TVAR (Threshold VAR)

[Scalper]

Will be the threshold value that share each regime??

[gordeza]

Hi Gareth,
Thanks a ot for helping thousands in this forum.
It would be greatly appreciated if you could elaborate a bit more on how to get the threshold value display in the TVAR output.
Also, how to edit the add-in code?
Thanks in advance.
Gordeza

[EViews Gareth]

The R program doesn't return the threshold value, so it can't be brought into EViews.

To edit any Add-in, simply open up the file in EViews. (File->Open->Program in Addins folder)

[agranda]

I get the very same error message that Mihaain and Gordeza mentioned early:
"R(for 'testtar<-TVAR(g, 3, include="const"model="TAR",thhresh=1)') returned an error: evaluation stopped because of an error".
Any idea? I don't find the error
(add my wf1)
Thanks in advance!

[EViews Gareth]

What was the specification you used?

[agranda]

the specification is

TVAR(options) 3 ipc_ln tc_v

[EViews Gareth]

Looks like it is just an issue with the R package for your particular data set.

[yasirjanjua]

Hello dear fellows, i have installed the R with rsc proxy and couldnot find the statconnDCOM version 3.1-1B2 if someone have plz attach as soon as possible.

[yasirjanjua]

Dear Fellows
i have run the program properly but having estimation error in the following command
> test<-TVAR(G, 2) which is not working properly
how can i resolve it
plz reply soon
thanks

[Potti_X]

Hi,

is there any way to use this addin for forecasting?

[EViews Gareth]

Not as far as I am aware.

[afham_acc]

I have a problem that says "R(for 'test<-TVAR(g, 1, include="const"model="TAR",nthresh=1)') returned an error: evaluation stopped because of an error" And I tested with the testtvar.wf1 gordeza has sent and it worked perfectly...So, i figured this must be a problem with my dataset...could u please help me with this?

I wanted to test the TVAR between g-inf and g-exrate...with lagged 1 and delay parameter 1...
is it because my sample size is too small?

[valeria]

Hello! I'm trying to estimate a TVAR with 3 endogenous variables and 2 lags. I chose the threshold variable, as a function of one of the variables of the model with 1 delay and I can make an assumption about the threshold value, setting it to 0 (as it is in the literature). Using the TVAR Add-in I'm able to activate the R-connection properly, but when I estimate the model the output displays all coefficients equal to 0. Can I specify somewhere in the estimation screen the threshold variable?
I would really appreciate help.
Thank you!
Valeria

[EViews Gareth]

No, I believe not.

[valeria]

So what is the meaning of this result?Am I doing something wrong?