. help spregdpd
. clear all
. sysuse spregdpd.dta, clear
. spregdpd y x1 x2 , nc(7) wmfile(SPWxt) model(sar) run(xtdhp) mfx(lin) test re
==============================================================================
*** Binary (0/1) Weight Matrix: 49x49 - NC=7 NT=7 (Non Normalized)
==============================================================================
* Spatial Lag Han-Philips Linear Dynamic Panel Data Regression
==============================================================================
y = w1y_y + x1 + x2
------------------------------------------------------------------------------
Sample Size = 42 | Cross Sections Number = 7
Wald Test = 52.2355 | P-Value > Chi2(4) = 0.0000
F-Test = 13.0589 | P-Value > F(4 , 38) = 0.0000
(Buse 1973) R2 = 0.5789 | Raw Moments R2 = 0.9661
(Buse 1973) R2 Adj = 0.5456 | Raw Moments R2 Adj = 0.9635
Root MSE (Sigma) = 13.3907 | Log Likelihood Function = -142.5329
------------------------------------------------------------------------------
- R2h= 0.4614 R2h Adj= 0.4188 F-Test = 7.92 P-Value > F(4 , 38) 0.0001
- R2v= 0.4431 R2v Adj= 0.3992 F-Test = 7.36 P-Value > F(4 , 38) 0.0002
------------------------------------------------------------------------------
y | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
y |
L1. | .0267714 .2782332 0.10 0.924 -.5364823 .5900251
|
w1y_y | -.1229202 .0692124 -1.78 0.084 -.2630333 .017193
x1 | -.2951249 .0833482 -3.54 0.001 -.4638545 -.1263953
x2 | -.8707571 .3158102 -2.76 0.009 -1.510081 -.2314327
_cons | 69.89907 7.676856 9.11 0.000 54.35808 85.44005
------------------------------------------------------------------------------
Rho Value = -0.1229 Chi2 Test = 3.154 P-Value > Chi2(1) 0.0837
------------------------------------------------------------------------------
==============================================================================
* Panel Model Selection Diagnostic Criteria
==============================================================================
- Log Likelihood Function LLF = -142.5329
---------------------------------------------------------------------------
- Akaike Information Criterion (1974) AIC = 75.9713
- Akaike Information Criterion (1973) Log AIC = 4.3304
---------------------------------------------------------------------------
- Schwarz Criterion (1978) SC = 105.7777
- Schwarz Criterion (1978) Log SC = 4.6613
---------------------------------------------------------------------------
- Amemiya Prediction Criterion (1969) FPE = 68.2952
- Hannan-Quinn Criterion (1979) HQ = 85.7704
- Rice Criterion (1984) Rice = 83.8455
- Shibata Criterion (1981) Shibata = 71.6775
- Craven-Wahba Generalized Cross Validation (1979) GCV = 79.2035
------------------------------------------------------------------------------
==============================================================================
*** Spatial Panel Aautocorrelation Tests
==============================================================================
Ho: Error has No Spatial AutoCorrelation
Ha: Error has Spatial AutoCorrelation
- GLOBAL Moran MI = 0.1519 P-Value > Z( 1.442) 0.1493
- GLOBAL Geary GC = 0.8314 P-Value > Z(-1.132) 0.2576
- GLOBAL Getis-Ords GO = -0.4341 P-Value > Z(-1.442) 0.1493
------------------------------------------------------------------------------
- Moran MI Error Test = 0.7885 P-Value > Z(6.648) 0.4304
------------------------------------------------------------------------------
- LM Error (Burridge) = 1.1335 P-Value > Chi2(1) 0.2870
- LM Error (Robust) = 4.2647 P-Value > Chi2(1) 0.0389
------------------------------------------------------------------------------
Ho: Spatial Lagged Dependent Variable has No Spatial AutoCorrelation
Ha: Spatial Lagged Dependent Variable has Spatial AutoCorrelation
- LM Lag (Anselin) = 0.2512 P-Value > Chi2(1) 0.6163
- LM Lag (Robust) = 3.3823 P-Value > Chi2(1) 0.0659
------------------------------------------------------------------------------
Ho: No General Spatial AutoCorrelation
Ha: General Spatial AutoCorrelation
* Marginal Effect - Elasticity (Model= sar): Linear *
+---------------------------------------------------------------------------+
| Variable | Marginal_Effect(B) | Elasticity(Es) | Mean |
|------------+--------------------+--------------------+--------------------|
| L.y | 0.0268 | 0.0265 | 34.7923 |
| w1y_y | -0.1229 | -0.3499 | 100.0064 |
| x1 | -0.2951 | -0.3229 | 38.4362 |
| x2 | -0.8708 | -0.3563 | 14.3749 |
+---------------------------------------------------------------------------+
Mean of Dependent Variable = 35.1288
tulipsliu 写于 2016年5月30日
从命令来看,是可以从网上下载的命令(看你使用的STATA版本,以前我用12版本,必须用search all 找到命令;到了STATA14版本,【这个版本我从论坛下载的】,直接在命令窗口输入 help spregdpd 就可以找到命令来源,可以下载);
一、数据整理:
两本部分,
1.1 经济数据,这个是面板数据格式,注意得保证数据是按找横切面的每一年,先2000年完整整理好这一年的,接着才是第二年,第三年的数据,如此类推。MATLAB里也是这样排列,数据可以先用 Excel 表格整理好。
2.2 空间权重W矩阵文件,这个命令里, wmfile(SPWxt) 这个选项会打开一个stata 格式的文件,实际上除了刚才必须先用 use 打开的数据文件,命令了调用了另一个数据文件,而第二个恰恰就是很重要的空间矩阵;
二、建模
依据命令开始估计模型;
三、观察模型的检验值
主要有 moran 检验
如果使用了 GMM 估计,甚至会有 sargan 过度识别检验;如下面这里粘贴的另一个调用时的结果:
Over Identification Restrictions Test
Ho: Over Identification Restrictions are Valid
- Sargan Over Identification LM Test = 18.987 P-Value > Chi2(17) 0.3293
------------------------------------------------------------------------------
大致如此,剩下的,是个人的论文里该怎么写和怎么使用程序估计了的结果的问题。




雷达卡





,我是stata小白
京公网安备 11010802022788号







