"""
Class for chaining gridders.
"""
from sklearn.utils.validation import check_is_fitted
from .base import BaseGridder
from .base.utils import check_data
from .coordinates import get_region
[docs]class Chain(BaseGridder):
"""
Chain filtering operations to fit on each subsequent output.
The :meth:`~verde.base.BaseGridder.filter` method of each element of the
set is called with the outputs of the previous one. For gridders and trend
estimators this means that each element fits the residuals (input data
minus predicted data) of the previous one.
When predicting data, the predictions of each step in the chain are added
together. Steps that don't implement a
:meth:`~verde.base.BaseGridder.predict` method are ignored.
This provides a convenient way to chaining operations like trend estimation
to a given gridder.
Parameters
----------
steps : list
A list of ``('name', step)`` pairs where ``step`` is any verde
class that implements a ``filter(coordinates, data, weights)`` method
(including ``Chain`` itself).
Attributes
----------
region_ : tuple
The boundaries (``[W, E, S, N]``) of the data used to fit the chain.
Used as the default region for the :meth:`~verde.Chain.grid` and
:meth:`~verde.Chain.scatter` methods.
named_steps : dict
A dictionary version of *steps* where the ``'name'`` strings are keys
and the estimator/gridder/processor objects are the values.
See also
--------
verde.Vector : Fit an estimator to each component of vector data
"""
def __init__(self, steps):
super().__init__()
self.steps = steps
@property
def named_steps(self):
"""
A dictionary version of steps.
"""
return dict(self.steps)
[docs] def fit(self, coordinates, data, weights=None):
"""
Fit the chained estimators to the given data.
Each estimator in the chain is fitted to the residuals of the previous
estimator. The coordinates are preserved. Only the data values are
changed.
The data region is captured and used as default for the
:meth:`~verde.Chain.grid` and :meth:`~verde.Chain.scatter` methods.
All input arrays must have the same shape.
Parameters
----------
coordinates : tuple of arrays
Arrays with the coordinates of each data point. Should be in the
following order: (easting, northing, vertical, ...).
data : array or tuple of arrays
the data values of each data point. if the data has more than one
component, *data* must be a tuple of arrays (one for each
component).
weights : none or array or tuple of arrays
if not none, then the weights assigned to each data point. if more
than one data component is provided, you must provide a weights
array for each data component (if not none).
Returns
-------
self
Returns this estimator instance for chaining operations.
"""
self.region_ = get_region(coordinates[:2])
args = coordinates, data, weights
for _, step in self.steps:
args = step.filter(*args)
return self
[docs] def predict(self, coordinates):
"""
Evaluates the data predicted by the chain on the given set of points.
Predictions from each step in the chain are added together. Any step
that doesn't implement a ``predict`` method is ignored.
Requires a fitted gridder (see :meth:`~verde.Chain.fit`).
Parameters
----------
coordinates : tuple of arrays
Arrays with the coordinates of each data point. Should be in the
following order: (easting, northing, vertical, ...).
Returns
-------
data : array
The data values predicted on the given points.
"""
check_is_fitted(self, ["region_"])
result = None
for _, step in self.steps:
if hasattr(step, "predict"):
predicted = check_data(step.predict(coordinates))
if result is None:
result = [0 for i in range(len(predicted))]
for i, pred in enumerate(predicted):
result[i] += pred
if len(result) == 1:
return result[0]
return tuple(result)