def ppf(self, x):
# convert the parameter to the right format
if isList(x):
x = DataVector(x)
elif isNumerical(x):
x = DataVector([x])
if isinstance(x, DataMatrix):
A = x
B = DataMatrix(A.getNrows(), A.getNcols())
B.setAll(0.0)
elif isinstance(x, DataVector):
A = DataMatrix(1, len(x))
A.setRow(0, x)
B = DataMatrix(1, len(x))
B.setAll(0)
# do the transformation
assert A.getNcols() == B.getNcols() == self.trainData.getNcols()
op = createOperationInverseRosenblattTransformationKDE(self.trainData)
op.doTransformation(A, B)
# transform the outcome
if isNumerical(x) or isinstance(x, DataVector):
return B.get(0, 0)
elif isinstance(x, DataMatrix):
return B.array()
def cdf(self, x):
# convert the parameter to the right format
if isList(x):
x = DataVector(x)
elif isNumerical(x):
x = DataVector([x])
elif isMatrix(x):
x = DataMatrix(x)
if isinstance(x, DataMatrix):
A = x
B = DataMatrix(A.getNrows(), A.getNcols())
B.setAll(0.0)
elif isinstance(x, DataVector):
A = DataMatrix(1, len(x))
A.setRow(0, x)
B = DataMatrix(1, len(x))
B.setAll(0)
# do the transformation
self.dist.cdf(A, B)
# transform the outcome
if isNumerical(x) or isinstance(x, DataVector):
return B.get(0, 0)
elif isinstance(x, DataMatrix):
return B.array()
def ppf(self, x):
# convert the parameter to the right format
if isList(x):
x = DataVector(x)
elif isNumerical(x):
x = DataVector([x])
elif isMatrix(x):
x = DataMatrix(x)
if isinstance(x, DataMatrix):
A = x
B = DataMatrix(A.getNrows(), A.getNcols())
B.setAll(0.0)
elif isinstance(x, DataVector):
A = DataMatrix(1, len(x))
A.setRow(0, x)
B = DataMatrix(1, len(x))
B.setAll(0)
# do the transformation
opInvRosen = createOperationInverseRosenblattTransformationKDE(self.dist)
opInvRosen.doTransformation(A, B)
# transform the outcome
if isNumerical(x) or isinstance(x, DataVector):
return B.get(0, 0)
elif isinstance(x, DataMatrix):
return B.array()
def __init__(self, numDims, trainData=None, bounds=None):
super(EstimatedDist, self).__init__()
self.trainData = None
self.dim = numDims
self.trans = None
if trainData is not None:
if isList(trainData) or len(trainData.shape) == 1:
trainData = np.array([trainData]).reshape(len(trainData), 1)
self.trainData = trainData
self.dim = trainData.shape[1]
if bounds is None:
# estimate bounds from data
bounds = np.vstack(
(np.min(trainData, axis=0), np.max(trainData, axis=0))).T
if bounds is None:
self.bounds = np.array([[0, 1]] * self.dim, dtype="float")
else:
self.bounds = bounds
self.trans = self.computeLinearTransformation(self.bounds)
def ppf(self, x):
# convert the parameter to the right format
if isList(x):
x = DataVector(x)
elif isNumerical(x):
x = DataVector([x])
elif isMatrix(x):
x = DataMatrix(x)
if isinstance(x, DataMatrix):
A = x
B = DataMatrix(A.getNrows(), A.getNcols())
B.setAll(0.0)
elif isinstance(x, DataVector):
A = DataMatrix(1, len(x))
A.setRow(0, x)
B = DataMatrix(1, len(x))
B.setAll(0)
# do the transformation
self.dist.ppf(A, B)
# transform the outcome
if isNumerical(x) or isinstance(x, DataVector):
return B.get(0, 0)
elif isinstance(x, DataMatrix):
return B.array()
def _convertEvalPoint(self, x):
# convert the parameter to the right format
if self.dim == 1:
if isNumerical(x):
x = np.array([[x]])
elif isList(x) or len(x.shape) == 1:
x = np.array([x]).reshape(len(x), 1)
else:
x = np.array(x)
if len(x.shape) == 1:
x = np.array([x])
return x
def pdf(self, x):
# convert the parameter to the right format
if isList(x):
x = DataVector(x)
elif isNumerical(x):
x = DataVector([x])
if isinstance(x, DataMatrix):
A = x
res = DataVector(A.getNrows())
res.setAll(0.0)
elif isinstance(x, DataVector):
A = DataMatrix(1, len(x))
A.setRow(0, x)
res = DataVector(1)
res.setAll(0)
self.dist.pdf(A, res)
if len(res) == 1:
return res[0]
else:
return res.array()
def pdf(self, x):
# convert the parameter to the right format
if isList(x):
x = DataVector(x)
elif isNumerical(x):
x = DataVector([x])
if isinstance(x, DataMatrix):
A = x
res = DataVector(A.getNrows())
res.setAll(0.0)
elif isinstance(x, DataVector):
A = DataMatrix(1, len(x))
A.setRow(0, x)
res = DataVector(1)
res.setAll(0)
self.dist.pdf(A, res)
if len(res) == 1:
return res[0]
else:
return res.array()
