.array([0.0, 0.1, 0.2, 0.3]),
"{prefix2}/attr/G-Ring_Latitude":
np.array([0.0, 0.1, 0.2, 0.3]),
"{prefix2}/attr/AggregateBeginningOrbitNumber":
"{0:d}".format(filename_info['orbit']),
"{prefix2}/attr/AggregateEndingOrbitNumber":
"{0:d}".format(filename_info['orbit']),
"{prefix1}/attr/Instrument_Short_Name":
"VIIRS",
"/attr/Platform_Short_Name":
"NPP",
}
if geo_prefix:
file_content['/attr/N_GEO_Ref'] = geo_prefix + filename[5:]
for k, v in list(file_content.items()):
file_content[k.format(prefix1=prefix1, prefix2=prefix2)] = v
if filename[:3] in ['SVM', 'SVI', 'SVD']:
if filename[2:5] in ['M{:02d}'.format(x)
for x in range(12)] + ['I01', 'I02', 'I03']:
keys = ['Radiance', 'Reflectance']
elif filename[2:5] in ['M{:02d}'.format(x)
for x in range(12, 17)] + ['I04', 'I05']:
keys = ['Radiance', 'BrightnessTemperature']
else:
# DNB
keys = ['Radiance']
for k in keys:
k = prefix3 + "/" + k
file_content[k] = DEFAULT_FILE_DATA.copy()
file_content[k + "/shape"] = DEFAULT_FILE_SHAPE
file_content[k + "Factors"] = DEFAULT_FILE_FACTORS.copy()
elif filename[0] == 'G':
if filename[:5] in ['GMODO', 'GIMGO']:
lon_data = np.linspace(
15, 55, DEFAULT_FILE_SHAPE[1]).astype(DEFAULT_FILE_DTYPE)
lat_data = np.linspace(
55, 75, DEFAULT_FILE_SHAPE[1]).astype(DEFAULT_FILE_DTYPE)
else:
lon_data = np.linspace(
5, 45, DEFAULT_FILE_SHAPE[1]).astype(DEFAULT_FILE_DTYPE)
lat_data = np.linspace(
45, 65, DEFAULT_FILE_SHAPE[1]).astype(DEFAULT_FILE_DTYPE)
for k in ["Latitude"]:
k = prefix3 + "/" + k
file_content[k] = lat_data
file_content[k] = np.repeat([file_content[k]],
DEFAULT_FILE_SHAPE[0],
axis=0)
file_content[k + "/shape"] = DEFAULT_FILE_SHAPE
for k in ["Longitude"]:
k = prefix3 + "/" + k
file_content[k] = lon_data
file_content[k] = np.repeat([file_content[k]],
DEFAULT_FILE_SHAPE[0],
axis=0)
file_content[k + "/shape"] = DEFAULT_FILE_SHAPE
# convert to xarrays
from xarray import DataArray
import dask.array as da
for key, val in file_content.items():
if isinstance(val, np.ndarray):
val = da.from_array(val, chunks=val.shape)
if val.ndim > 1:
file_content[key] = DataArray(val, dims=('y', 'x'))
else:
file_content[key] = DataArray(val)
return file_content
def hotspots(raster, kernel):
"""
Identify statistically significant hot spots and cold spots in an
input raster. To be a statistically significant hot spot, a feature
will have a high value and be surrounded by other features with
high values as well.
Neighborhood of a feature defined by the input kernel, which
currently support a shape of circle, annulus, or custom kernel.
The result should be a raster with the following 7 values:
- 90 for 90% confidence high value cluster
- 95 for 95% confidence high value cluster
- 99 for 99% confidence high value cluster
- 90 for 90% confidence low value cluster
- 95 for 95% confidence low value cluster
- 99 for 99% confidence low value cluster
- 0 for no significance
Parameters
----------
raster : xarray.DataArray
2D Input raster image with `raster.shape` = (height, width).
Can be a NumPy backed, CuPy backed, or Dask with NumPy backed DataArray
kernel : Numpy Array
2D array where values of 1 indicate the kernel.
Returns
-------
hotspots_agg : xarray.DataArray of same type as `raster`
2D array of hotspots with values indicating confidence level.
Examples
--------
.. sourcecode:: python
>>> import numpy as np
>>> import xarray as xr
>>> from xrspatial.convolution import custom_kernel
>>> kernel = custom_kernel(np.array([[1, 1, 0]]))
>>> data = np.array([
... [0, 1000, 1000, 0, 0, 0],
... [0, 0, 0, -1000, -1000, 0],
... [0, -900, -900, 0, 0, 0],
... [0, 100, 1000, 0, 0, 0]])
>>> from xrspatial.focal import hotspots
>>> hotspots(xr.DataArray(data), kernel)
array([[ 0, 0, 95, 0, 0, 0],
[ 0, 0, 0, 0, -90, 0],
[ 0, 0, -90, 0, 0, 0],
[ 0, 0, 0, 0, 0, 0]], dtype=int8)
Dimensions without coordinates: dim_0, dim_1
"""
# validate raster
if not isinstance(raster, DataArray):
raise TypeError("`raster` must be instance of DataArray")
if raster.ndim != 2:
raise ValueError("`raster` must be 2D")
mapper = ArrayTypeFunctionMapping(
numpy_func=_hotspots_numpy,
cupy_func=_hotspots_cupy,
dask_func=_hotspots_dask_numpy,
dask_cupy_func=lambda *args: not_implemented_func(
*args,
messages=
'hotspots() does not support dask with cupy backed DataArray.'
), # noqa
)
out = mapper(raster)(raster, kernel)
attrs = copy.deepcopy(raster.attrs)
attrs['unit'] = '%'
return DataArray(out, coords=raster.coords, dims=raster.dims, attrs=attrs)
def from_series_or_scalar(se):
try:
return DataArray.from_series(se)
except AttributeError: # scalar case
return DataArray(se)
def raster(self,
source,
band=1,
upsample_method='linear',
downsample_method='mean'):
"""Sample a raster dataset by canvas size and bounds. Note: requires
`rasterio`. Missing values (those having the value indicated by the
"nodata" attribute of the raster) are replaced with `NaN` if floats, and
0 if int.
Parameters
----------
source : xarray.DataArray
input datasource most likely obtain from `xr.open_rasterio()`.
band : int (unused)
source band number : optional default=1. Not yet implemented.
upsample_method : str, optional default=linear
resample mode when upsampling raster.
options include: nearest, linear.
downsample_method : str, optional default=mean
resample mode when downsampling raster.
options include: first, last, mean, mode, var, std
Returns
-------
data : xarray.Dataset
Notes
-------
requires `rasterio`.
"""
try:
import rasterio as rio
except ImportError:
raise ImportError('install rasterio to use this feature')
upsample_methods = dict(nearest=US_NEAREST, linear=US_LINEAR)
downsample_methods = dict(first=DS_FIRST,
last=DS_LAST,
mean=DS_MEAN,
mode=DS_MODE,
var=DS_VAR,
std=DS_STD)
if upsample_method not in upsample_methods.keys():
raise ValueError(
'Invalid upsample method: options include {}'.format(
list(upsample_methods.keys())))
if downsample_method not in downsample_methods.keys():
raise ValueError(
'Invalid downsample method: options include {}'.format(
list(downsample_methods.keys())))
res = calc_res(source)
left, bottom, right, top = calc_bbox(source.x.values, source.y.values,
res)
# window coodinates
xmin = max(self.x_range[0], left)
ymin = max(self.y_range[0], bottom)
xmax = min(self.x_range[1], right)
ymax = min(self.y_range[1], top)
width_ratio = (xmax - xmin) / (self.x_range[1] - self.x_range[0])
height_ratio = (ymax - ymin) / (self.y_range[1] - self.y_range[0])
if np.isclose(width_ratio, 0) or np.isclose(height_ratio, 0):
raise ValueError(
'Canvas x_range or y_range values do not match closely-enough with the data source to be able to accurately rasterize. Please provide ranges that are more accurate.'
)
w = int(np.ceil(self.plot_width * width_ratio))
h = int(np.ceil(self.plot_height * height_ratio))
data = resample_2d(source.values[0].astype(np.float32),
w,
h,
ds_method=downsample_methods[downsample_method],
us_method=upsample_methods[upsample_method])
if w != self.plot_width or h != self.plot_height:
num_height = self.plot_height - h
num_width = self.plot_width - w
lpad = xmin - self.x_range[0]
rpad = self.x_range[1] - xmax
lpct = lpad / (lpad + rpad) if lpad + rpad > 0 else 0
left = int(np.ceil(num_width * lpct))
right = num_width - left
left_pad = np.empty(shape=(self.plot_height, left)).astype(
source.dtype) * np.nan
right_pad = np.empty(shape=(self.plot_height, right)).astype(
source.dtype) * np.nan
tpad = ymin - self.y_range[0]
bpad = self.y_range[1] - ymax
tpct = tpad / (tpad + bpad) if tpad + bpad > 0 else 0
top = int(np.ceil(num_height * tpct))
bottom = num_height - top
top_pad = np.empty(shape=(top, w)).astype(source.dtype) * np.nan
bottom_pad = np.empty(shape=(bottom, w)).astype(
source.dtype) * np.nan
data = np.concatenate((bottom_pad, data, top_pad), axis=0)
data = np.concatenate((left_pad, data, right_pad), axis=1)
data = np.flipud(data)
attrs = dict(res=res[0], nodata=source._file_obj.nodata)
return DataArray(data, dims=['x', 'y'], attrs=attrs)
def test_cftime_datetime_mean_dask_error():
times = cftime_range('2000', periods=4)
da = DataArray(times, dims=['time']).chunk()
with pytest.raises(NotImplementedError):
da.mean()
def test_download_from_github(self):
ds = tutorial.open_dataset(self.testfile).load()
tiny = DataArray(range(5), name="tiny").to_dataset()
assert_identical(ds, tiny)
def bump(width: int,
height: int,
count: Optional[int] = None,
height_func=None,
spread: int = 1) -> xr.DataArray:
"""
Generate a simple bump map to simulate the appearance of land
features.
Using a user-defined height function, determines at what elevation
a specific bump height is acceptable. Bumps of number `count` are
applied over the area `width` x `height`.
Parameters
----------
width : int
Total width, in pixels, of the image.
height : int
Total height, in pixels, of the image.
count : int
Number of bumps to generate.
height_func : function which takes x, y and returns a height value
Function used to apply varying bump heights to different
elevations.
spread : int, default=1
Number of pixels to spread on all sides.
Returns
-------
bump_agg : xarray.DataArray
2D aggregate array of calculated bump heights.
References
----------
- ICA: http://www.mountaincartography.org/mt_hood/pdfs/nighbert_bump1.pdf # noqa
Examples
--------
.. plot::
:include-source:
from functools import partial
import matplotlib.pyplot as plt
import numpy as np
import xarray as xr
from xrspatial import generate_terrain, bump
# Generate Example Terrain
W = 500
H = 300
template_terrain = xr.DataArray(np.zeros((H, W)))
x_range=(-20e6, 20e6)
y_range=(-20e6, 20e6)
terrain_agg = generate_terrain(
template_terrain, x_range=x_range, y_range=y_range
)
# Edit Attributes
terrain_agg = terrain_agg.assign_attrs(
{
'Description': 'Example Terrain',
'units': 'km',
'Max Elevation': '4000',
}
)
terrain_agg = terrain_agg.rename({'x': 'lon', 'y': 'lat'})
terrain_agg = terrain_agg.rename('Elevation')
# Create Height Function
def heights(locations, src, src_range, height = 20):
num_bumps = locations.shape[0]
out = np.zeros(num_bumps, dtype = np.uint16)
for r in range(0, num_bumps):
loc = locations[r]
x = loc[0]
y = loc[1]
val = src[y, x]
if val >= src_range[0] and val < src_range[1]:
out[r] = height
return out
# Create Bump Map Aggregate Array
bump_count = 10000
src = terrain_agg.data
# Short Bumps from z = 1000 to z = 1300
bump_agg = bump(width = W, height = H, count = bump_count,
height_func = partial(heights, src = src,
src_range = (1000, 1300),
height = 5))
# Tall Bumps from z = 1300 to z = 1700
bump_agg += bump(width = W, height = H, count = bump_count // 2,
height_func = partial(heights, src = src,
src_range = (1300, 1700),
height=20))
# Short Bumps from z = 1700 to z = 2000
bump_agg += bump(width = W, height = H, count = bump_count // 3,
height_func = partial(heights, src = src,
src_range = (1700, 2000),
height=5))
# Edit Attributes
bump_agg = bump_agg.assign_attrs({'Description': 'Example Bump Map',
'units': 'km'})
bump_agg = bump_agg.rename('Bump Height')
# Rename Coordinates
bump_agg = bump_agg.assign_coords({'x': terrain_agg.coords['lon'].data,
'y': terrain_agg.coords['lat'].data})
# Remove zeros
bump_agg.data[bump_agg.data == 0] = np.nan
# Plot Terrain
terrain_agg.plot(cmap = 'terrain', aspect = 2, size = 4)
plt.title("Terrain")
plt.ylabel("latitude")
plt.xlabel("longitude")
# Plot Bump Map
bump_agg.plot(cmap = 'summer', aspect = 2, size = 4)
plt.title("Bump Map")
plt.ylabel("latitude")
plt.xlabel("longitude")
.. sourcecode:: python
>>> print(terrain_agg[200:203, 200:202])
<xarray.DataArray 'Elevation' (lat: 3, lon: 2)>
array([[1264.02296597, 1261.947921 ],
[1285.37105519, 1282.48079719],
[1306.02339636, 1303.4069579 ]])
Coordinates:
* lon (lon) float64 -3.96e+06 -3.88e+06
* lat (lat) float64 6.733e+06 6.867e+06 7e+06
Attributes:
res: (80000.0, 133333.3333333333)
Description: Example Terrain
units: km
Max Elevation: 4000
.. sourcecode:: python
>>> print(bump_agg[200:205, 200:206])
<xarray.DataArray 'Bump Height' (y: 5, x: 6)>
array([[nan, nan, nan, nan, 5., 5.],
[nan, nan, nan, nan, nan, 5.],
[nan, nan, nan, nan, nan, nan],
[nan, nan, nan, nan, nan, nan],
[nan, nan, nan, nan, nan, nan]])
Coordinates:
* x (x) float64 -3.96e+06 -3.88e+06 -3.8e+06 ... -3.64e+06 -3.56e+06
* y (y) float64 6.733e+06 6.867e+06 7e+06 7.133e+06 7.267e+06
Attributes:
res: 1
Description: Example Bump Map
units: km
"""
linx = range(width)
liny = range(height)
if count is None:
count = width * height // 10
if height_func is None:
height_func = lambda bumps: np.ones(len(bumps)) # noqa
# create 2d array of random x, y for bump locations
locs = np.empty((count, 2), dtype=np.uint16)
locs[:, 0] = np.random.choice(linx, count)
locs[:, 1] = np.random.choice(liny, count)
heights = height_func(locs)
bumps = _finish_bump(width, height, locs, heights, spread)
return DataArray(bumps, dims=['y', 'x'], attrs=dict(res=1))
def get_test_content(self, filename, filename_info, filetype_info):
"""Mimic reader input file content"""
file_content = {
'/attr/time_coverage_start':
filename_info['start_time'].strftime('%Y-%m-%dT%H:%M:%S.%fZ'),
'/attr/time_coverage_end':
filename_info['end_time'].strftime('%Y-%m-%dT%H:%M:%S.%fZ'),
'/attr/start_orbit_number':
1,
'/attr/end_orbit_number':
2,
'/attr/platform_name':
'NPP',
'/attr/instrument_name':
'CrIS, ATMS, VIIRS',
}
for k, units, standard_name in [
('Solar_Zenith', 'degrees', 'solar_zenith_angle'),
('Topography', 'meters', ''),
('Land_Fraction', '1', ''),
('Surface_Pressure', 'mb', ''),
('Skin_Temperature', 'Kelvin', 'surface_temperature'),
]:
file_content[k] = DEFAULT_FILE_DATA
file_content[k + '/shape'] = DEFAULT_FILE_SHAPE
file_content[k + '/attr/units'] = units
file_content[k + '/attr/valid_range'] = (0., 120.)
file_content[k + '/attr/_FillValue'] = -9999.
if standard_name:
file_content[k + '/attr/standard_name'] = standard_name
for k, units, standard_name in [
('Temperature', 'Kelvin', 'air_temperature'),
('H2O', '1', ''),
('H2O_MR', 'g/g', ''),
('O3', '1', ''),
('O3_MR', '1', ''),
('Liquid_H2O', '1', ''),
('Liquid_H2O_MR', 'g/g', 'cloud_liquid_water_mixing_ratio'),
('CO', '1', ''),
('CO_MR', '1', ''),
('CH4', '1', ''),
('CH4_MR', '1', ''),
('CO2', '1', ''),
('HNO3', '1', ''),
('HNO3_MR', '1', ''),
('N2O', '1', ''),
('N2O_MR', '1', ''),
('SO2', '1', ''),
('SO2_MR', '1', ''),
]:
file_content[k] = DEFAULT_PRES_FILE_DATA
file_content[k + '/shape'] = DEFAULT_PRES_FILE_SHAPE
file_content[k + '/attr/units'] = units
file_content[k + '/attr/valid_range'] = (0., 120.)
file_content[k + '/attr/_FillValue'] = -9999.
if standard_name:
file_content[k + '/attr/standard_name'] = standard_name
k = 'Pressure'
file_content[k] = ALL_PRESSURE_LEVELS
file_content[k + '/shape'] = DEFAULT_PRES_FILE_SHAPE
file_content[k + '/attr/units'] = 'mb'
file_content[k + '/attr/valid_range'] = (0., 2000.)
file_content[k + '/attr/_FillValue'] = -9999.
k = 'Quality_Flag'
file_content[k] = DEFAULT_FILE_DATA.astype(np.int32)
file_content[k + '/shape'] = DEFAULT_FILE_SHAPE
file_content[k + '/attr/valid_range'] = (0, 31)
file_content[k + '/attr/_FillValue'] = -9999.
k = 'Longitude'
file_content[k] = DEFAULT_LON_DATA
file_content[k + '/shape'] = DEFAULT_FILE_SHAPE
file_content[k + '/attr/units'] = 'degrees_east'
file_content[k + '/attr/valid_range'] = (-180., 180.)
file_content[k + '/attr/standard_name'] = 'longitude'
file_content[k + '/attr/_FillValue'] = -9999.
k = 'Latitude'
file_content[k] = DEFAULT_LAT_DATA
file_content[k + '/shape'] = DEFAULT_FILE_SHAPE
file_content[k + '/attr/units'] = 'degrees_north'
file_content[k + '/attr/valid_range'] = (-90., 90.)
file_content[k + '/attr/standard_name'] = 'latitude'
file_content[k + '/attr/_FillValue'] = -9999.
# convert to xarrays
from xarray import DataArray
for key, val in file_content.items():
if isinstance(val, np.ndarray):
attrs = {}
for a in [
'_FillValue', 'flag_meanings', 'flag_values', 'units'
]:
if key + '/attr/' + a in file_content:
attrs[a] = file_content[key + '/attr/' + a]
if val.ndim == 1:
file_content[key] = DataArray(val,
dims=('number_of_FORs', ),
attrs=attrs)
elif val.ndim > 1:
file_content[key] = DataArray(val,
dims=('number_of_FORs',
'number_of_p_levels'),
attrs=attrs)
else:
file_content[key] = DataArray(val, attrs=attrs)
return file_content
def raster(
self,
source,
layer=None,
upsample_method='linear', # Deprecated as of datashader=0.6.4
downsample_method=rd.mean(), # Deprecated as of datashader=0.6.4
nan_value=None,
agg=None,
interpolate=None,
chunksize=None,
max_mem=None):
"""Sample a raster dataset by canvas size and bounds.
Handles 2D or 3D xarray DataArrays, assuming that the last two
array dimensions are the y- and x-axis that are to be
resampled. If a 3D array is supplied a layer may be specified
to resample to select the layer along the first dimension to
resample.
Missing values (those having the value indicated by the
"nodata" attribute of the raster) are replaced with `NaN` if
floats, and 0 if int.
Also supports resampling out-of-core DataArrays backed by dask
Arrays. By default it will try to maintain the same chunksize
in the output array but a custom chunksize may be provided.
If there are memory constraints they may be defined using the
max_mem parameter, which determines how large the chunks in
memory may be.
Parameters
----------
source : xarray.DataArray or xr.Dataset
2D or 3D labelled array (if Dataset, the agg reduction must
define the data variable).
layer : float
For a 3D array, value along the z dimension : optional default=None
ds_method : str (optional)
Grid cell aggregation method for a possible downsampling.
us_method : str (optional)
Grid cell interpolation method for a possible upsampling.
nan_value : int or float, optional
Optional nan_value which will be masked out when applying
the resampling.
agg : Reduction, optional default=mean()
Resampling mode when downsampling raster.
options include: first, last, mean, mode, var, std, min, max
Accepts an executable function, function object, or string name.
interpolate : str, optional default=linear
Resampling mode when upsampling raster.
options include: nearest, linear.
chunksize : tuple(int, int) (optional)
Size of the output chunks. By default this the chunk size is
inherited from the *src* array.
max_mem : int (optional)
The maximum number of bytes that should be loaded into memory
during the regridding operation.
Returns
-------
data : xarray.Dataset
"""
# For backwards compatibility
if agg is None: agg = downsample_method
if interpolate is None: interpolate = upsample_method
upsample_methods = ['nearest', 'linear']
downsample_methods = {
'first': 'first',
rd.first: 'first',
'last': 'last',
rd.last: 'last',
'mode': 'mode',
rd.mode: 'mode',
'mean': 'mean',
rd.mean: 'mean',
'var': 'var',
rd.var: 'var',
'std': 'std',
rd.std: 'std',
'min': 'min',
rd.min: 'min',
'max': 'max',
rd.max: 'max'
}
if interpolate not in upsample_methods:
raise ValueError(
'Invalid interpolate method: options include {}'.format(
upsample_methods))
if not isinstance(source, (DataArray, Dataset)):
raise ValueError('Expected xarray DataArray or Dataset as '
'the data source, found %s.' %
type(source).__name__)
column = None
if isinstance(agg, rd.Reduction):
agg, column = type(agg), agg.column
if (isinstance(source, DataArray) and column is not None
and source.name != column):
agg_repr = '%s(%r)' % (agg.__name__, column)
raise ValueError('DataArray name %r does not match '
'supplied reduction %s.' %
(source.name, agg_repr))
if isinstance(source, Dataset):
data_vars = list(source.data_vars)
if column is None:
raise ValueError('When supplying a Dataset the agg reduction '
'must specify the variable to aggregate. '
'Available data_vars include: %r.' %
data_vars)
elif column not in source.data_vars:
raise KeyError('Supplied reduction column %r not found '
'in Dataset, expected one of the following '
'data variables: %r.' % (column, data_vars))
source = source[column]
if agg not in downsample_methods.keys():
raise ValueError(
'Invalid aggregation method: options include {}'.format(
list(downsample_methods.keys())))
ds_method = downsample_methods[agg]
if source.ndim not in [2, 3]:
raise ValueError('Raster aggregation expects a 2D or 3D '
'DataArray, found %s dimensions' % source.ndim)
res = calc_res(source)
ydim, xdim = source.dims[-2:]
xvals, yvals = source[xdim].values, source[ydim].values
left, bottom, right, top = calc_bbox(xvals, yvals, res)
if layer is not None:
source = source.sel(**{source.dims[0]: layer})
array = orient_array(source, res)
dtype = array.dtype
if nan_value is not None:
mask = array == nan_value
array = np.ma.masked_array(array, mask=mask, fill_value=nan_value)
fill_value = nan_value
else:
fill_value = np.NaN
if self.x_range is None: self.x_range = (left, right)
if self.y_range is None: self.y_range = (bottom, top)
# window coordinates
xmin = max(self.x_range[0], left)
ymin = max(self.y_range[0], bottom)
xmax = min(self.x_range[1], right)
ymax = min(self.y_range[1], top)
width_ratio = min((xmax - xmin) / (self.x_range[1] - self.x_range[0]),
1)
height_ratio = min((ymax - ymin) / (self.y_range[1] - self.y_range[0]),
1)
if np.isclose(width_ratio, 0) or np.isclose(height_ratio, 0):
raise ValueError(
'Canvas x_range or y_range values do not match closely enough with the data source to be able to accurately rasterize. Please provide ranges that are more accurate.'
)
w = max(int(round(self.plot_width * width_ratio)), 1)
h = max(int(round(self.plot_height * height_ratio)), 1)
cmin, cmax = get_indices(xmin, xmax, xvals, res[0])
rmin, rmax = get_indices(ymin, ymax, yvals, res[1])
kwargs = dict(w=w,
h=h,
ds_method=ds_method,
us_method=interpolate,
fill_value=fill_value)
if array.ndim == 2:
source_window = array[rmin:rmax + 1, cmin:cmax + 1]
if ds_method in ['var', 'std']:
source_window = source_window.astype('f')
if isinstance(source_window, da.Array):
data = resample_2d_distributed(source_window,
chunksize=chunksize,
max_mem=max_mem,
**kwargs)
else:
data = resample_2d(source_window, **kwargs)
layers = 1
else:
source_window = array[:, rmin:rmax + 1, cmin:cmax + 1]
if ds_method in ['var', 'std']:
source_window = source_window.astype('f')
arrays = []
for arr in source_window:
if isinstance(arr, da.Array):
arr = resample_2d_distributed(arr,
chunksize=chunksize,
max_mem=max_mem,
**kwargs)
else:
arr = resample_2d(arr, **kwargs)
arrays.append(arr)
data = np.dstack(arrays)
layers = len(arrays)
if w != self.plot_width or h != self.plot_height:
num_height = self.plot_height - h
num_width = self.plot_width - w
lpad = xmin - self.x_range[0]
rpad = self.x_range[1] - xmax
lpct = lpad / (lpad + rpad) if lpad + rpad > 0 else 0
left = max(int(np.ceil(num_width * lpct)), 0)
right = max(num_width - left, 0)
lshape, rshape = (self.plot_height, left), (self.plot_height,
right)
if layers > 1:
lshape, rshape = lshape + (layers, ), rshape + (layers, )
left_pad = np.full(lshape, fill_value, source_window.dtype)
right_pad = np.full(rshape, fill_value, source_window.dtype)
tpad = ymin - self.y_range[0]
bpad = self.y_range[1] - ymax
tpct = tpad / (tpad + bpad) if tpad + bpad > 0 else 0
top = max(int(np.ceil(num_height * tpct)), 0)
bottom = max(num_height - top, 0)
tshape, bshape = (top, w), (bottom, w)
if layers > 1:
tshape, bshape = tshape + (layers, ), bshape + (layers, )
top_pad = np.full(tshape, fill_value, source_window.dtype)
bottom_pad = np.full(bshape, fill_value, source_window.dtype)
concat = da.concatenate if isinstance(data,
da.Array) else np.concatenate
if top_pad.shape[0] > 0:
data = concat((top_pad, data, bottom_pad), axis=0)
if left_pad.shape[1] > 0:
data = concat((left_pad, data, right_pad), axis=1)
# Reorient array to original orientation
if res[1] > 0: data = data[::-1]
if res[0] < 0: data = data[:, ::-1]
# Restore nan_value from masked array
if nan_value is not None:
data = data.filled()
# Restore original dtype
if dtype != data.dtype:
data = data.astype(dtype)
# Compute DataArray metadata
xs, ys = compute_coords(self.plot_width, self.plot_height,
self.x_range, self.y_range, res)
coords = {xdim: xs, ydim: ys}
dims = [ydim, xdim]
attrs = dict(res=res[0])
if source._file_obj is not None and hasattr(source._file_obj,
'nodata'):
attrs['nodata'] = source._file_obj.nodata
# Handle DataArray with layers
if data.ndim == 3:
data = data.transpose([2, 0, 1])
layer_dim = source.dims[0]
coords[layer_dim] = source.coords[layer_dim]
dims = [layer_dim] + dims
return DataArray(data, coords=coords, dims=dims, attrs=attrs)
def manhattan(data, colora="#5689AC", colorb="#21334F",
anno_pv_max=None, pts_kws=None, ax=None):
"""
Produce a manhattan plot.
Parameters
----------
data : DataFrame, dict
DataFrame containing the chromosome, base-pair positions, and
p-values.
colora : matplotlib color
Points color of the first group.
colorb : matplotlib color
Points color of the second group.
anno_pv_max : float
Threshold of maximum p value for annotating data['id'] in the plot.
pts_kws : dict, optional
Keyword arguments forwarded to the matplotlib function used for
plotting the points.
ax : matplotlib Axes, optional
The target handle for this figure. If ``None``, the current axes is
set.
Example
-------
.. plot::
>>> import limix_plot as lp
>>> from numpy import log10
>>>
>>> df = lp.load_dataset('gwas')
>>> df = df.rename(columns={"chr": "chrom"})
>>> print(df.head())
chrom pos pv
234 10 224239 0.00887
239 10 229681 0.00848
253 10 240788 0.00721
258 10 246933 0.00568
266 10 255222 0.00593
>>> lp.manhattan(df)
>>> plt = lp.get_pyplot()
>>> _ = plt.axhline(-log10(1e-7), color='red')
>>> _ = plt.ylim(2, plt.ylim()[1])
"""
from numpy import log10, unique, where
from xarray import DataArray
import pandas as pd
plt = get_pyplot()
if isinstance(data, pd.DataFrame):
data = DataArray(
data=data["pv"],
dims=["candidate"],
coords={k: ("candidate", data[k]) for k in data.columns},
)
else:
data = DataArray(data=data)
if len(data) == 0:
raise ValueError("DataFrame is empty.")
if pts_kws is None:
pts_kws = dict()
ax = plt.gca() if ax is None else ax
data["chrom"] = data["chrom"].astype(str)
data["pos"] = data["pos"].astype(int)
chr_order = _chr_precedence(data)
data["order"] = ("candidate", [chr_order[i] for i in data["chrom"].values])
data = data.sortby(["order", "pos"])
data = _abs_pos(data)
if "markersize" not in pts_kws:
pts_kws["markersize"] = 2
if "marker" not in pts_kws:
pts_kws["marker"] = "."
if "linestyle" not in pts_kws:
pts_kws["linestyle"] = ""
colors = {0: colora, 1: colorb}
for i, c in enumerate(unique(data["order"])):
ok = data["order"] == c
pts_kws["color"] = colors[i % 2]
x = data.loc[ok]["abs_pos"]
y = -log10(data.loc[ok].values)
ax.plot(x, y, **pts_kws)
if anno_pv_max is not None:
_idx = where(y > -log10(anno_pv_max))[0]
for _ii in _idx:
if 'id' in data.coords:
_txt = data['id'].loc[ok].loc[_ii].values
else:
_txt = (str(data['chrom'].loc[ok].loc[_ii].values) + "_" +
str(data['pos'].loc[ok].loc[_ii].values))
ax.annotate(_txt, (x[_ii], y[_ii]), ha='center')
ax.set_xlim(data["abs_pos"].min(), data["abs_pos"].max())
ax.set_ylim(0, ax.get_ylim()[1])
ax.set_ylabel("-log$_{10}$pv")
ax.set_xlabel("chromosome")
u = unique(data["chrom"].values)
chrom_labels = sorted(u, key=lambda x: chr_order[x])
_set_ticks(ax, _chrom_bounds(data), chrom_labels)
def test_nonnumeric_index_raises_typeerror(self):
a = DataArray([1, 2, 3], {'letter': ['a', 'b', 'c']}, dims='letter')
with raises_regex(TypeError, r'[Pp]lot'):
a.plot.line()
