zipline.pipeline.factors._ExponentialWeightedFactor.from_halflife() - Code Metrics - Inspection of "New factors" - quantopian/zipline - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Pull Request — master (#910)

unknown

created 2015-12-12 03:20 UTC

from_halflife() A

↳ Parent: zipline.pipeline.factors._ExponentialWeightedFactor

Complexity

Conditions

Size

Total Lines

Duplication

Lines	0
Ratio	0 %

Metric	Value
cc	3
dl	0
loc	20
rs	9.4286

"""
Technical Analysis Factors
--------------------------
"""
from bottleneck import (
    nanargmax,
    nanmax,
    nanmean,
    nansum,
)
from numbers import Number
from numpy import (
    abs,
    arange,
    average,
    clip,
    diff,
    exp,
    fmax,
    full,
    inf,
    isnan,
    log,
    NINF,
    sqrt,
    sum as np_sum,
)
from numexpr import evaluate

from zipline.pipeline.data import USEquityPricing
from zipline.pipeline.mixins import SingleInputMixin
from zipline.utils.control_flow import ignore_nanwarnings
from zipline.utils.input_validation import expect_types
from .factor import CustomFactor


class Returns(CustomFactor):
    """
    Calculates the percent change in close price over the given window_length.

    **Default Inputs**: [USEquityPricing.close]
    """
    inputs = [USEquityPricing.close]

    def compute(self, today, assets, out, close):
        out[:] = (close[-1] - close[0]) / close[0]


class RSI(CustomFactor, SingleInputMixin):
    """
    Relative Strength Index

    **Default Inputs**: [USEquityPricing.close]

    **Default Window Length**: 15
    """
    window_length = 15
    inputs = (USEquityPricing.close,)

    def compute(self, today, assets, out, closes):
        diffs = diff(closes, axis=0)
        ups = nanmean(clip(diffs, 0, inf), axis=0)
        downs = abs(nanmean(clip(diffs, -inf, 0), axis=0))
        return evaluate(
            "100 - (100 / (1 + (ups / downs)))",
            local_dict={'ups': ups, 'downs': downs},
            global_dict={},
            out=out,
        )


class SimpleMovingAverage(CustomFactor, SingleInputMixin):
    """
    Average Value of an arbitrary column

    **Default Inputs**: None

    **Default Window Length**: None
    """
    # numpy's nan functions throw warnings when passed an array containing only
    # nans, but they still returns the desired value (nan), so we ignore the
    # warning.
    ctx = ignore_nanwarnings()

    def compute(self, today, assets, out, data):
        out[:] = nanmean(data, axis=0)


class WeightedAverageValue(CustomFactor):
    """
    Helper for VWAP-like computations.

    **Default Inputs:** None

    **Default Window Length:** None
    """
    def compute(self, today, assets, out, base, weight):
        out[:] = nansum(base * weight, axis=0) / nansum(weight, axis=0)


class VWAP(WeightedAverageValue):
    """
    Volume Weighted Average Price

    **Default Inputs:** [USEquityPricing.close, USEquityPricing.volume]

    **Default Window Length:** None
    """
    inputs = (USEquityPricing.close, USEquityPricing.volume)


class MaxDrawdown(CustomFactor, SingleInputMixin):
    """
    Max Drawdown

    **Default Inputs:** None

    **Default Window Length:** None
    """
    ctx = ignore_nanwarnings()

    def compute(self, today, assets, out, data):
        drawdowns = fmax.accumulate(data, axis=0) - data
        drawdowns[isnan(drawdowns)] = NINF
        drawdown_ends = nanargmax(drawdowns, axis=0)

        # TODO: Accelerate this loop in Cython or Numba.
        for i, end in enumerate(drawdown_ends):
            peak = nanmax(data[:end + 1, i])
            out[i] = (peak - data[end, i]) / data[end, i]


def DollarVolume():
    """
    Returns a Factor computing the product of most recent close price and
    volume.
    """
    return USEquityPricing.close.latest * USEquityPricing.volume.latest


class _ExponentialWeightedFactor(SingleInputMixin, CustomFactor):
    """
    Base class for factors implementing exponential-weighted operations.

    **Default Inputs:** None
    **Default Window Length:** None

    Parameters
    ----------
    inputs : length-1 list or tuple of BoundColumn
        The expression over which to compute the average.
    window_length : int > 0
        Length of the lookback window over which to compute the average.
    decay_rate : float, 0 < decay_rate <= 1
        Weighting factor by which to discount past observations.

        When calculating historical averages, rows are multiplied by the
        sequence::

            decay_rate, decay_rate ** 2, decay_rate ** 3, ...

    Methods
    -------
    weights
    from_span
    from_halflife
    from_center_of_mass
    """
    params = ('decay_rate',)

    @staticmethod
    def weights(length, decay_rate):
        """
        Return weighting vector for an exponential moving statistic on `length`
        rows with a decay rate of `decay_rate`.
        """
        return full(length, decay_rate) ** arange(length + 1, 1, -1)

    @classmethod
    @expect_types(span=Number)
    def from_span(cls, inputs, window_length, span):
        """
        Convenience constructor for passing `decay_rate` in terms of `span`.

        Forwards `decay_rate` as `1 - (2.0 / (1 + span))`.  This provides the
        behavior equivalent to passing `span` to pandas.ewma.
        """
        if span <= 1:
            raise ValueError(
                "`span` must be a positive number. %s was passed." % span
            )

        decay_rate = (1.0 - (2.0 / (1.0 + span)))
        assert 0.0 < decay_rate <= 1.0

        return cls(
            inputs=inputs,
            window_length=window_length,
            decay_rate=decay_rate,
        )

    @classmethod
    @expect_types(halflife=Number)
    def from_halflife(cls, inputs, window_length, halflife):
        """
        Convenience constructor for passing `decay_rate` in terms of half life.

        Forwards `decay_rate` as `exp(log(.5) / halflife)`.  This provides
        the behavior equivalent to passing `halflife` to pandas.ewma.
        """
        if halflife <= 0:
            raise ValueError(
                "`span` must be a positive number. %s was passed." % halflife
            )
        decay_rate = exp(log(.5) / halflife)
        assert 0.0 < decay_rate <= 1.0

        return cls(
            inputs=inputs,
            window_length=window_length,
            decay_rate=decay_rate,
        )

    @classmethod
    def from_center_of_mass(cls, inputs, window_length, center_of_mass):
        """
        Convenience constructor for passing `decay_rate` in terms of center of
        mass.

        Forwards `decay_rate` as `1 - (1 / center_of_mass)`.  This provides
        behavior equivalent to passing `center_of_mass` to pandas.ewma.
        """
        return cls(
            inputs=inputs,
            window_length=window_length,
            decay_rate=(1.0 - (1.0 / (1.0 + center_of_mass))),
        )


class ExponentialWeightedMovingAverage(_ExponentialWeightedFactor):
    """
    Exponentially Weighted Moving Average

    **Default Inputs:** None
    **Default Window Length:** None

    Parameters
    ----------
    inputs : length-1 list/tuple of BoundColumn
        The expression over which to compute the average.
    window_length : int > 0
        Length of the lookback window over which to compute the average.
    decay_rate : float, 0 < decay_rate <= 1
        Weighting factor by which to discount past observations.

        When calculating historical averages, rows are multiplied by the
        sequence::

            decay_rate, decay_rate ** 2, decay_rate ** 3, ...

    See Also
    --------
    pandas.ewma
    """
    def compute(self, today, assets, out, data, decay_rate):
        out[:] = average(
            data,
            axis=0,
            weights=self.weights(len(data), decay_rate),
        )


class ExponentialWeightedStandardDeviation(_ExponentialWeightedFactor):
    """
    Exponentially Weighted Moving Standard Deviation

    **Default Inputs:** None
    **Default Window Length:** None

    Parameters
    ----------
    inputs : length-1 list/tuple of BoundColumn
        The expression over which to compute the average.
    window_length : int > 0
        Length of the lookback window over which to compute the average.
    decay_rate : float, 0 < decay_rate <= 1
        Weighting factor by which to discount past observations.

        When calculating historical averages, rows are multiplied by the
        sequence::

            decay_rate, decay_rate ** 2, decay_rate ** 3, ...

    See Also
    --------
    pandas.ewmstd
    """

    def compute(self, today, assets, out, data, decay_rate):
        weights = self.weights(len(data), decay_rate)

        mean = average(data, axis=0, weights=weights)
        variance = average((data - mean) ** 2, axis=0, weights=weights)

        squared_weight_sum = (np_sum(weights) ** 2)
        bias_correction = (
            squared_weight_sum / (squared_weight_sum - np_sum(weights ** 2))
        )
        out[:] = sqrt(variance * bias_correction)


# Convenience aliases.
EWMA = ExponentialWeightedMovingAverage
EWMSTD = ExponentialWeightedStandardDeviation


1			"""
2			Technical Analysis Factors
3			--------------------------
4			"""
5			from bottleneck import (
6			nanargmax,
7			nanmax,
8			nanmean,
9			nansum,
10			)
11			from numbers import Number
12			from numpy import (
13			abs,
14			arange,
15			average,
16			clip,
17			diff,
18			exp,
19			fmax,
20			full,
21			inf,
22			isnan,
23			log,
24			NINF,
25			sqrt,
26			sum as np_sum,
27			)
28			from numexpr import evaluate
29
30			from zipline.pipeline.data import USEquityPricing
31			from zipline.pipeline.mixins import SingleInputMixin
32			from zipline.utils.control_flow import ignore_nanwarnings
33			from zipline.utils.input_validation import expect_types
34			from .factor import CustomFactor
35
36
37			class Returns(CustomFactor):
38			"""
39			Calculates the percent change in close price over the given window_length.
40
41			Default Inputs: [USEquityPricing.close]
42			"""
43			inputs = [USEquityPricing.close]
44
45			def compute(self, today, assets, out, close):
46			out[:] = (close[-1] - close[0]) / close[0]
47
48
49			class RSI(CustomFactor, SingleInputMixin):
50			"""
51			Relative Strength Index
52
53			Default Inputs: [USEquityPricing.close]
54
55			Default Window Length: 15
56			"""
57			window_length = 15
58			inputs = (USEquityPricing.close,)
59
60			def compute(self, today, assets, out, closes):
61			diffs = diff(closes, axis=0)
62			ups = nanmean(clip(diffs, 0, inf), axis=0)
63			downs = abs(nanmean(clip(diffs, -inf, 0), axis=0))
64			return evaluate(
65			"100 - (100 / (1 + (ups / downs)))",
66			local_dict={'ups': ups, 'downs': downs},
67			global_dict={},
68			out=out,
69			)
70
71
72			class SimpleMovingAverage(CustomFactor, SingleInputMixin):
73			"""
74			Average Value of an arbitrary column
75
76			Default Inputs: None
77
78			Default Window Length: None
79			"""
80			# numpy's nan functions throw warnings when passed an array containing only
81			# nans, but they still returns the desired value (nan), so we ignore the
82			# warning.
83			ctx = ignore_nanwarnings()
84
85			def compute(self, today, assets, out, data):
86			out[:] = nanmean(data, axis=0)
87
88
89			class WeightedAverageValue(CustomFactor):
90			"""
91			Helper for VWAP-like computations.
92
93			Default Inputs: None
94
95			Default Window Length: None
96			"""
97			def compute(self, today, assets, out, base, weight):
98			out[:] = nansum(base * weight, axis=0) / nansum(weight, axis=0)
99
100
101			class VWAP(WeightedAverageValue):
102			"""
103			Volume Weighted Average Price
104
105			Default Inputs: [USEquityPricing.close, USEquityPricing.volume]
106
107			Default Window Length: None
108			"""
109			inputs = (USEquityPricing.close, USEquityPricing.volume)
110
111
112			class MaxDrawdown(CustomFactor, SingleInputMixin):
113			"""
114			Max Drawdown
115
116			Default Inputs: None
117
118			Default Window Length: None
119			"""
120			ctx = ignore_nanwarnings()
121
122			def compute(self, today, assets, out, data):
123			drawdowns = fmax.accumulate(data, axis=0) - data
124			drawdowns[isnan(drawdowns)] = NINF
125			drawdown_ends = nanargmax(drawdowns, axis=0)
126
127			# TODO: Accelerate this loop in Cython or Numba.
128			for i, end in enumerate(drawdown_ends):
129			peak = nanmax(data[:end + 1, i])
130			out[i] = (peak - data[end, i]) / data[end, i]
131
132
133			def DollarVolume():
134			"""
135			Returns a Factor computing the product of most recent close price and
136			volume.
137			"""
138			return USEquityPricing.close.latest * USEquityPricing.volume.latest
139
140
141			class _ExponentialWeightedFactor(SingleInputMixin, CustomFactor):
142			"""
143			Base class for factors implementing exponential-weighted operations.
144
145			Default Inputs: None
146			Default Window Length: None
147
148			Parameters
149			----------
150			inputs : length-1 list or tuple of BoundColumn
151			The expression over which to compute the average.
152			window_length : int > 0
153			Length of the lookback window over which to compute the average.
154			decay_rate : float, 0 < decay_rate <= 1
155			Weighting factor by which to discount past observations.
156
157			When calculating historical averages, rows are multiplied by the
158			sequence::
159
160			decay_rate, decay_rate 2, decay_rate 3, ...
161
162			Methods
163			-------
164			weights
165			from_span
166			from_halflife
167			from_center_of_mass
168			"""
169			params = ('decay_rate',)
170
171			@staticmethod
172			def weights(length, decay_rate):
173			"""
174			Return weighting vector for an exponential moving statistic on `length`
175			rows with a decay rate of `decay_rate`.
176			"""
177			return full(length, decay_rate) ** arange(length + 1, 1, -1)
178
179			@classmethod
180			@expect_types(span=Number)
181			def from_span(cls, inputs, window_length, span):
182			"""
183			Convenience constructor for passing `decay_rate` in terms of `span`.
184
185			Forwards `decay_rate` as `1 - (2.0 / (1 + span))`. This provides the
186			behavior equivalent to passing `span` to pandas.ewma.
187			"""
188			if span <= 1:
189			raise ValueError(
190			"`span` must be a positive number. %s was passed." % span
191			)
192
193			decay_rate = (1.0 - (2.0 / (1.0 + span)))
194			assert 0.0 < decay_rate <= 1.0
195
196			return cls(
197			inputs=inputs,
198			window_length=window_length,
199			decay_rate=decay_rate,
200			)
201
202			@classmethod
203			@expect_types(halflife=Number)
204			def from_halflife(cls, inputs, window_length, halflife):
205			"""
206			Convenience constructor for passing `decay_rate` in terms of half life.
207
208			Forwards `decay_rate` as `exp(log(.5) / halflife)`. This provides
209			the behavior equivalent to passing `halflife` to pandas.ewma.
210			"""
211			if halflife <= 0:
212			raise ValueError(
213			"`span` must be a positive number. %s was passed." % halflife
214			)
215			decay_rate = exp(log(.5) / halflife)
216			assert 0.0 < decay_rate <= 1.0
217
218			return cls(
219			inputs=inputs,
220			window_length=window_length,
221			decay_rate=decay_rate,
222			)
223
224			@classmethod
225			def from_center_of_mass(cls, inputs, window_length, center_of_mass):
226			"""
227			Convenience constructor for passing `decay_rate` in terms of center of
228			mass.
229
230			Forwards `decay_rate` as `1 - (1 / center_of_mass)`. This provides
231			behavior equivalent to passing `center_of_mass` to pandas.ewma.
232			"""
233			return cls(
234			inputs=inputs,
235			window_length=window_length,
236			decay_rate=(1.0 - (1.0 / (1.0 + center_of_mass))),
237			)
238
239
240			class ExponentialWeightedMovingAverage(_ExponentialWeightedFactor):
241			"""
242			Exponentially Weighted Moving Average
243
244			Default Inputs: None
245			Default Window Length: None
246
247			Parameters
248			----------
249			inputs : length-1 list/tuple of BoundColumn
250			The expression over which to compute the average.
251			window_length : int > 0
252			Length of the lookback window over which to compute the average.
253			decay_rate : float, 0 < decay_rate <= 1
254			Weighting factor by which to discount past observations.
255
256			When calculating historical averages, rows are multiplied by the
257			sequence::
258
259			decay_rate, decay_rate 2, decay_rate 3, ...
260
261			See Also
262			--------
263			pandas.ewma
264			"""
265			def compute(self, today, assets, out, data, decay_rate):
266			out[:] = average(
267			data,
268			axis=0,
269			weights=self.weights(len(data), decay_rate),
270			)
271
272
273			class ExponentialWeightedStandardDeviation(_ExponentialWeightedFactor):
274			"""
275			Exponentially Weighted Moving Standard Deviation
276
277			Default Inputs: None
278			Default Window Length: None
279
280			Parameters
281			----------
282			inputs : length-1 list/tuple of BoundColumn
283			The expression over which to compute the average.
284			window_length : int > 0
285			Length of the lookback window over which to compute the average.
286			decay_rate : float, 0 < decay_rate <= 1
287			Weighting factor by which to discount past observations.
288
289			When calculating historical averages, rows are multiplied by the
290			sequence::
291
292			decay_rate, decay_rate 2, decay_rate 3, ...
293
294			See Also
295			--------
296			pandas.ewmstd
297			"""
298
299			def compute(self, today, assets, out, data, decay_rate):
300			weights = self.weights(len(data), decay_rate)
301
302			mean = average(data, axis=0, weights=weights)
303			variance = average((data - mean) ** 2, axis=0, weights=weights)
304
305			squared_weight_sum = (np_sum(weights) ** 2)
306			bias_correction = (
307			squared_weight_sum / (squared_weight_sum - np_sum(weights ** 2))
308			)
309			out[:] = sqrt(variance * bias_correction)
310
311
312			# Convenience aliases.
313			EWMA = ExponentialWeightedMovingAverage
314			EWMSTD = ExponentialWeightedStandardDeviation
315

quantopian / zipline

Pull Request — master (#910)

from_halflife() A

Complexity

Size

Duplication

Duplication Side-by-Side

Filter issues like