Spatial Weights

Options

Geometry column

The name of the geometry column in the input data.

ID column

Select the column which contains for each observation in the input data a unique ID, it should be an integer column. If 'none' is selected, the IDs will be automatically generated from 0 to the number of rows flowing the order of the input data. The IDs of this column must match with the values of the ID column selected in subsequent ESDA or spatial modeling nodes.

Weights category

The type of spatial weights to construct. Defaults to 'Queen'. Other options are 'Rook', 'Binary Distance Band', 'Inverse Distance', 'Lattice', 'K nearest', 'Kernel', and 'Get spatial weights matrix from file'.

• Queen which will construct a queen contiguity weights matrix, is more robust and more suitable for areal unit data. The queen criterion is somewhat more encompassing and defines neighbors as spatial units sharing a common edge or a common vertex.
• The Rook criterion defines neighbors by the existence of a common edge between two spatial units. Therefore, the number of neighbors according to the queen criterion will always be at least as large as for the rook criterion.
• When choosing K nearest, select the nearest number 'Nearest k' in the following options. K-nearest are often used for point data.
• When selecting Binary Distance Band, please select the distance threshold 'Threshold' in the following options.
• When selecting Inverse Distance, please select the distance threshold 'Threshold' and the corresponding power 'Power' in the following options.
• When 'Your own' is selected, please enter the path of the spatial weights matrix in CSV format in the following options.
• More details about spatial weights, please see the GeoDa center website.

Order

The order of the weight matrix is 1 by default. Users can change the order of the weights, higher order weights will treat further units as neighbors.

Threshold for Inverse Distance or Binary Distance Band

The distance threshold for constructing binary distance band and inverse distance weights. Defaults to 1

Rows for Lattice

The number of rows for constructing a lattice spatial weights matrix. Defaults to 5.

Columns for Lattice

The number of columns for constructing a lattice spatial weights matrix. Defaults to 5.

Nearest k

The number of nearest neighbors to use for constructing k-nearest neighbors weights. Defaults to 4.

Kernel K

The number of nearest neighbors to use for determining the bandwidth in kernel weights. Defaults to 12.

Kernel type

The type of kernel to use in constructing kernel weights. Defaults to 'triangular'

Kernel bandwidth

The type of kernel bandwidth to use in constructing kernel weights. The bandwidth of the kernel. The default is fixed. If adaptive then bandwidth is adaptive across observations.

Get spatial weights matrix from file

The file path of a user-defined spatial weights matrix in CSV format. Defaults to ''. Please enter the path of the spatial weights matrix in CSV format in the following options. The weights matrix must be in matrix format and in the order of the samples.

Views

This node has no views

Workflows

• No workflows found

Links

• No links available

Developers