利用matlab进行空间计量模型检验中,出现NaN,请问如何解决这个问题?
Ordinary Least-squares Estimates
Dependent Variable = logTC
R-squared = 0.9890
Rbar-squared = 0.9888
sigma^2 = 0.0729
Durbin-Watson = 2.0690
Nobs, Nvars = 480, 7
***************************************************************
Variable Coefficient t-statistic t-probability
intercept -0.059096 -1.243576 0.214272
logP 0.774525 51.670712 0.000000
logA 0.847940 32.705430 0.000000
logIND -0.307303 -5.217023 0.000000
logSV -0.779493 -15.645382 0.000000
logURB -0.073516 -1.904153 0.057497
logEI 0.564915 17.469041 0.000000
loglikols =
-49.0345
LM test no spatial lag, probability = NaN, NaN
robust LM test no spatial lag, probability = NaN, NaN
LM test no spatial error, probability = NaN, NaN
robust LM test no spatial error, probability = NaN, NaN
Wrong # of variable names in prt_reg -- check vnames argument
will use generic variable names
prt_reg(results,vnames,1); 这里有问题吗