AnniesLasso.CannonModel.train() - Code Metrics - Inspection of "Code status" - andycasey/AnniesLasso - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Pull Request — master (#1)

by Andy

created 2015-11-21 10:06 UTC

AnniesLasso.CannonModel.train() B

↳ Parent: AnniesLasso.CannonModel

Complexity

Conditions

Size

Total Lines

Duplication

Lines	0
Ratio	0 %

Metric	Value
dl	0
loc	46
rs	7.5748
cc	6

#!/usr/bin/env python
# -*- coding: utf-8 -*-

"""
A pedestrian version of The Cannon.
"""

from __future__ import (division, print_function, absolute_import,
                        unicode_literals)

__all__ = ["CannonModel"]

import logging
import numpy as np
import scipy.optimize as op

from . import (model, utils)

logger = logging.getLogger(__name__)


class CannonModel(model.BaseCannonModel):
    """
    A generalised Cannon model for the estimation of arbitrary stellar labels.

    :param labels:
        A table with columns as labels, and stars as rows.

    :type labels:
        :class:`~astropy.table.Table` or numpy structured array

    :param fluxes:
        An array of fluxes for stars in the training set, given as shape
        `(num_stars, num_pixels)`. The `num_stars` should match the number of
        rows in `labels`.

    :type fluxes:
        :class:`np.ndarray`

    :param flux_uncertainties:
        An array of 1-sigma flux uncertainties for stars in the training set,
        The shape of the `flux_uncertainties` should match `fluxes`. 

    :type flux_uncertainties:
        :class:`np.ndarray`

    :param dispersion: [optional]
        The dispersion values corresponding to the given pixels. If provided, 
        this should have length `num_pixels`.

    :param live_dangerously: [optional]
        If enabled then no checks will be made on the label names, prohibiting
        the user to input human-readable forms of the label vector.
    """

    _descriptive_attributes = ["_label_vector"]
    _trained_attributes = ["_coefficients", "_scatter"]
    _data_attributes = \
        ["training_labels", "training_fluxes", "training_flux_uncertainties"]
    
    def __init__(self, *args, **kwargs):
        super(CannonModel, self).__init__(*args, **kwargs)


    @model.requires_label_vector
    def train(self, **kwargs):
        """
        Train the model based on the training set and the description of the
        label vector.
        """
        
        # Initialise the scatter and coefficient arrays.
        N_px = len(self.dispersion)
        scatter = np.nan * np.ones(N_px)
        label_vector_array = self.label_vector_array
        theta = np.nan * np.ones((N_px, label_vector_array.shape[0]))

        # Details for the progressbar.
        pb_kwds = {
            "message": "Training Cannon model from {0} stars with {1} pixels "
                       "each".format(len(self.training_labels), N_px),
            "size": 100 if kwargs.pop("progressbar", True) else -1
        }
        
        if self.pool is None:
            for pixel in utils.progressbar(range(N_px), **pb_kwds):
                theta[pixel, :], scatter[pixel] = _fit_pixel(
                    self.training_fluxes[:, pixel], 
                    self.training_flux_uncertainties[:, pixel],
                    label_vector_array, **kwargs)

        else:
            # Not as nice as just mapping, but necessary for a progress bar.
            process = { pixel: self.pool.apply_async(
                    _fit_pixel,
                    args=(
                        self.training_fluxes[:, pixel], 
                        self.training_flux_uncertainties[:, pixel],
                        label_vector_array
                    ),
                    kwds=kwargs) \
                for pixel in range(N_px) }

            for pixel, proc in utils.progressbar(process.items(), **pb_kwds):
                theta[pixel, :], scatter[pixel] = proc.get()

        self.coefficients, self.scatter = (theta, scatter)
        self._trained = True

        return (theta, scatter)


    @model.requires_training_wheels
    def predict(self, labels=None, **labels_as_kwargs):
        """
        Predict spectra from the trained model, given the labels.

        :param labels: [optional]
            The labels required for the trained model. This should be a N-length
            list matching the number of unique terms in the model, in the order
            given by `self.labels`. Alternatively, labels can be explicitly 
            given as keyword arguments.
        """

        labels = self._format_input_labels(labels, **labels_as_kwargs)

        return np.dot(self.coefficients, 
            model._build_label_vector_rows(self.label_vector, labels)).flatten()


    @model.requires_training_wheels
    def fit(self, fluxes, flux_uncertainties, **kwargs):
        """
        Solve the labels for given pixel fluxes and uncertainties.

        :param fluxes:
            The normalised fluxes. These should be on the same dispersion scale
            as the trained data.

        :param flux_uncertainties:
            The 1-sigma uncertainties in the fluxes. This should have the same
            shape as `fluxes`.

        :returns:
            The labels and covariance matrix.
        """

        label_indices = self._get_lowest_order_label_indices()
        fluxes, flux_uncertainties = map(np.array, (fluxes, flux_uncertainties))

        # TODO: Consider parallelising this, which would mean factoring
        # _fit out of the model class, which gets messy.
        # Since solving for labels is not a big bottleneck (yet), let's leave
        # this.

        full_output = kwargs.pop("full_output", False)
        if fluxes.ndim == 1:
            labels, covariance = \
                self._fit(fluxes, flux_uncertainties, label_indices, **kwargs)
        else:
            N_stars, N_labels = (fluxes.shape[0], len(self.labels))
            labels = np.empty((N_stars, N_labels), dtype=float)
            covariance = np.empty((N_stars, N_labels, N_labels), dtype=float)

            for i, (f, u) in enumerate(zip(fluxes, flux_uncertainties)):
                labels[i, :], covariance[i, :] = \
                    self._fit(f, u, label_indices, **kwargs)

        if full_output:
            return (labels, covariance)
        return labels


    def _fit(self, fluxes, flux_uncertainties, label_indices, **kwargs):
        """
        Solve the labels for given pixel fluxes and uncertainties
        for a single star.

        :param fluxes:
            The normalised fluxes. These should be on the same dispersion scale
            as the trained data.

        :param flux_uncertainties:
            The 1-sigma uncertainties in the fluxes. This should have the same
            shape as `fluxes`.

        :returns:
            The labels and covariance matrix.
        """

        # Check which pixels to use, then just use those.
        use = (flux_uncertainties < kwargs.get("max_uncertainty", 1)) \
            * np.isfinite(self.coefficients[:, 0] * fluxes * flux_uncertainties)

        fluxes = fluxes.copy()[use]
        flux_uncertainties = flux_uncertainties.copy()[use]
        scatter, coefficients = self.scatter[use], self.coefficients[use]

        Cinv = 1.0 / (scatter**2 + flux_uncertainties**2)
        A = np.dot(coefficients.T, Cinv[:, None] * coefficients)
        B = np.dot(coefficients.T, Cinv * fluxes)
        theta_p0 = np.linalg.solve(A, B)

        # Need to match the initial theta coefficients back to label values.
        # (Maybe this should use some general non-linear simultaneous solver?)
        initial = {}
        for index in label_indices:
            if index is None: continue
            label, order = self.label_vector[index][0]
            # The +1 index offset is because the first theta is a scaling.
            value = abs(theta_p0[1 + index])**(1./order)
            if not np.isfinite(value): continue
            initial[label] = value

        missing = set(self.labels).difference(initial)
        if missing:
            # There must be some coefficients that are only used in cross-terms.
            # We could solve for them, or just take the mean of the training
            # set as the initial guess.
            initial.update({ label: \
                np.nanmean(self.training_labels[label]) for label in missing
            })
            
        # Create and test the generating function.
        def function(coeffs, *labels):
            return np.dot(coeffs, 
                model._build_label_vector_rows(self.label_vector, 
                    { label: [v] for label, v in zip(self.labels, labels) }
                )).flatten()

        # Solve for the parameters.
        kwds = {
            "p0": np.array([initial[label] for label in self.labels]),
            "maxfev": 10000,
            "sigma": 1.0/np.sqrt(Cinv),
            "absolute_sigma": True
        }
        kwds.update(kwargs)
        labels_opt, cov = op.curve_fit(function, coefficients, fluxes, **kwds)

        return (labels_opt, cov)



def _fit_pixel(fluxes, flux_uncertainties, label_vector_array, **kwargs):
    """
    Return the optimal label vector coefficients and scatter for a pixel, given
    the fluxes, uncertainties, and the label vector array.

    :param fluxes:
        The fluxes for the given pixel, from all stars.

    :param flux_uncertainties:
        The 1-sigma flux uncertainties for the given pixel, from all stars.

    :param label_vector_array:
        The label vector array. This should have shape `(N_stars, N_terms + 1)`.

    :returns:
        The optimised label vector coefficients and scatter for this pixel.
    """

    _ = kwargs.get("max_uncertainty", 1)
    failed_response = (np.nan * np.ones(label_vector_array.shape[0]), _)
    if np.all(flux_uncertainties > _):
        return failed_response

    # Get an initial guess of the scatter.
    scatter = np.var(fluxes) - np.median(flux_uncertainties)**2
    scatter = np.sqrt(scatter) if scatter >= 0 else np.std(fluxes)
    
    # Optimise the scatter, and at each scatter value we will calculate the
    # optimal vector coefficients.
    op_scatter, fopt, direc, n_iter, n_funcs, warnflag = op.fmin_powell(
        _pixel_scatter_nll, scatter,
        args=(fluxes, flux_uncertainties, label_vector_array),
        disp=False, full_output=True)

    if warnflag > 0:
        logger.warning("Warning: {}".format([
            "Maximum number of function evaluations made during optimisation.",
            "Maximum number of iterations made during optimisation."
            ][warnflag - 1]))

    # Calculate the coefficients at the optimal scatter value.
    # Note that if we can't solve for the coefficients, we should just set them
    # as zero and send back a giant variance.
    try:
        coefficients, ATCiAinv, variance = _fit_coefficients(
            fluxes, flux_uncertainties, op_scatter, label_vector_array)

    except np.linalg.linalg.LinAlgError:
        logger.exception("Failed to calculate coefficients")
        if kwargs.get("debug", False): raise

        return failed_response

    else:
        return (coefficients, op_scatter)


def _pixel_scatter_nll(scatter, fluxes, flux_uncertainties, label_vector_array,
    **kwargs):
    """
    Return the negative log-likelihood for the scatter in a single pixel.

    :param scatter:
        The model scatter in the pixel.

    :param fluxes:
        The fluxes for a given pixel (in many stars).

    :param flux_uncertainties:
        The 1-sigma uncertainties in the fluxes for a given pixel. This should
        have the same shape as `fluxes`.

    :param label_vector_array:
        The label vector array for each star, for the given pixel.

    :returns:
        The log-likelihood of the log scatter, given the fluxes and the label
        vector array.

    :raises np.linalg.linalg.LinAlgError:
        If there was an error in inverting a matrix, and `debug` is set to True.
    """

    if 0 > scatter:
        return np.inf

    try:
        # Calculate the coefficients for the given level of scatter.
        theta, ATCiAinv, variance = _fit_coefficients(
            fluxes, flux_uncertainties, scatter, label_vector_array)

    except np.linalg.linalg.LinAlgError:
        if kwargs.get("debug", False): raise
        return np.inf

    model = np.dot(theta, label_vector_array)

    return 0.5 * np.sum((fluxes - model)**2 / variance) \
        +  0.5 * np.sum(np.log(variance))


def _fit_coefficients(fluxes, flux_uncertainties, scatter, label_vector_array):
    """
    Fit model coefficients and scatter to a given set of normalised fluxes for a
    single pixel.

    :param fluxes:
        The normalised fluxes for a single pixel (in many stars).

    :param flux_uncertainties:
        The 1-sigma uncertainties in normalised fluxes. This should have the
        same shape as `fluxes`.

    :param label_vector_array:
        The label vector array for each pixel.

    :returns:
        The label vector coefficients for the pixel, the inverse variance matrix
        and the total pixel variance.
    """

    variance = flux_uncertainties**2 + scatter**2
    CiA = label_vector_array.T * \
        np.tile(1./variance, (label_vector_array.shape[0], 1)).T
    ATCiAinv = np.linalg.inv(np.dot(label_vector_array, CiA))

    ATY = np.dot(label_vector_array, fluxes/variance)
    theta = np.dot(ATCiAinv, ATY)

    return (theta, ATCiAinv, variance)


1			#!/usr/bin/env python
2			# -- coding: utf-8 --
3
4			"""
5			A pedestrian version of The Cannon.
6			"""
7
8			from __future__ import (division, print_function, absolute_import,
9			unicode_literals)
10
11			__all__ = ["CannonModel"]
12
13			import logging
14			import numpy as np
15			import scipy.optimize as op
16
17			from . import (model, utils)
18
19			logger = logging.getLogger(__name__)
20
21
22			class CannonModel(model.BaseCannonModel):
23			"""
24			A generalised Cannon model for the estimation of arbitrary stellar labels.
25
26			:param labels:
27			A table with columns as labels, and stars as rows.
28
29			:type labels:
30			:class:`~astropy.table.Table` or numpy structured array
31
32			:param fluxes:
33			An array of fluxes for stars in the training set, given as shape
34			`(num_stars, num_pixels)`. The `num_stars` should match the number of
35			rows in `labels`.
36
37			:type fluxes:
38			:class:`np.ndarray`
39
40			:param flux_uncertainties:
41			An array of 1-sigma flux uncertainties for stars in the training set,
42			The shape of the `flux_uncertainties` should match `fluxes`.
43
44			:type flux_uncertainties:
45			:class:`np.ndarray`
46
47			:param dispersion: [optional]
48			The dispersion values corresponding to the given pixels. If provided,
49			this should have length `num_pixels`.
50
51			:param live_dangerously: [optional]
52			If enabled then no checks will be made on the label names, prohibiting
53			the user to input human-readable forms of the label vector.
54			"""
55
56			_descriptive_attributes = ["_label_vector"]
57			_trained_attributes = ["_coefficients", "_scatter"]
58			_data_attributes = \
59			["training_labels", "training_fluxes", "training_flux_uncertainties"]
60
61			def __init__(self, args, *kwargs):
62			super(CannonModel, self).__init__(args, *kwargs)
63
64
65			@model.requires_label_vector
66			def train(self, **kwargs):
67			"""
68			Train the model based on the training set and the description of the
69			label vector.
70			"""
71
72			# Initialise the scatter and coefficient arrays.
73			N_px = len(self.dispersion)
74			scatter = np.nan * np.ones(N_px)
75			label_vector_array = self.label_vector_array
76			theta = np.nan * np.ones((N_px, label_vector_array.shape[0]))
77
78			# Details for the progressbar.
79			pb_kwds = {
80			"message": "Training Cannon model from {0} stars with {1} pixels "
81			"each".format(len(self.training_labels), N_px),
82			"size": 100 if kwargs.pop("progressbar", True) else -1
83			}
84
85			if self.pool is None:
86			for pixel in utils.progressbar(range(N_px), **pb_kwds):
87			theta[pixel, :], scatter[pixel] = _fit_pixel(
88			self.training_fluxes[:, pixel],
89			self.training_flux_uncertainties[:, pixel],
90			label_vector_array, **kwargs)
91
92			else:
93			# Not as nice as just mapping, but necessary for a progress bar.
94			process = { pixel: self.pool.apply_async(
95			_fit_pixel,
96			args=(
97			self.training_fluxes[:, pixel],
98			self.training_flux_uncertainties[:, pixel],
99			label_vector_array
100			),
101			kwds=kwargs) \
102			for pixel in range(N_px) }
103
104			for pixel, proc in utils.progressbar(process.items(), **pb_kwds):
105			theta[pixel, :], scatter[pixel] = proc.get()
106
107			self.coefficients, self.scatter = (theta, scatter)
108			self._trained = True
109
110			return (theta, scatter)
111
112
113			@model.requires_training_wheels
114			def predict(self, labels=None, **labels_as_kwargs):
115			"""
116			Predict spectra from the trained model, given the labels.
117
118			:param labels: [optional]
119			The labels required for the trained model. This should be a N-length
120			list matching the number of unique terms in the model, in the order
121			given by `self.labels`. Alternatively, labels can be explicitly
122			given as keyword arguments.
123			"""
124
125			labels = self._format_input_labels(labels, **labels_as_kwargs)
126
127			return np.dot(self.coefficients,
128			model._build_label_vector_rows(self.label_vector, labels)).flatten()
129
130
131			@model.requires_training_wheels
132			def fit(self, fluxes, flux_uncertainties, **kwargs):
133			"""
134			Solve the labels for given pixel fluxes and uncertainties.
135
136			:param fluxes:
137			The normalised fluxes. These should be on the same dispersion scale
138			as the trained data.
139
140			:param flux_uncertainties:
141			The 1-sigma uncertainties in the fluxes. This should have the same
142			shape as `fluxes`.
143
144			:returns:
145			The labels and covariance matrix.
146			"""
147
148			label_indices = self._get_lowest_order_label_indices()
149			fluxes, flux_uncertainties = map(np.array, (fluxes, flux_uncertainties))
150
151			# TODO: Consider parallelising this, which would mean factoring
152			# _fit out of the model class, which gets messy.
153			# Since solving for labels is not a big bottleneck (yet), let's leave
154			# this.
155
156			full_output = kwargs.pop("full_output", False)
157			if fluxes.ndim == 1:
158			labels, covariance = \
159			self._fit(fluxes, flux_uncertainties, label_indices, **kwargs)
160			else:
161			N_stars, N_labels = (fluxes.shape[0], len(self.labels))
162			labels = np.empty((N_stars, N_labels), dtype=float)
163			covariance = np.empty((N_stars, N_labels, N_labels), dtype=float)
164
165			for i, (f, u) in enumerate(zip(fluxes, flux_uncertainties)):
166			labels[i, :], covariance[i, :] = \
167			self._fit(f, u, label_indices, **kwargs)
168
169			if full_output:
170			return (labels, covariance)
171			return labels
172
173
174			def _fit(self, fluxes, flux_uncertainties, label_indices, **kwargs):
175			"""
176			Solve the labels for given pixel fluxes and uncertainties
177			for a single star.
178
179			:param fluxes:
180			The normalised fluxes. These should be on the same dispersion scale
181			as the trained data.
182
183			:param flux_uncertainties:
184			The 1-sigma uncertainties in the fluxes. This should have the same
185			shape as `fluxes`.
186
187			:returns:
188			The labels and covariance matrix.
189			"""
190
191			# Check which pixels to use, then just use those.
192			use = (flux_uncertainties < kwargs.get("max_uncertainty", 1)) \
193			* np.isfinite(self.coefficients[:, 0] * fluxes * flux_uncertainties)
194
195			fluxes = fluxes.copy()[use]
196			flux_uncertainties = flux_uncertainties.copy()[use]
197			scatter, coefficients = self.scatter[use], self.coefficients[use]
198
199			Cinv = 1.0 / (scatter2 + flux_uncertainties2)
200			A = np.dot(coefficients.T, Cinv[:, None] * coefficients)
201			B = np.dot(coefficients.T, Cinv * fluxes)
202			theta_p0 = np.linalg.solve(A, B)
203
204			# Need to match the initial theta coefficients back to label values.
205			# (Maybe this should use some general non-linear simultaneous solver?)
206			initial = {}
207			for index in label_indices:
208			if index is None: continue
209			label, order = self.label_vector[index][0]
210			# The +1 index offset is because the first theta is a scaling.
211			value = abs(theta_p0[1 + index])**(1./order)
212			if not np.isfinite(value): continue
213			initial[label] = value
214
215			missing = set(self.labels).difference(initial)
216			if missing:
217			# There must be some coefficients that are only used in cross-terms.
218			# We could solve for them, or just take the mean of the training
219			# set as the initial guess.
220			initial.update({ label: \
221			np.nanmean(self.training_labels[label]) for label in missing
222			})
223
224			# Create and test the generating function.
225			def function(coeffs, *labels):
226			return np.dot(coeffs,
227			model._build_label_vector_rows(self.label_vector,
228			{ label: [v] for label, v in zip(self.labels, labels) }
229			)).flatten()
230
231			# Solve for the parameters.
232			kwds = {
233			"p0": np.array([initial[label] for label in self.labels]),
234			"maxfev": 10000,
235			"sigma": 1.0/np.sqrt(Cinv),
236			"absolute_sigma": True
237			}
238			kwds.update(kwargs)
239			labels_opt, cov = op.curve_fit(function, coefficients, fluxes, **kwds)
240
241			return (labels_opt, cov)
242
243
244
245			def _fit_pixel(fluxes, flux_uncertainties, label_vector_array, **kwargs):
246			"""
247			Return the optimal label vector coefficients and scatter for a pixel, given
248			the fluxes, uncertainties, and the label vector array.
249
250			:param fluxes:
251			The fluxes for the given pixel, from all stars.
252
253			:param flux_uncertainties:
254			The 1-sigma flux uncertainties for the given pixel, from all stars.
255
256			:param label_vector_array:
257			The label vector array. This should have shape `(N_stars, N_terms + 1)`.
258
259			:returns:
260			The optimised label vector coefficients and scatter for this pixel.
261			"""
262
263			_ = kwargs.get("max_uncertainty", 1)
264			failed_response = (np.nan * np.ones(label_vector_array.shape[0]), _)
265			if np.all(flux_uncertainties > _):
266			return failed_response
267
268			# Get an initial guess of the scatter.
269			scatter = np.var(fluxes) - np.median(flux_uncertainties)**2
270			scatter = np.sqrt(scatter) if scatter >= 0 else np.std(fluxes)
271
272			# Optimise the scatter, and at each scatter value we will calculate the
273			# optimal vector coefficients.
274			op_scatter, fopt, direc, n_iter, n_funcs, warnflag = op.fmin_powell(
275			_pixel_scatter_nll, scatter,
276			args=(fluxes, flux_uncertainties, label_vector_array),
277			disp=False, full_output=True)
278
279			if warnflag > 0:
280			logger.warning("Warning: {}".format([
281			"Maximum number of function evaluations made during optimisation.",
282			"Maximum number of iterations made during optimisation."
283			][warnflag - 1]))
284
285			# Calculate the coefficients at the optimal scatter value.
286			# Note that if we can't solve for the coefficients, we should just set them
287			# as zero and send back a giant variance.
288			try:
289			coefficients, ATCiAinv, variance = _fit_coefficients(
290			fluxes, flux_uncertainties, op_scatter, label_vector_array)
291
292			except np.linalg.linalg.LinAlgError:
293			logger.exception("Failed to calculate coefficients")
294			if kwargs.get("debug", False): raise
295
296			return failed_response
297
298			else:
299			return (coefficients, op_scatter)
300
301
302			def _pixel_scatter_nll(scatter, fluxes, flux_uncertainties, label_vector_array,
303			**kwargs):
304			"""
305			Return the negative log-likelihood for the scatter in a single pixel.
306
307			:param scatter:
308			The model scatter in the pixel.
309
310			:param fluxes:
311			The fluxes for a given pixel (in many stars).
312
313			:param flux_uncertainties:
314			The 1-sigma uncertainties in the fluxes for a given pixel. This should
315			have the same shape as `fluxes`.
316
317			:param label_vector_array:
318			The label vector array for each star, for the given pixel.
319
320			:returns:
321			The log-likelihood of the log scatter, given the fluxes and the label
322			vector array.
323
324			:raises np.linalg.linalg.LinAlgError:
325			If there was an error in inverting a matrix, and `debug` is set to True.
326			"""
327
328			if 0 > scatter:
329			return np.inf
330
331			try:
332			# Calculate the coefficients for the given level of scatter.
333			theta, ATCiAinv, variance = _fit_coefficients(
334			fluxes, flux_uncertainties, scatter, label_vector_array)
335
336			except np.linalg.linalg.LinAlgError:
337			if kwargs.get("debug", False): raise
338			return np.inf
339
340			model = np.dot(theta, label_vector_array)
341
342			return 0.5 * np.sum((fluxes - model)**2 / variance) \
343			+ 0.5 * np.sum(np.log(variance))
344
345
346			def _fit_coefficients(fluxes, flux_uncertainties, scatter, label_vector_array):
347			"""
348			Fit model coefficients and scatter to a given set of normalised fluxes for a
349			single pixel.
350
351			:param fluxes:
352			The normalised fluxes for a single pixel (in many stars).
353
354			:param flux_uncertainties:
355			The 1-sigma uncertainties in normalised fluxes. This should have the
356			same shape as `fluxes`.
357
358			:param label_vector_array:
359			The label vector array for each pixel.
360
361			:returns:
362			The label vector coefficients for the pixel, the inverse variance matrix
363			and the total pixel variance.
364			"""
365
366			variance = flux_uncertainties2 + scatter2
367			CiA = label_vector_array.T * \
368			np.tile(1./variance, (label_vector_array.shape[0], 1)).T
369			ATCiAinv = np.linalg.inv(np.dot(label_vector_array, CiA))
370
371			ATY = np.dot(label_vector_array, fluxes/variance)
372			theta = np.dot(ATCiAinv, ATY)
373
374			return (theta, ATCiAinv, variance)
375

andycasey / AnniesLasso

Pull Request — master (#1)

AnniesLasso.CannonModel.train() B

Complexity

Size

Duplication

Duplication Side-by-Side

Filter issues like