__author__ = "Xun Li <lixun910@gmail.com>"
from ..libgeoda import gda_distance_weights, gda_knn_weights
from .weight import Weight
[docs]def kernel_weights(geoda_obj, bandwidth, kernel, **kwargs):
'''Distance-based Kernel Spatial Weights
Create a kernel weights by specifying a bandwidth and a kernel method
Args:
geoda_obj (geoda): An instance of geoda class.
bandwidth (float): A float value represents the distance of bandwidth used in kernel function.
For example, one can use the pygeoda.min_distthreshold() to get a distance that
guarantees that every observation has at least 1 neighbor.
kernel (str): A string value, which has to be one of {'triangular', 'uniform', 'epanechnikov', 'quartic', 'gaussian'}
use_kernel_diagonals (bool, optional): A bool flag indicates whether or not the lower order neighbors should be
included in the weights structure. Defaults to False.
is_inverse (bool, optional): A bool flag indicates whether or not to apply inverse on distance value. Defaults to False.
power (float, optional): The power (or exponent) of a number indicates how many times to use the number in a multiplication.
is_arc (bool, optional): A bool flag indicates if compute arc distance or Euclidean distance. Defaults to False (Euclidean distance)
is_mile (bool, optional): A bool flag indicates if the distance unit is mile or km. Defaults to True (mile).
Returns:
Weight: An instance of Weight class
'''
if kernel not in ['triangular', 'uniform', 'epanechnikov', 'quartic', 'gaussian']:
raise("The parameter 'kernel' has to be one of {'triangular', 'uniform', 'epanechnikov', 'quartic', 'gaussian'}")
use_kernel_diagonals = False if 'use_kernel_diagonals' not in kwargs else kwargs['use_kernel_diagonals']
is_arc = False if 'is_arc' not in kwargs else kwargs['is_arc']
is_mile = True if 'is_mile' not in kwargs else kwargs['is_mile']
power = 1.0 if 'power' not in kwargs else kwargs['power']
is_inverse = False if 'is_inverse' not in kwargs else kwargs['is_inverse']
poly_id = ""
gda_w = gda_distance_weights(geoda_obj.gda, bandwidth, poly_id, power, is_inverse, is_arc, is_mile, kernel, use_kernel_diagonals)
return Weight(gda_w)
[docs]def kernel_knn_weights(geoda_obj, k, kernel, **kwargs):
'''K-NN Kernel Spatial Weights
Create a kernel weights by specifying k-nearest neighbors and a kernel method
Args:
geoda_obj (geoda): An instance of geoda class.
k (int): a positive integer number for k-nearest neighbors
kernel (str): A string value, which has to be one of {'triangular', 'uniform', 'epanechnikov', 'quartic', 'gaussian'}
Adaptive_bandwidth (bool, optional): True(default) or False: True - use adaptive bandwidth calculated using distance of k-nearest neithbors, False - use max distance of all observation to their k-nearest neighbors
use_kernel_diagonals (bool, optional): A bool flag indicates whether or not the lower order neighbors should be
included in the weights structure. Defaults to False.
is_inverse (bool, optional): A bool flag indicates whether or not to apply inverse on distance value. Defaults to False.
power (float, optional): The power (or exponent) of a number indicates how many times to use the number in a multiplication.
is_arc (bool, optional): A bool flag indicates if compute arc distance or Euclidean distance. Defaults to False (Euclidean distance)
is_mile (bool, optional): A bool flag indicates if the distance unit is mile or km. Defaults to True (mile).
Returns:
Weight: An instance of Weight class
'''
if kernel not in ['triangular', 'uniform', 'epanechnikov', 'quartic', 'gaussian']:
raise("The parameter 'kernel' has to be one of 'triangular', 'uniform', 'epanechnikov', 'quartic', 'gaussian'")
adaptive_bandwidth = True if 'adaptive_bandwidth' not in kwargs else kwargs['adaptive_bandwidth']
use_kernel_diagonals = False if 'use_kernel_diagonals' not in kwargs else kwargs['use_kernel_diagonals']
power = 1.0 if 'power' not in kwargs else kwargs['power']
is_inverse = False if 'is_inverse' not in kwargs else kwargs['is_inverse']
is_arc = False if 'is_arc' not in kwargs else kwargs['is_arc']
is_mile = True if 'is_mile' not in kwargs else kwargs['is_mile']
bandwidth = 0
polyid = ""
gda_w = gda_knn_weights(geoda_obj.gda, k, power, is_inverse, is_arc, is_mile, kernel, bandwidth, adaptive_bandwidth, use_kernel_diagonals, polyid)
return Weight(gda_w)