|
Reconstructs a field averaging over several analog patterns
|
Methods
|
|
__getitem__
__init__
returnAnalogs
returnAverage
returnWeightedAverage
|
|
|
__getitem__
|
__getitem__ ( self, idx )
Slide access to the analog record indices
|
|
|
__init__
|
__init__ (
self,
ANALOGobj,
patterns,
smoothing=1,
weightexp=2.0,
report="",
)
Constructor for the ANALOGSelector class
Arguments:
-
ANALOGobj
- An instance of an analog class (
EOFANALOG or
CCAANALOG)
-
patterns
- Base patterns (along the first axis) for which the
analogs are searched
Optional arguments:
-
smoothing
- The number of analogs to average. Defaults to 1 (i.e.
single analog search)
-
weightexp
- In the weighted averages sets the power to raise the
inverse euclidean distance in the weights. Defaults
to 2 (inverse squared distance weights).
-
report
- Filename where to dump a detailed report of the search.
The default behavior is not to dump any report.
|
Exceptions
|
|
pex.ANALOGNoMatchingShape( patterns.shape [ 1 : ], ANALOGobj.originalshape )
|
|
|
|
returnAnalogs
|
returnAnalogs ( self, field=None )
Returns the single analog (smoothing=1) reconstructed field
The output is exactly the same as that of the method returnAverage()
It's just a notation matter. There is no need to use this specific
method for the case smoothing=1 if you are, for instante, into a loop
over several smoothing factors.
Optional argument:
-
field
- To reconstruct the analogs in a field different from the
library dataset. The first axis dimension must match that
of the library dataset.
|
Exceptions
|
|
pex.ANALOGNoMatchingLength(len( field ), len( self.ANALOGobj.dataset ) )
|
|
|
|
returnAverage
|
returnAverage ( self, field=None )
Returns the average reconstructed field
Optional argument:
-
field
- To reconstruct the analogs in a field different from the
library dataset. The first axis dimension must match that
of the library dataset.
|
|
|
returnWeightedAverage
|
returnWeightedAverage ( self, field=None )
Returns the weighted average reconstructed field
Optional argument:
-
field
- To reconstruct the analogs in a field different from the
library dataset. The first axis dimension must match that
of the library dataset.
|
|