|
Analog search in the CCA space
|
Methods
|
|
__init__
getCoords
|
|
|
__init__
|
__init__ (
self,
dataset,
theotherdataset,
neofs=None,
spherized=1,
)
Constructor for a CCA-space search
Arguments:
-
dataset
- NumPy array with the library dataset
(time is first dimension)
-
theotherdataset
- Another array with the complementary field. This is
actually the target field to be reconstructed.
-
neofs
- Tuple with the number of EOFs to retain in each field in the
PCA prefilter for the CCA. This sets the degrees of
freedom of the search space (these are the minimum of this tuple).
Defaults to the number of EOFs accounting for a 70% of the
total variance in each field.
spherized -- Bit indicating if the search in performed in an
spherized space (1) or an inverse correlation scaled
one (0). Default: 1
|
|
|
getCoords
|
getCoords ( self, field )
Returns the coordinates of field in the CCA space
|
Exceptions
|
|
pex.ANALOGNoMatchingShape( field.shape, self.originalshape )
|
|
|