呵呵，我是个新手，麻烦问下该怎么求助版主啊？

对于横截面自相关你还是先找到文献
查查计算公式吧

逍遥梦蝶  说的没错，空间自相关好解释，
一般的横截面数据 你能说明没什么有自相关吗？

所以，你最好先查好理论，然后再说软件的事情
许多情况下需要自己更加公式编程

---------------------------------------------------------------------------------------------------------------------------------
help for spatcorr                                                                                                 (STB-60: sg162)
---------------------------------------------------------------------------------------------------------------------------------

Spatial correlogram

        spatcorr varname , bands(numlist) xcoord(varname) ycoord(varname) [ geary cumulative twotail graph needle
              savegraph(filename [, replace]) ]


Description

spatcorr computes and optionally plots Moran's I or Geary's c spatial correlogram based on two or more consecutive or cumulative
distance bands. For each distance band, spatcorr computes and displays in tabular form the requested statistic, the expected
value of the statistic under the null hypothesis of global spatial independence, the standard deviation of the statistic, the
z-value, and the corresponding 1-tail or 2-tail p-value.


Options

bands(numlist) is always required. It specifies the sequence of lower and upper bounds of the distance bands to be used in the
    calculation of the spatial correlogram. For example, if option bands(0 2 4 6) is specified and option cumulative is not
    specified, then the spatial correlogram will be computed on the following consecutive distance bands: (0-2], (2-4], (4-6]. On
    the other hand, if option cumulative is specified, then the spatial correlogram will be computed on the following cumulative
    distance bands: (0-2], (0-4], (0-6].

xcoord(varname) is always required. It specifies the name of the variable containing the x-coordinate of each location object of
    analysis.

ycoord(varname) is always required. It specifies the name of the variable containing the y-coordinate of each location object of
    analysis. Both the x-coordinate and the y-coordinate must be expressed in projected units, e.g., meters, kilometers, miles,
    or arbitrary digitizing units.

geary requests that Geary's c spatial correlogram be computed and optionally plotted instead of the default Moran's I spatial
    correlogram.

cumulative requests that cumulative distance bands be used instead of the default consecutive distance bands.

twotail requests that 2-tail p-values be computed and displayed instead of the default 1-tail p-values.

graph requests that the spatial correlogram be plotted.

needle requests that the spatial correlogram be plotted using vertical lines from zero (if Moran's I) or one (if Geary's c) to
    the computed statistics instead of lines that connect the computed statistics.

savegraph(filename [, replace]) requests that the graph be saved in filename. If filename exists, an error will occur unless
    replace is also specified. filename must have one of the following extensions: .ps, .eps, .prn, or .wmf.


Examples

        . spatcorr crime, bands(0(1)5) xcoord(x) ycoord(y)

        . spatcorr crime, bands(0(1)5) xcoord(x) ycoord(y) cumulative

        . spatcorr crime, bands(0(1)5) xcoord(x) ycoord(y) graph


Author

        Maurizio Pisati
        Department of Sociology and Social Research
        University of Milano Bicocca - Italy
        maurizio.pisati@galactica.it


Also see

On-line:  help for spatwmat, spatgsa, spatlsa, spatdiag, spatreg if installed

DW检验 在stata 中，只在时间序列中才能进行序列相关检验，因此截面数据中仍然要生成一个伪时 间变量，并且把数据设置为时间序列。而时间序列中则可以直接进行dw 检验。DW 检验只 适用于一阶自相关。 ————————————模板———————————— quietly regress 被解释变量 解释变量若干 generate t=_n tset t estat dwatson ————————————模板———————————— 当数据时截面数据时，需要generate 和tset 语句： Generate 生成伪时间变量t；tset 将数据设置为时间序列数据。

请问空间自相关是怎么解释的啊？截面数据如何检验是否有自相关，然后如何解决呢？用R