Table 1.
The threshold values of Moran’s index, Geary’s coefficient and the revised results.
Table 2.
Comparison of the advantages and disadvantages of the four methods.
Figure 1.
A flow chart of data processing, parameter estimation, and autocorrelation analysis.
Figure 2.
A sketch map of the geographic relationship of the principal cities of China.
Table 3.
Classification of spatial autocorrelation based on population size of the principal cities in China (2000).
Figure 3.
The improved Moran’s scatterplots with trendlines of spatial autocorrelation for the principal cities of China (2000).
[a. Based on inverse power function; b. Based on negative exponential function].
Figure 4.
The inverse Moran’s scatterplots with trendlines of spatial autocorrelation for the principal cities of China (2000).
[a. Based on inverse power function; b. Based on negative exponential function].
Figure 5.
The normal histograms of error distributions based on different spatial weight functions for China’s cities (2000).
(Note: These graphs are created using standardized error series. The filled bars represent the actual distributions based on observed values, while the unfilled bars with double frames represent the normal distributions based on the expected values, which form bell-shaped histograms.) [a. Based on inverse power function; b. Based on negative exponential function].
Table 4.
Moran’s index and Geary’s coefficient values based on spatial population and sample of Chinese cities (2000).