I am trying to analyze spatial data of Japanese c. 1,800 municipalities, using SAR models. Japan is an archipelago, which has numerous islands. One municipality may have island areas, and it is currently impossible to calculate weighting matirx (W) and make one to one correspondence between the shapefile data and another data set for municipalities. Is there any goog methods for choosing the largest polygon or other ways out? Thank you for reading this question.
The following is a message when I tried to calculate W.
. spmatrix create contiguity Wc
J(): 3900 unable to allocate real <tmp>[121158,121158]
_SPMATRIX_contiguity_create(): - function returned error
_SPMATRIX_contiguity(): - function returned error
<istmt>: - function returned error
r(3900);
. spmatrix create idistance Widist
J(): 3900 unable to allocate real <tmp>[121158,121158]
_SPMATRIX_Euclidean_dist(): - function returned error
_SPMATRIX_idistance_compute_dist(): - function returned error
_SPMATRIX_idistance(): - function returned error
<istmt>: - function returned error
r(3900);
The following is a message when I tried to calculate W.
. spmatrix create contiguity Wc
J(): 3900 unable to allocate real <tmp>[121158,121158]
_SPMATRIX_contiguity_create(): - function returned error
_SPMATRIX_contiguity(): - function returned error
<istmt>: - function returned error
r(3900);
. spmatrix create idistance Widist
J(): 3900 unable to allocate real <tmp>[121158,121158]
_SPMATRIX_Euclidean_dist(): - function returned error
_SPMATRIX_idistance_compute_dist(): - function returned error
_SPMATRIX_idistance(): - function returned error
<istmt>: - function returned error
r(3900);
Comment