Thanks for all the helpful thoughts and comments everyone! : )
On Tue, Dec 13, 2011 at 9:11 PM, John C Frain wrote:
> One should not even think of estimating an 11 variable VAR(4) with 100
> observations. Standard errors will be so large that any null will be
> acceptable an tests will have n
Allin Cottrell,The scatterplot with v1 on y axis and v2 on x axis gives y=17.9+0.0453xUsing ols v1 v2 OR ols v2 v1 gives 0.181744 and 5.49601 coefficient which is differentthe scatterplotAnd the graph for selecting v1 against v2 fiitted vs obs is different from the graph from the scatterplotWanted
You forgot the constant.
-sven
On 12/14/2011 08:25 AM, clarodina(a)lycos.com wrote:
> Allin Cottrell,
>
> The scatterplot with v1 on y axis and v2 on x axis gives y=17.9+0.0453x
>
> Using ols v1 v2 OR ols v2 v1 gives 0.181744 and 5.49601 coefficient
> which is different
> the scatterplot
>
> A
Hi,
I have the following problem. I should estimate a VAR model in which some
parameters are constrained. How can I restrict some coefficients equal to 0 in
gretl?
Best regards
Alessandro
On 12/14/2011 11:24 AM, alexkakashi(a)libero.it wrote:
> Hi,
> I have the following problem. I should estimate a VAR model in which some
> parameters are constrained. How can I restrict some coefficients equal to 0
> in
> gretl?
>
set up a SUR system:
system method=sur
equation ...
equati
Don't forget to add the constant in the regression!
Jan
-Original Message-
From: gretl-users-bounces(a)lists.wfu.edu
[mailto:gretl-users-bounces(a)lists.wfu.edu] On Behalf Of
gretl-users-request(a)lists.wfu.edu
Sent: woensdag 14 december 2011 8:26
To: gretl-users(a)lists.wfu.edu
Subjec
On Wed, 14 Dec 2011, Sven Schreiber wrote:
> On 12/14/2011 11:24 AM, alexkakashi(a)libero.it wrote:
>> Hi,
>> I have the following problem. I should estimate a VAR model in which some
>> parameters are constrained. How can I restrict some coefficients equal to 0
>> in
>> gretl?
>
> set up a SUR s
On Wed, 14 Dec 2011, clarodina(a)lycos.com wrote:
[ nothing readable other than attached HTML ]
Please do not send HTML mail to the gretl list.
But in relation to "Wanted to save the residual against v2
from the scatterplot and have the value on the gretl How to do
using GUI rather the command
The model for adf is (1-L)y = b0 + (a-1)*y(-1) + eBut isn't adf is using reg of change of y to lag of y? The (1-L)y is not refering to change of y Clicking the use first difference generate the same (1-L)y rather using change of y What does L represent?Clarodina
Clarodina,
L is the lag operator: Ly(t) = y(t-1). So (1-L)y = y(t) – Ly(t) = y(t) –
y(t-1). The left-hand side of the adf regression is then just the first
difference of y(t).
HTH,
PS
From: gretl-users-bounces(a)lists.wfu.edu
[mailto:gretl-users-bounces(a)lists.wfu.edu] On Behalf Of clarodin
You should look for "L operator".
On Wed, Dec 14, 2011 at 4:17 PM, clarodina(a)lycos.com
wrote:
> The model for adf is (1-L)y = b0 + (a-1)*y(-1) + e
>
> But isn't adf is using reg of change of y to lag of y?
>
> The (1-L)y is not refering to change of y
>
> Clicking the use first difference gen
On 12/14/2011 07:17 PM, clarodina(a)lycos.com wrote:
> The model for adf is (1-L)y = b0 + (a-1)*y(-1) + e
When you start a new topic/thread, I think it's good practice to start
with "hello", or "hi", or something like that...
>
> But isn't adf is using reg of change of y to lag of y?
yes
>
>
Thanks for your answer. In my equations the restrictions are non-linear. For
example, can I use the following restriction
restrict varsys
> b[1,2]*b[2,2] = 0
>end restrict
Best regards.
Alessandro
>Messaggio originale
>Da: gretl-users-request(a)lists.wfu.edu
>Data: 14/12/2011 18.00
>
arset="us-ascii"An HTML attachment was scrubbed...URL: http://lists.wfu.edu/pipermail/gretl-users/attachments/20111214/ba378a5e/attachment-0001.html --Message: 2Date: Wed, 14 Dec 2011 18:22:09 +From: "Summers, Peter" <psumm...@highpo
No, non-linear restrictions are not available for system estimators
(yet), AFAIK.
Of course there are workarounds, for example specifying an equivalent
GMM block, where I think it is possible now to test nonlinear
restrictions. But this is admittedly not the most friendly way.
cheers,
sven
On 12
On 12/14/2011 08:44 PM, clarodina(a)lycos.com wrote:
> Should adf be done on the residual against time or the residual from
> scatter plot?
>
I think this is the right moment to refer you to the great user guide
that gretl has (aka RTFM).
happy testing,
sven
On Wed, 14 Dec 2011, Sven Schreiber wrote:
> No, non-linear restrictions are not available for system estimators
> (yet), AFAIK.
>
> Of course there are workarounds, for example specifying an equivalent
> GMM block, where I think it is possible now to test nonlinear
> restrictions. But this is adm
17 matches
Mail list logo
