> library(tsDyn)
> A <- read.table("data.txt",col.names=c("date","t","price1","price2","price3","price4","price5","price6","price7","price8","price9","price10","price11"))
> mod.lstar <- lstar(A$price4,m=1,control=list(maxit=3000))
Using maximum autoregressive order for low regime: mL = 1
Using maximum autoregressive order for high regime: mH = 1
Using default threshold variable: thDelay=0
Performing grid search for starting values...
Starting values fixed: gamma =  3.538462 , th =  9.236948 ; SSE =  21060.39
Optimization algorithm converged
Optimized values fixed for regime 2  : gamma =  37.7138 , th =  12.17005 ; SSE =  20357.32
> summary(mod.lstar)

Non linear autoregressive model

LSTAR model
Coefficients:
Low regime:
   const1    phi1.1
0.2974328 0.3633679

High regime:
   const2    phi2.1
37.751052 -2.509213

Smoothing parameter: gamma = 37.71

Threshold
Variable: Z(t) = + (1) X(t)

Value: 12.17

Residuals:
      Min        1Q    Median        3Q       Max
-21.36528  -3.90579  -0.26034   4.02009  22.78858

Fit:
residuals variance = 40.55,  AIC = 1871, MAPE = 120.7%

Coefficient(s):
          Estimate  Std. Error    t value  Pr(>|z|)   
const1  0.29743282  0.29347916     1.0135    0.3108   
phi1.1  0.36336788  0.04858111     7.4796 7.461e-14 ***
const2 37.75105188  6.41464076     5.8851 3.977e-09 ***
phi2.1 -2.50921301  0.39707585    -6.3192 2.629e-10 ***
gamma  37.71380290          NA         NA        NA   
th     12.17005093  0.00097296 12508.2756 < 2.2e-16 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Non-linearity test of full-order LSTAR model against full-order AR model
F = 11.015 ; p-value = 0.00097013

Threshold
Variable: Z(t) = + (1) X(t)
Warning messages:
1: In vcov.lstar(object) : Hessian negative-semi definite

2: In sqrt(diag(vc)) : NaNs produced

此次运行结果是重启了R，再运行的代码。
之前对lstar.R进行过修改，并运行过，但我把它放在另一个包了。
为什么会出现这个警告信息呢？怎么修改呢？（在其他电脑上运行，没有这个警告信息，可能是修改的文件影响了结果？）