# st: =?utf-8?Q?st:_variance-covariance_of_nonlinear_combination_in_matrix_form?==?utf-8?Q?=E2=80=8F=E2=80=8F?=

 From =?utf-8?Q?=E7=94=B0=E4=B8=AD_=E8=89=AF=E5=AD=90?= <[email protected]> To <[email protected]> Subject st: =?utf-8?Q?st:_variance-covariance_of_nonlinear_combination_in_matrix_form?==?utf-8?Q?=E2=80=8F=E2=80=8F?= Date Fri, 21 Dec 2007 21:52:55 +0900

```Dear Statalist

I need to test nonlinear combination shown in MATRIX form.

I estimated the following equation.

A(t)=aX(t)+cA(t-1)     (1)

A(t): 7*1 dependent variables
a:     7*24 parameters
X(t): 24*1 independent variables
c:     7*7 parameters
A(t-1): 7*1 independent variables

using the estimation results for eq.(1), i want to test inv(I-c)a in eq.(2), which is parameters for X(t)

given A(t)=A(t-1), (1) becomes
(I-c)A(t)=aX(t)
A(t)=inv(I-c)aX(t)        (2)

there is a command "nlcom" for nonlinear combination of parameters, but, as you know, it is not for matrix, but scalar.
i can calculate each element for inv(I-c) to use it in "nlcom." but the expression for each element is too long to be available in "nlcom."
could anyone please solve this problem? i would be soooo happy if you could do that.

yoshi

_________________________________________________________________

http://match.jp.msn.com/channel/index.aspx?avo=1&trackingid=1034447

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
```