def normalized_power(arg, mask=None, value=0.0, **kwargs):
"""Normalized power
Parameters
----------
arg : array-like
Power stream
mask: array-like of bool, optional
default=None, which means no masking
value : number, optional
Value to use for replacement, default=0.0
type : {"xPower", "NP}
Determines calculation method to use, default='xPower'
Returns
-------
number
"""
if kwargs.get('type', 'NP') == 'xPower':
_rolling_mean = rolling_mean(arg, window=25, mask=mask, value=value, type='emwa')
else:
_rolling_mean = rolling_mean(arg, window=30, mask=mask, value=value)
if type(_rolling_mean) == list:
_rolling_mean = np.asarray(_rolling_mean, dtype=np.float)
rv = np.mean(np.power(_rolling_mean, 4)) ** (1/4)
return rv
def normalized_power(arg, mask=None, value=0.0, **kwargs):
"""Normalized power
Parameters
----------
arg : array-like
Power stream
mask: array-like of bool, optional
default=None, which means no masking
value : number, optional
Value to use for replacement, default=0.0
type : {"xPower", "NP}
Determines calculation method to use, default='xPower'
Returns
-------
number
"""
if kwargs.get('type', 'NP') == 'xPower':
_rolling_mean = rolling_mean(arg,
window=25,
mask=mask,
value=value,
type='emwa')
else:
_rolling_mean = rolling_mean(arg, window=30, mask=mask, value=value)
if type(_rolling_mean) == list:
_rolling_mean = np.asarray(_rolling_mean, dtype=np.float)
rv = np.mean(np.power(_rolling_mean, 4))**(1 / 4)
return rv
def best_interval(arg, window, mask=None, value=0.0, **kwargs):
"""Compute best interval of the stream
Masking with replacement is controlled by keyword arguments
Parameters
----------
arg: array-like
window : int
Duration of the interval in seconds
mask : array-like of bool, optional
default=None, which means no masking
value : number, optional
Value to use for replacement, default=0.0
Returns
-------
float
"""
y = rolling_mean(arg, window=window, mask=mask, value=value, **kwargs)
rv = np.max(y)
return rv
def best_interval(arg, window, mask=None, value=0.0, **kwargs):
"""Compute best interval of the stream
Masking with replacement is controlled by keyword arguments
Parameters
----------
arg: array-like
window : int
Duration of the interval in seconds
mask : array-like of bool, optional
default=None, which means no masking
value : number, optional
Value to use for replacement, default=0.0
Returns
-------
float
"""
y = rolling_mean(arg, window=window, mask=mask, value=value, **kwargs)
rv = np.max(y)
return rv
def test_rolling_mean_list():
stream = [1, 2, 3, 4, 5]
expected = [1, 1.5, 2.5, 3.5, 4.5]
rv = streams.rolling_mean(stream, 2)
assert type(rv) == list
assert rv == expected
def test_rolling_mean_real_data(test_stream):
rv = streams.rolling_mean(test_stream['watts'],
mask=test_stream['moving'],
window=1)
assert type(rv) == list
assert rv == test_stream['watts']
def test_rolling_mean_list_emwa():
stream = list(np.ones(30))
expected = list(np.ones(30))
rv = streams.rolling_mean(stream, 2, type='ewma')
assert type(rv) == list
assert rv == expected
def test_rolling_mean_list():
stream = [1, 2, 3, 4, 5]
expected = [1, 1.5, 2.5, 3.5, 4.5]
rv = streams.rolling_mean(stream, 2)
assert type(rv) == list
assert rv == expected
def test_rolling_mean_real_data(test_stream):
rv = streams.rolling_mean(test_stream['watts'],
mask=test_stream['moving'],
window=1)
assert type(rv) == list
assert rv == test_stream['watts']
def test_rolling_mean_list_emwa():
stream = list(np.ones(30))
expected = list(np.ones(30))
rv = streams.rolling_mean(stream, 2, type='ewma')
assert type(rv) == list
assert rv == expected
def test_rolling_mean_list_with_mask():
stream = [1, 2, 3, 4, 5]
mask = [True, True, False, True, True]
expected = [1, 1.5, 1.0, 2.0, 4.5]
rv = streams.rolling_mean(stream, window=2, mask=mask)
assert type(rv) == list
assert rv == expected
def test_rolling_mean_ndarray():
stream = np.asarray([1, 2, 3, 4, 5])
expected = np.asarray([1, 1.5, 2.5, 3.5, 4.5])
rv = streams.rolling_mean(stream, 2)
assert type(rv) == np.ndarray
assert (rv == expected).all()
def test_rolling_mean_with_mask_ndarray():
stream = np.asarray([1, 2, 3, 4, 5])
mask = np.asarray([True, True, False, True, True], dtype=bool)
expected = np.asarray([1, 1.5, 1.0, 2.0, 4.5])
rv = streams.rolling_mean(stream, window=2, mask=mask)
assert type(rv) == np.ndarray
assert (rv == expected).all()
def test_rolling_mean_list_with_mask():
stream = [1, 2, 3, 4, 5]
mask = [True, True, False, True, True]
expected = [1, 1.5, 1.0, 2.0, 4.5]
rv = streams.rolling_mean(stream, window=2, mask=mask)
assert type(rv) == list
assert rv == expected
def test_rolling_mean_ndarray():
stream = np.asarray([1, 2, 3, 4, 5])
expected = np.asarray([1, 1.5, 2.5, 3.5, 4.5])
rv = streams.rolling_mean(stream, 2)
assert type(rv) == np.ndarray
assert (rv == expected).all()
def test_rolling_mean_with_mask_ndarray():
stream = np.asarray([1, 2, 3, 4, 5])
mask = np.asarray([True, True, False, True, True], dtype=bool)
expected = np.asarray([1, 1.5, 1.0, 2.0, 4.5])
rv = streams.rolling_mean(stream, window=2, mask=mask)
assert type(rv) == np.ndarray
assert (rv == expected).all()