sir,

i thought that the model and respondent estimation is available and feasible on condition that it is imposed with

a set of assumptions .

unluckly,i've never exployed GAUSS as my econometric tool.

u can email me with the related papers u quoted in ur research if u really demand much a discussion

about it.

best wishes!

sorry for my poor english:

1. here is some webs for spacial economics:http://www.statistical.org/;

2. and i have some e-books introduce spacial economics used matlab as main tool;but this web doesn't allowed upload the paper size upto 150k.

3. pro. kuan-pinlin has done some works in his web http://www.econ.pdx.edu/staff/KPL/, so, i like use GAUSS to study econometrics. there are more free codes for panel date and spacial analysis.

so. my questions:

if it is very difficult to expend panel data analysis to spacial analysis?

which software do you select as your econometric tool? why?

sir,iam zhangang

for u first question,in my opinion. the answer is yes.it is very difficult to estimate the spatial model for panel data unless the model is imposed with a set of assumption such as it that at least the spatial lagged term or the spatial autocorrelated error term and the time lagged or error term would be parameterized,and even if the two sides may be parameterized,we would face the question about how to specify the model.

in a word ,it's really a nuisance!

for u second question. SAS is the only one for my econnometric study.i'd employed SAS successfully to transform the spatial data to spatial weight matrix and with it,i've finished the estimation and test for two typical models in spatial econometrics.

P.S.

some one had stolen zhangang's id so that u may email the details in ur rresearch

to baicheng2003@126.com if u wanna a deep discussion.

best regards!

zhangang