def __init__(self, models, transforms=None, stats=None, algorithm=None):
self._models = models
if transforms is None:
self._transforms = TransformList()
else:
self._transforms = transforms
self._stats = stats
self._algorithm = algorithm
def _prepare(self, X):
X = vstack([X, sqrt(self.c) * self.R])
estimator = TikhonovPseudoInv(X,
self.nPenalties,
intercept=self._intercept)
transforms = TransformList()
return estimator, transforms
def __init__(self, models, transforms=None, stats=None, algorithm=None):
self._models = models
if transforms is None:
self._transforms = TransformList()
else:
self._transforms = transforms
self._stats = stats
self._algorithm = algorithm
class RegressionModel(object):
"""
Class for fitted regression models.
Backed by an RDD of of key-value pairs. The keys are the tuple identifiers from the
Series of response variables used in the fit. The values are per-record regression
models that contain the coefficients from the fit, though these models are not directly
exposed.
"""
def __init__(self, models, transforms=None, stats=None, algorithm=None):
self._models = models
if transforms is None:
self._transforms = TransformList()
else:
self._transforms = transforms
self._stats = stats
self._algorithm = algorithm
def __repr__(self):
lines = []
lines.append(self.__class__.__name__)
if self._transforms.transforms is None:
t = "None"
else:
t = ", ".join([str(x.__class__.__name__) for x in self._transforms.transforms])
lines.append("transformations: " + t)
lines.append("algorithm: " + str(self._algorithm))
return "\n".join(lines)
@property
def coeffs(self):
"""
Series containing the coefficients of the model.
"""
if not hasattr(self, "_coeffs"):
self._coeffs = Series(self._models.mapValues(lambda v: v.betas))
return self._coeffs
@property
def stats(self):
"""
Series containing the R-squared values from the original fit of the model.
"""
if self._stats is None:
return self._stats
else:
return Series(self._stats, index="R2")
def predict(self, X):
"""
Predicts the responses given a design matrix
Parameters
----------
X: array
Design matrix of shape n x k, where n is the number of samples and k is the
number of regressors. Even if an intercept term was fit, should NOT include
a column of ones.
Returns
-------
yhat: Series
Series of predictions (each of length n)
"""
X = self._transforms.apply(X)
return Series(self._models.mapValues(lambda v: v.predict(X)))
def score(self, X, y):
"""
Computes R-squared values for a single design matrix and multiple responses.
Parameters
----------
X: array
Design matrix of shape n x k, where n is the number of samples and k is the
number of regressors. Even if an intercept term was fit, should NOT include
a column of ones.
y: Series
Series of response variables where each record is a vector of length n, where
n is the number of samples.
Returns
-------
scores: Series
Series of R-squared values.
"""
X = self._transforms.apply(X)
joined = self._models.join(y.rdd)
newrdd = joined.mapValues(lambda (model, y): model.stats(X, y))
return Series(newrdd)
def predictAndScore(self, X, y):
X = self._transforms.apply(X)
joined = self._models.join(y.rdd)
results = joined.mapValues(lambda (model, y): model.predictWithStats(X, y))
yhat = results.mapValues(lambda v: v[0])
stats = results.mapValues(lambda v: v[1])
return Series(yhat), Series(stats)
class RegressionModel(object):
"""
Class for fitted regression models.
Backed by an RDD of of key-value pairs. The keys are the tuple identifiers from the
Series of response variables used in the fit. The values are per-record regression
models that contain the coefficients from the fit, though these models are not directly
exposed.
"""
def __init__(self, models, transforms=None, stats=None, algorithm=None):
self._models = models
if transforms is None:
self._transforms = TransformList()
else:
self._transforms = transforms
self._stats = stats
self._algorithm = algorithm
def __repr__(self):
lines = []
lines.append(self.__class__.__name__)
if self._transforms.transforms is None:
t = 'None'
else:
t = ', '.join([
str(x.__class__.__name__) for x in self._transforms.transforms
])
lines.append('transformations: ' + t)
lines.append('algorithm: ' + str(self._algorithm))
return '\n'.join(lines)
@property
def coeffs(self):
"""
Series containing the coefficients of the model.
"""
if not hasattr(self, '_coeffs'):
self._coeffs = Series(self._models.mapValues(lambda v: v.betas))
return self._coeffs
@property
def stats(self):
"""
Series containing the R-squared values from the original fit of the model.
"""
if self._stats is None:
return self._stats
else:
return Series(self._stats, index='R2')
def predict(self, X):
"""
Predicts the responses given a design matrix
Parameters
----------
X: array
Design matrix of shape n x k, where n is the number of samples and k is the
number of regressors. Even if an intercept term was fit, should NOT include
a column of ones.
Returns
-------
yhat: Series
Series of predictions (each of length n)
"""
X = self._transforms.apply(X)
return Series(self._models.mapValues(lambda v: v.predict(X)))
def score(self, X, y):
"""
Computes R-squared values for a single design matrix and multiple responses.
Parameters
----------
X: array
Design matrix of shape n x k, where n is the number of samples and k is the
number of regressors. Even if an intercept term was fit, should NOT include
a column of ones.
y: Series
Series of response variables where each record is a vector of length n, where
n is the number of samples.
Returns
-------
scores: Series
Series of R-squared values.
"""
X = self._transforms.apply(X)
joined = self._models.join(y.rdd)
newrdd = joined.mapValues(lambda (model, y): model.stats(X, y))
return Series(newrdd)
def predictAndScore(self, X, y):
X = self._transforms.apply(X)
joined = self._models.join(y.rdd)
results = joined.mapValues(lambda
(model, y): model.predictWithStats(X, y))
yhat = results.mapValues(lambda v: v[0])
stats = results.mapValues(lambda v: v[1])
return Series(yhat), Series(stats)
def _prepare(self, X):
if self._intercept:
X = AddConstant().transform(X)
estimator = PseudoInv(X)
transforms = TransformList()
return estimator, transforms
