不知道论坛里有没有空间计量方面的高手,最近在用stata做空间计量方面的东西。help了一下xsmle发现这个能做很多东西,看了一下帮助发现:做SAR模型的固定效应时是可以分成SAR的个体固定效应、SAR的时间固定效应以及SAR的双向固定效应的,Syntax
Spatial Autoregressive (SAR) model
xsmle depvar [indepvars] [if] [in] [weight] , wmat(name) model(sar) [SAR_options]
Spatial Durbin (SDM) model
xsmle depvar [indepvars] [if] [in] [weight] , wmat(name) model(sdm) [SDM_options]
Spatial Autocorrelation (SAC) model
xsmle depvar [indepvars] [if] [in] [weight] , wmat(name) emat(name) model(sac) [SAC_options]
Spatial Error (SEM) model
xsmle depvar [indepvars] [if] [in] [weight] , emat(name) model(sem) [SEM_options]
Generalised Spatial Panel Random Effects (GSPRE) model
xsmle depvar [indepvars] [if] [in] [weight] , wmat(name) model(gspre) [emat(name) GSPRE_options]
SAR_options Description
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Model
wmat(name) specify the weight matrix for the spatial-autoregressive term
re use the random effects estimator; the default
fe use the fixed effects estimator
type(type_options[, leeyu]) specify fixed-effects type
Type options Description
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
ind for individual fixed effects effects
time for time fixed effect
both for both time and individual fixed effects
然后按照例子输入了一下命令:xsmle x lnp lnnights lnincome lnage lneducation, wmat(w) model(sar) fe ind
xsmle x lnp lnnights lnincome lnage lneducation, wmat(w) model(sar) fe time
xsmle x lnp lnnights lnincome lnage lneducation, wmat(w) model(sar) fe both
但是这个三个做出来的结果是一模一样,我不知道哪里错了,哪位高手能够指点一下?