6:08 p.m.
Dear gretl users,
I would like to obtain simulated data by bootstrapping the residuals of a
VAR model. I have gave a look to the gretl guide, and I found something
related to resampling and bootstrapping (section 7.4), actually, I would
like to adapt the following example - 7.1 in the guide - (with ols) to the
VAR framework:
ols y 0 x
genr ui = $uhat
scalar ybar = mean(y)
scalar replics = 2000
scalar tcount = 0
series ysim = 0
loop replics
ysim = ybar + resample(ui)
ols ysim 0 x
scalar tsim = abs($ coeff (x) / $stderr(x))
tcount +=(tsim > 2.5)
endloop
I have started to write a code.. but my "first" problem is how to save the
matrix of residuals and then generate the new dataset. The next one I
guess will be how to store 2000 Irfs! But let's go step by step..
Would someone have any suggestion?
Thanks in advance.
Gabriela
6:40 p.m.
On Fri, 14 Jun 2013, Gabriela Nodari wrote:
I would like to obtain simulated data by bootstrapping the residuals
of a
VAR model. I have gave a look to the gretl guide, and I found something
related to resampling and bootstrapping (section 7.4) [...]
Check out the varsimul() function. Here's a simple example of
use:
<hansl>
open data9-7
var 4 PRIME UNEMP --silent
matrix A = $compan[1:2,]
# residuals + const
matrix U = $uhat .+ $coeff[1,]
# initial values
matrix y0 = {PRIME, UNEMP}[1:4,]
# simulate the VAR
matrix X = varsimul(A, U, y0)
# check that we got back the original data
# (to within machine precision)
eval X - {PRIME, UNEMP}
# now resample U
matrix Us = resample(U)
matrix Xs = varsimul(A, Us, y0)
# look at the simulated data
print Xs
</hansl>
Use "set seed" if you want to be able to replicate the
resampling. Put the last few lines in a loop for multiple
simulations.
Allin Cottrell
6:46 p.m.
Thank you very much! I will give a look to the varsimul function!!
On 14/06/2013 8:40 PM, "Allin Cottrell" <cottrell(a)wfu.edu> wrote:
On Fri, 14 Jun 2013, Gabriela Nodari wrote:
> I would like to obtain simulated data by bootstrapping the residuals of a
> VAR model. I have gave a look to the gretl guide, and I found something
> related to resampling and bootstrapping (section 7.4) [...]
Check out the varsimul() function. Here's a simple example of
use:
<hansl>
open data9-7
var 4 PRIME UNEMP --silent
matrix A = $compan[1:2,]
# residuals + const
matrix U = $uhat .+ $coeff[1,]
# initial values
matrix y0 = {PRIME, UNEMP}[1:4,]
# simulate the VAR
matrix X = varsimul(A, U, y0)
# check that we got back the original data
# (to within machine precision)
eval X - {PRIME, UNEMP}
# now resample U
matrix Us = resample(U)
matrix Xs = varsimul(A, Us, y0)
# look at the simulated data
print Xs
</hansl>
Use "set seed" if you want to be able to replicate the
resampling. Put the last few lines in a loop for multiple
simulations.
Allin Cottrell
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
11:24 a.m.
Dear all,
I have run the varsimul() function and it works well!
Now I am having some problems to estimate the var with simulated data..
The code I am running is the following:
scalar p = 6
list y = 8 13 12 1 2 3
var p y -- silent
matrix Ac = $compan[1:6,]
matrix U = $uhat.+$coeff[1,]
matrix y0 = {var1, var2, var3, var4, var5, var6}[1:6,]
list y1 = 8 13 12 1 2
list x = 3
var p y1; x
matrix Ac1 = $compan[1:5,]
matrix U1 = $uhat.+$coeff[1,]
matrix y01 = {var1, var2, var3, var4, var5}[1:6,]
loop (n. of replics)
matrix Usim = resample(U)
matrix Ysim = varsimul(Ac, Usim, y0)
matrix Usim1 = resample(U1)
matrix Ysim1 = vasimul(Ac1,Usim1,y01)
series frs = Ysim[,1]
series ips= Ysim[,2]
...
series baas=Ysim[,6]
var p frs ips ... baas
....
Here I get error: invalid option '--' var p frs...
I guess it is because I cannot enter these series in the var command..
Could someone help me? How can I estimate the var with the simulated data?
Thanks in advance..
Gabriela
2013/6/14 Gabriela Nodari <gabriela.nodari(a)gmail.com>
Thank you very much! I will give a look to the varsimul function!!
On 14/06/2013 8:40 PM, "Allin Cottrell" <cottrell(a)wfu.edu> wrote:
> On Fri, 14 Jun 2013, Gabriela Nodari wrote:
>
> > I would like to obtain simulated data by bootstrapping the residuals of
> a
> > VAR model. I have gave a look to the gretl guide, and I found something
> > related to resampling and bootstrapping (section 7.4) [...]
>
> Check out the varsimul() function. Here's a simple example of
> use:
>
> <hansl>
> open data9-7
> var 4 PRIME UNEMP --silent
> matrix A = $compan[1:2,]
> # residuals + const
> matrix U = $uhat .+ $coeff[1,]
> # initial values
> matrix y0 = {PRIME, UNEMP}[1:4,]
> # simulate the VAR
> matrix X = varsimul(A, U, y0)
> # check that we got back the original data
> # (to within machine precision)
> eval X - {PRIME, UNEMP}
> # now resample U
> matrix Us = resample(U)
> matrix Xs = varsimul(A, Us, y0)
> # look at the simulated data
> print Xs
> </hansl>
>
> Use "set seed" if you want to be able to replicate the
> resampling. Put the last few lines in a loop for multiple
> simulations.
>
> Allin Cottrell
>
>
> _______________________________________________
> Gretl-users mailing list
> Gretl-users(a)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-users
>
11:32 a.m.
You should delete the space in "var p y -- silent". So it should be "var p
y --silent" instead.
Artur
2013/6/17 Gabriela Nodari <gabriela.nodari(a)gmail.com>
Dear all,
I have run the varsimul() function and it works well!
Now I am having some problems to estimate the var with simulated data..
The code I am running is the following:
scalar p = 6
list y = 8 13 12 1 2 3
var p y -- silent
matrix Ac = $compan[1:6,]
matrix U = $uhat.+$coeff[1,]
matrix y0 = {var1, var2, var3, var4, var5, var6}[1:6,]
list y1 = 8 13 12 1 2
list x = 3
var p y1; x
matrix Ac1 = $compan[1:5,]
matrix U1 = $uhat.+$coeff[1,]
matrix y01 = {var1, var2, var3, var4, var5}[1:6,]
loop (n. of replics)
matrix Usim = resample(U)
matrix Ysim = varsimul(Ac, Usim, y0)
matrix Usim1 = resample(U1)
matrix Ysim1 = vasimul(Ac1,Usim1,y01)
series frs = Ysim[,1]
series ips= Ysim[,2]
...
series baas=Ysim[,6]
var p frs ips ... baas
....
Here I get error: invalid option '--' var p frs...
I guess it is because I cannot enter these series in the var command..
Could someone help me? How can I estimate the var with the simulated data?
Thanks in advance..
Gabriela
2013/6/14 Gabriela Nodari <gabriela.nodari(a)gmail.com>
> Thank you very much! I will give a look to the varsimul function!!
> On 14/06/2013 8:40 PM, "Allin Cottrell" <cottrell(a)wfu.edu> wrote:
>
>> On Fri, 14 Jun 2013, Gabriela Nodari wrote:
>>
>> > I would like to obtain simulated data by bootstrapping the residuals
>> of a
>> > VAR model. I have gave a look to the gretl guide, and I found something
>> > related to resampling and bootstrapping (section 7.4) [...]
>>
>> Check out the varsimul() function. Here's a simple example of
>> use:
>>
>> <hansl>
>> open data9-7
>> var 4 PRIME UNEMP --silent
>> matrix A = $compan[1:2,]
>> # residuals + const
>> matrix U = $uhat .+ $coeff[1,]
>> # initial values
>> matrix y0 = {PRIME, UNEMP}[1:4,]
>> # simulate the VAR
>> matrix X = varsimul(A, U, y0)
>> # check that we got back the original data
>> # (to within machine precision)
>> eval X - {PRIME, UNEMP}
>> # now resample U
>> matrix Us = resample(U)
>> matrix Xs = varsimul(A, Us, y0)
>> # look at the simulated data
>> print Xs
>> </hansl>
>>
>> Use "set seed" if you want to be able to replicate the
>> resampling. Put the last few lines in a loop for multiple
>> simulations.
>>
>> Allin Cottrell
>>
>>
>> _______________________________________________
>> Gretl-users mailing list
>> Gretl-users(a)lists.wfu.edu
>> http://lists.wfu.edu/mailman/listinfo/gretl-users
>>
>
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
11:45 a.m.
right!! thanks..
Now the problem is with
set horizon = 60
error message: syntax error in command line...
2013/6/17 Artur Tarassow <artur.tarassow(a)googlemail.com>
You should delete the space in "var p y -- silent". So it
should be "var p
y --silent" instead.
Artur
2013/6/17 Gabriela Nodari <gabriela.nodari(a)gmail.com>
> Dear all,
>
> I have run the varsimul() function and it works well!
>
> Now I am having some problems to estimate the var with simulated data..
>
> The code I am running is the following:
>
> scalar p = 6
> list y = 8 13 12 1 2 3
>
> var p y -- silent
> matrix Ac = $compan[1:6,]
> matrix U = $uhat.+$coeff[1,]
> matrix y0 = {var1, var2, var3, var4, var5, var6}[1:6,]
>
> list y1 = 8 13 12 1 2
> list x = 3
>
> var p y1; x
> matrix Ac1 = $compan[1:5,]
> matrix U1 = $uhat.+$coeff[1,]
> matrix y01 = {var1, var2, var3, var4, var5}[1:6,]
>
> loop (n. of replics)
> matrix Usim = resample(U)
> matrix Ysim = varsimul(Ac, Usim, y0)
> matrix Usim1 = resample(U1)
> matrix Ysim1 = vasimul(Ac1,Usim1,y01)
>
> series frs = Ysim[,1]
> series ips= Ysim[,2]
> ...
> series baas=Ysim[,6]
>
> var p frs ips ... baas
> ....
>
> Here I get error: invalid option '--' var p frs...
>
> I guess it is because I cannot enter these series in the var command..
>
>
> Could someone help me? How can I estimate the var with the simulated
> data?
>
> Thanks in advance..
> Gabriela
>
>
>
> 2013/6/14 Gabriela Nodari <gabriela.nodari(a)gmail.com>
>
>> Thank you very much! I will give a look to the varsimul function!!
>> On 14/06/2013 8:40 PM, "Allin Cottrell" <cottrell(a)wfu.edu>
wrote:
>>
>>> On Fri, 14 Jun 2013, Gabriela Nodari wrote:
>>>
>>> > I would like to obtain simulated data by bootstrapping the residuals
>>> of a
>>> > VAR model. I have gave a look to the gretl guide, and I found
>>> something
>>> > related to resampling and bootstrapping (section 7.4) [...]
>>>
>>> Check out the varsimul() function. Here's a simple example of
>>> use:
>>>
>>> <hansl>
>>> open data9-7
>>> var 4 PRIME UNEMP --silent
>>> matrix A = $compan[1:2,]
>>> # residuals + const
>>> matrix U = $uhat .+ $coeff[1,]
>>> # initial values
>>> matrix y0 = {PRIME, UNEMP}[1:4,]
>>> # simulate the VAR
>>> matrix X = varsimul(A, U, y0)
>>> # check that we got back the original data
>>> # (to within machine precision)
>>> eval X - {PRIME, UNEMP}
>>> # now resample U
>>> matrix Us = resample(U)
>>> matrix Xs = varsimul(A, Us, y0)
>>> # look at the simulated data
>>> print Xs
>>> </hansl>
>>>
>>> Use "set seed" if you want to be able to replicate the
>>> resampling. Put the last few lines in a loop for multiple
>>> simulations.
>>>
>>> Allin Cottrell
>>>
>>>
>>> _______________________________________________
>>> Gretl-users mailing list
>>> Gretl-users(a)lists.wfu.edu
>>> http://lists.wfu.edu/mailman/listinfo/gretl-users
>>>
>>
>
> _______________________________________________
> Gretl-users mailing list
> Gretl-users(a)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-users
>
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
11:47 a.m.
It should be simply
set horizon 60
2013/6/17 Gabriela Nodari <gabriela.nodari(a)gmail.com>
right!! thanks..
Now the problem is with
set horizon = 60
error message: syntax error in command line...
2013/6/17 Artur Tarassow <artur.tarassow(a)googlemail.com>
> You should delete the space in "var p y -- silent". So it should be
"var
> p y --silent" instead.
>
> Artur
>
>
> 2013/6/17 Gabriela Nodari <gabriela.nodari(a)gmail.com>
>
>> Dear all,
>>
>> I have run the varsimul() function and it works well!
>>
>> Now I am having some problems to estimate the var with simulated data..
>>
>> The code I am running is the following:
>>
>> scalar p = 6
>> list y = 8 13 12 1 2 3
>>
>> var p y -- silent
>> matrix Ac = $compan[1:6,]
>> matrix U = $uhat.+$coeff[1,]
>> matrix y0 = {var1, var2, var3, var4, var5, var6}[1:6,]
>>
>> list y1 = 8 13 12 1 2
>> list x = 3
>>
>> var p y1; x
>> matrix Ac1 = $compan[1:5,]
>> matrix U1 = $uhat.+$coeff[1,]
>> matrix y01 = {var1, var2, var3, var4, var5}[1:6,]
>>
>> loop (n. of replics)
>> matrix Usim = resample(U)
>> matrix Ysim = varsimul(Ac, Usim, y0)
>> matrix Usim1 = resample(U1)
>> matrix Ysim1 = vasimul(Ac1,Usim1,y01)
>>
>> series frs = Ysim[,1]
>> series ips= Ysim[,2]
>> ...
>> series baas=Ysim[,6]
>>
>> var p frs ips ... baas
>> ....
>>
>> Here I get error: invalid option '--' var p frs...
>>
>> I guess it is because I cannot enter these series in the var command..
>>
>>
>> Could someone help me? How can I estimate the var with the simulated
>> data?
>>
>> Thanks in advance..
>> Gabriela
>>
>>
>>
>> 2013/6/14 Gabriela Nodari <gabriela.nodari(a)gmail.com>
>>
>>> Thank you very much! I will give a look to the varsimul function!!
>>> On 14/06/2013 8:40 PM, "Allin Cottrell" <cottrell(a)wfu.edu>
wrote:
>>>
>>>> On Fri, 14 Jun 2013, Gabriela Nodari wrote:
>>>>
>>>> > I would like to obtain simulated data by bootstrapping the
residuals
>>>> of a
>>>> > VAR model. I have gave a look to the gretl guide, and I found
>>>> something
>>>> > related to resampling and bootstrapping (section 7.4) [...]
>>>>
>>>> Check out the varsimul() function. Here's a simple example of
>>>> use:
>>>>
>>>> <hansl>
>>>> open data9-7
>>>> var 4 PRIME UNEMP --silent
>>>> matrix A = $compan[1:2,]
>>>> # residuals + const
>>>> matrix U = $uhat .+ $coeff[1,]
>>>> # initial values
>>>> matrix y0 = {PRIME, UNEMP}[1:4,]
>>>> # simulate the VAR
>>>> matrix X = varsimul(A, U, y0)
>>>> # check that we got back the original data
>>>> # (to within machine precision)
>>>> eval X - {PRIME, UNEMP}
>>>> # now resample U
>>>> matrix Us = resample(U)
>>>> matrix Xs = varsimul(A, Us, y0)
>>>> # look at the simulated data
>>>> print Xs
>>>> </hansl>
>>>>
>>>> Use "set seed" if you want to be able to replicate the
>>>> resampling. Put the last few lines in a loop for multiple
>>>> simulations.
>>>>
>>>> Allin Cottrell
>>>>
>>>>
>>>> _______________________________________________
>>>> Gretl-users mailing list
>>>> Gretl-users(a)lists.wfu.edu
>>>> http://lists.wfu.edu/mailman/listinfo/gretl-users
>>>>
>>>
>>
>> _______________________________________________
>> Gretl-users mailing list
>> Gretl-users(a)lists.wfu.edu
>> http://lists.wfu.edu/mailman/listinfo/gretl-users
>>
>
>
> _______________________________________________
> Gretl-users mailing list
> Gretl-users(a)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-users
>
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
11:53 a.m.
Ok!! thank you so much!!!
Now I have a problem with the last line...
tcount += (Dsim > D)
error: unknown variable name in command..
2013/6/17 Artur Tarassow <artur.tarassow(a)googlemail.com>
It should be simply
set horizon 60
2013/6/17 Gabriela Nodari <gabriela.nodari(a)gmail.com>
> right!! thanks..
>
> Now the problem is with
>
> set horizon = 60
>
> error message: syntax error in command line...
>
>
> 2013/6/17 Artur Tarassow <artur.tarassow(a)googlemail.com>
>
>> You should delete the space in "var p y -- silent". So it should be
"var
>> p y --silent" instead.
>>
>> Artur
>>
>>
>> 2013/6/17 Gabriela Nodari <gabriela.nodari(a)gmail.com>
>>
>>> Dear all,
>>>
>>> I have run the varsimul() function and it works well!
>>>
>>> Now I am having some problems to estimate the var with simulated data..
>>>
>>> The code I am running is the following:
>>>
>>> scalar p = 6
>>> list y = 8 13 12 1 2 3
>>>
>>> var p y -- silent
>>> matrix Ac = $compan[1:6,]
>>> matrix U = $uhat.+$coeff[1,]
>>> matrix y0 = {var1, var2, var3, var4, var5, var6}[1:6,]
>>>
>>> list y1 = 8 13 12 1 2
>>> list x = 3
>>>
>>> var p y1; x
>>> matrix Ac1 = $compan[1:5,]
>>> matrix U1 = $uhat.+$coeff[1,]
>>> matrix y01 = {var1, var2, var3, var4, var5}[1:6,]
>>>
>>> loop (n. of replics)
>>> matrix Usim = resample(U)
>>> matrix Ysim = varsimul(Ac, Usim, y0)
>>> matrix Usim1 = resample(U1)
>>> matrix Ysim1 = vasimul(Ac1,Usim1,y01)
>>>
>>> series frs = Ysim[,1]
>>> series ips= Ysim[,2]
>>> ...
>>> series baas=Ysim[,6]
>>>
>>> var p frs ips ... baas
>>> ....
>>>
>>> Here I get error: invalid option '--' var p frs...
>>>
>>> I guess it is because I cannot enter these series in the var command..
>>>
>>>
>>> Could someone help me? How can I estimate the var with the simulated
>>> data?
>>>
>>> Thanks in advance..
>>> Gabriela
>>>
>>>
>>>
>>> 2013/6/14 Gabriela Nodari <gabriela.nodari(a)gmail.com>
>>>
>>>> Thank you very much! I will give a look to the varsimul function!!
>>>> On 14/06/2013 8:40 PM, "Allin Cottrell"
<cottrell(a)wfu.edu> wrote:
>>>>
>>>>> On Fri, 14 Jun 2013, Gabriela Nodari wrote:
>>>>>
>>>>> > I would like to obtain simulated data by bootstrapping the
>>>>> residuals of a
>>>>> > VAR model. I have gave a look to the gretl guide, and I found
>>>>> something
>>>>> > related to resampling and bootstrapping (section 7.4) [...]
>>>>>
>>>>> Check out the varsimul() function. Here's a simple example of
>>>>> use:
>>>>>
>>>>> <hansl>
>>>>> open data9-7
>>>>> var 4 PRIME UNEMP --silent
>>>>> matrix A = $compan[1:2,]
>>>>> # residuals + const
>>>>> matrix U = $uhat .+ $coeff[1,]
>>>>> # initial values
>>>>> matrix y0 = {PRIME, UNEMP}[1:4,]
>>>>> # simulate the VAR
>>>>> matrix X = varsimul(A, U, y0)
>>>>> # check that we got back the original data
>>>>> # (to within machine precision)
>>>>> eval X - {PRIME, UNEMP}
>>>>> # now resample U
>>>>> matrix Us = resample(U)
>>>>> matrix Xs = varsimul(A, Us, y0)
>>>>> # look at the simulated data
>>>>> print Xs
>>>>> </hansl>
>>>>>
>>>>> Use "set seed" if you want to be able to replicate the
>>>>> resampling. Put the last few lines in a loop for multiple
>>>>> simulations.
>>>>>
>>>>> Allin Cottrell
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Gretl-users mailing list
>>>>> Gretl-users(a)lists.wfu.edu
>>>>> http://lists.wfu.edu/mailman/listinfo/gretl-users
>>>>>
>>>>
>>>
>>> _______________________________________________
>>> Gretl-users mailing list
>>> Gretl-users(a)lists.wfu.edu
>>> http://lists.wfu.edu/mailman/listinfo/gretl-users
>>>
>>
>>
>> _______________________________________________
>> Gretl-users mailing list
>> Gretl-users(a)lists.wfu.edu
>> http://lists.wfu.edu/mailman/listinfo/gretl-users
>>
>
>
> _______________________________________________
> Gretl-users mailing list
> Gretl-users(a)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-users
>
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
11:58 a.m.
ok, i guess it was because I haven't define the variable at the beginning..
now should be ok...
Thanks again.
Gabriela
2013/6/17 Gabriela Nodari <gabriela.nodari(a)gmail.com>
Ok!! thank you so much!!!
Now I have a problem with the last line...
tcount += (Dsim > D)
error: unknown variable name in command..
2013/6/17 Artur Tarassow <artur.tarassow(a)googlemail.com>
> It should be simply
> set horizon 60
>
>
> 2013/6/17 Gabriela Nodari <gabriela.nodari(a)gmail.com>
>
>> right!! thanks..
>>
>> Now the problem is with
>>
>> set horizon = 60
>>
>> error message: syntax error in command line...
>>
>>
>> 2013/6/17 Artur Tarassow <artur.tarassow(a)googlemail.com>
>>
>>> You should delete the space in "var p y -- silent". So it should
be
>>> "var p y --silent" instead.
>>>
>>> Artur
>>>
>>>
>>> 2013/6/17 Gabriela Nodari <gabriela.nodari(a)gmail.com>
>>>
>>>> Dear all,
>>>>
>>>> I have run the varsimul() function and it works well!
>>>>
>>>> Now I am having some problems to estimate the var with simulated data..
>>>>
>>>> The code I am running is the following:
>>>>
>>>> scalar p = 6
>>>> list y = 8 13 12 1 2 3
>>>>
>>>> var p y -- silent
>>>> matrix Ac = $compan[1:6,]
>>>> matrix U = $uhat.+$coeff[1,]
>>>> matrix y0 = {var1, var2, var3, var4, var5, var6}[1:6,]
>>>>
>>>> list y1 = 8 13 12 1 2
>>>> list x = 3
>>>>
>>>> var p y1; x
>>>> matrix Ac1 = $compan[1:5,]
>>>> matrix U1 = $uhat.+$coeff[1,]
>>>> matrix y01 = {var1, var2, var3, var4, var5}[1:6,]
>>>>
>>>> loop (n. of replics)
>>>> matrix Usim = resample(U)
>>>> matrix Ysim = varsimul(Ac, Usim, y0)
>>>> matrix Usim1 = resample(U1)
>>>> matrix Ysim1 = vasimul(Ac1,Usim1,y01)
>>>>
>>>> series frs = Ysim[,1]
>>>> series ips= Ysim[,2]
>>>> ...
>>>> series baas=Ysim[,6]
>>>>
>>>> var p frs ips ... baas
>>>> ....
>>>>
>>>> Here I get error: invalid option '--' var p frs...
>>>>
>>>> I guess it is because I cannot enter these series in the var command..
>>>>
>>>>
>>>> Could someone help me? How can I estimate the var with the simulated
>>>> data?
>>>>
>>>> Thanks in advance..
>>>> Gabriela
>>>>
>>>>
>>>>
>>>> 2013/6/14 Gabriela Nodari <gabriela.nodari(a)gmail.com>
>>>>
>>>>> Thank you very much! I will give a look to the varsimul function!!
>>>>> On 14/06/2013 8:40 PM, "Allin Cottrell"
<cottrell(a)wfu.edu> wrote:
>>>>>
>>>>>> On Fri, 14 Jun 2013, Gabriela Nodari wrote:
>>>>>>
>>>>>> > I would like to obtain simulated data by bootstrapping the
>>>>>> residuals of a
>>>>>> > VAR model. I have gave a look to the gretl guide, and I
found
>>>>>> something
>>>>>> > related to resampling and bootstrapping (section 7.4) [...]
>>>>>>
>>>>>> Check out the varsimul() function. Here's a simple example
of
>>>>>> use:
>>>>>>
>>>>>> <hansl>
>>>>>> open data9-7
>>>>>> var 4 PRIME UNEMP --silent
>>>>>> matrix A = $compan[1:2,]
>>>>>> # residuals + const
>>>>>> matrix U = $uhat .+ $coeff[1,]
>>>>>> # initial values
>>>>>> matrix y0 = {PRIME, UNEMP}[1:4,]
>>>>>> # simulate the VAR
>>>>>> matrix X = varsimul(A, U, y0)
>>>>>> # check that we got back the original data
>>>>>> # (to within machine precision)
>>>>>> eval X - {PRIME, UNEMP}
>>>>>> # now resample U
>>>>>> matrix Us = resample(U)
>>>>>> matrix Xs = varsimul(A, Us, y0)
>>>>>> # look at the simulated data
>>>>>> print Xs
>>>>>> </hansl>
>>>>>>
>>>>>> Use "set seed" if you want to be able to replicate the
>>>>>> resampling. Put the last few lines in a loop for multiple
>>>>>> simulations.
>>>>>>
>>>>>> Allin Cottrell
>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> Gretl-users mailing list
>>>>>> Gretl-users(a)lists.wfu.edu
>>>>>> http://lists.wfu.edu/mailman/listinfo/gretl-users
>>>>>>
>>>>>
>>>>
>>>> _______________________________________________
>>>> Gretl-users mailing list
>>>> Gretl-users(a)lists.wfu.edu
>>>> http://lists.wfu.edu/mailman/listinfo/gretl-users
>>>>
>>>
>>>
>>> _______________________________________________
>>> Gretl-users mailing list
>>> Gretl-users(a)lists.wfu.edu
>>> http://lists.wfu.edu/mailman/listinfo/gretl-users
>>>
>>
>>
>> _______________________________________________
>> Gretl-users mailing list
>> Gretl-users(a)lists.wfu.edu
>> http://lists.wfu.edu/mailman/listinfo/gretl-users
>>
>
>
> _______________________________________________
> Gretl-users mailing list
> Gretl-users(a)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-users
>
12:01 p.m.
Nice that it works now ;-)
Artur
2013/6/17 Gabriela Nodari <gabriela.nodari(a)gmail.com>
ok, i guess it was because I haven't define the variable at the
beginning.. now should be ok...
Thanks again.
Gabriela
2013/6/17 Gabriela Nodari <gabriela.nodari(a)gmail.com>
> Ok!! thank you so much!!!
>
> Now I have a problem with the last line...
>
> tcount += (Dsim > D)
>
> error: unknown variable name in command..
>
>
> 2013/6/17 Artur Tarassow <artur.tarassow(a)googlemail.com>
>
>> It should be simply
>> set horizon 60
>>
>>
>> 2013/6/17 Gabriela Nodari <gabriela.nodari(a)gmail.com>
>>
>>> right!! thanks..
>>>
>>> Now the problem is with
>>>
>>> set horizon = 60
>>>
>>> error message: syntax error in command line...
>>>
>>>
>>> 2013/6/17 Artur Tarassow <artur.tarassow(a)googlemail.com>
>>>
>>>> You should delete the space in "var p y -- silent". So it
should be
>>>> "var p y --silent" instead.
>>>>
>>>> Artur
>>>>
>>>>
>>>> 2013/6/17 Gabriela Nodari <gabriela.nodari(a)gmail.com>
>>>>
>>>>> Dear all,
>>>>>
>>>>> I have run the varsimul() function and it works well!
>>>>>
>>>>> Now I am having some problems to estimate the var with simulated
>>>>> data..
>>>>>
>>>>> The code I am running is the following:
>>>>>
>>>>> scalar p = 6
>>>>> list y = 8 13 12 1 2 3
>>>>>
>>>>> var p y -- silent
>>>>> matrix Ac = $compan[1:6,]
>>>>> matrix U = $uhat.+$coeff[1,]
>>>>> matrix y0 = {var1, var2, var3, var4, var5, var6}[1:6,]
>>>>>
>>>>> list y1 = 8 13 12 1 2
>>>>> list x = 3
>>>>>
>>>>> var p y1; x
>>>>> matrix Ac1 = $compan[1:5,]
>>>>> matrix U1 = $uhat.+$coeff[1,]
>>>>> matrix y01 = {var1, var2, var3, var4, var5}[1:6,]
>>>>>
>>>>> loop (n. of replics)
>>>>> matrix Usim = resample(U)
>>>>> matrix Ysim = varsimul(Ac, Usim, y0)
>>>>> matrix Usim1 = resample(U1)
>>>>> matrix Ysim1 = vasimul(Ac1,Usim1,y01)
>>>>>
>>>>> series frs = Ysim[,1]
>>>>> series ips= Ysim[,2]
>>>>> ...
>>>>> series baas=Ysim[,6]
>>>>>
>>>>> var p frs ips ... baas
>>>>> ....
>>>>>
>>>>> Here I get error: invalid option '--' var p frs...
>>>>>
>>>>> I guess it is because I cannot enter these series in the var
>>>>> command..
>>>>>
>>>>>
>>>>> Could someone help me? How can I estimate the var with the simulated
>>>>> data?
>>>>>
>>>>> Thanks in advance..
>>>>> Gabriela
>>>>>
>>>>>
>>>>>
>>>>> 2013/6/14 Gabriela Nodari <gabriela.nodari(a)gmail.com>
>>>>>
>>>>>> Thank you very much! I will give a look to the varsimul
function!!
>>>>>> On 14/06/2013 8:40 PM, "Allin Cottrell"
<cottrell(a)wfu.edu> wrote:
>>>>>>
>>>>>>> On Fri, 14 Jun 2013, Gabriela Nodari wrote:
>>>>>>>
>>>>>>> > I would like to obtain simulated data by bootstrapping
the
>>>>>>> residuals of a
>>>>>>> > VAR model. I have gave a look to the gretl guide, and I
found
>>>>>>> something
>>>>>>> > related to resampling and bootstrapping (section 7.4)
[...]
>>>>>>>
>>>>>>> Check out the varsimul() function. Here's a simple
example of
>>>>>>> use:
>>>>>>>
>>>>>>> <hansl>
>>>>>>> open data9-7
>>>>>>> var 4 PRIME UNEMP --silent
>>>>>>> matrix A = $compan[1:2,]
>>>>>>> # residuals + const
>>>>>>> matrix U = $uhat .+ $coeff[1,]
>>>>>>> # initial values
>>>>>>> matrix y0 = {PRIME, UNEMP}[1:4,]
>>>>>>> # simulate the VAR
>>>>>>> matrix X = varsimul(A, U, y0)
>>>>>>> # check that we got back the original data
>>>>>>> # (to within machine precision)
>>>>>>> eval X - {PRIME, UNEMP}
>>>>>>> # now resample U
>>>>>>> matrix Us = resample(U)
>>>>>>> matrix Xs = varsimul(A, Us, y0)
>>>>>>> # look at the simulated data
>>>>>>> print Xs
>>>>>>> </hansl>
>>>>>>>
>>>>>>> Use "set seed" if you want to be able to replicate
the
>>>>>>> resampling. Put the last few lines in a loop for multiple
>>>>>>> simulations.
>>>>>>>
>>>>>>> Allin Cottrell
>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> Gretl-users mailing list
>>>>>>> Gretl-users(a)lists.wfu.edu
>>>>>>> http://lists.wfu.edu/mailman/listinfo/gretl-users
>>>>>>>
>>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Gretl-users mailing list
>>>>> Gretl-users(a)lists.wfu.edu
>>>>> http://lists.wfu.edu/mailman/listinfo/gretl-users
>>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> Gretl-users mailing list
>>>> Gretl-users(a)lists.wfu.edu
>>>> http://lists.wfu.edu/mailman/listinfo/gretl-users
>>>>
>>>
>>>
>>> _______________________________________________
>>> Gretl-users mailing list
>>> Gretl-users(a)lists.wfu.edu
>>> http://lists.wfu.edu/mailman/listinfo/gretl-users
>>>
>>
>>
>> _______________________________________________
>> Gretl-users mailing list
>> Gretl-users(a)lists.wfu.edu
>> http://lists.wfu.edu/mailman/listinfo/gretl-users
>>
>
>
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
12:30 p.m.
Yes, the code works! but I'm not sure the result is correct..
In practice, what I am trying to do is to compare two VAR models. I have
estimated one model with 6 variables, and the other one with 5 variables +
an exogenous. (the 6 variables are the same).
I have then computed then IRFs, and calculated the difference btw IRFs of
the two models
Next, I wanted to compute simulated IRFs' differences, so I have simulated
data and re-estimated the 2 models. I have done 2000 replications.
For each replication, I then compared the maximum difference I get from the
two simulated IRFs, with the difference between the 2 original IRFs.
But over 2000 replications, only once I got a simulated difference greater
than the benchmark (original) difference.
Sorry, I have explained in a bad way, hope you understand.. here is the
complete code:
D=max(abs(IP))
tcount=0
y = 8 13 12 1 2 3
y1 = 8 13 12 1 2
x = 3
#MODEL 1
var 6 y --silent
matrix Ac = $compan[1:6,]
matrix U = $uhat. +$coeff[1,]
#Initial values
matrix y0 = {var1, var2, var3, var4, var5, var6}[1:6,]
#MODEL 2
var 6 y1; x --silent
matrix Ac1 = $compan[1:5,]
matrix U1 = $uhat.+coeff[1,]
#Initial values
matrix y01={var1, var2, var3, var4, var5}[1:6,]
loop 2000
# Simulation data model 1
matrix Usim = resample (U)
matrix Ysim = varsimul(Ac, Usim, y0)
series frs=Ysim[,1]
series ips=Ysim[,2]
...
series bs=Ysim[,6]
# model 1 with simulated data
var 6 frs ips ... bs --silent
set horizon 60
matrix K = cholesky($sigma)
matrix V = $vma
#matrix of IRFs from simulated data
matrix IRFsim = V * (K**(I(6))
# IRF of variable of interest
matrix IRFip_b = IRFsim[,2]
# Simulation data model 2
matrix Usim1 = resample(U1)
matrix Ysim1 = varsimul(Ac1, Usim1, y01)
series frs1 = Ysim1[,1]
series ips1 = Ysim1[,2]
...
series fs1 = Ysim1[,5]
var 6 frs1 ips1 ... fs1; x --silent
set horizon 60
matrix K1 = cholesky($sigma)
matrix V1 = $vma
matrix IRFsim1 = V1 * (K1**I(5))
matrix IRFip_b1=IRFsim1[,2]
#Difference between IRFs from simulated models 1 and 2
matrix Diff = IRFip_b1 - IRFip_b
Dsim = max(abs(Diff))
if Dsim > D
tcount = tcount + 1
endif
endloop
print count
I would be very grateful if someone could provide any suggestion about it..
Thank you again,
Gabriela
1:32 p.m.
On Mon, Jun 17, 2013 at 7:30 AM, Gabriela Nodari
<gabriela.nodari(a)gmail.com> wrote:
Yes, the code works! but I'm not sure the result is correct.
[...]
here is the complete code:
D=max(abs(IP))
tcount=0
y = 8 13 12 1 2 3
y1 = 8 13 12 1 2
x = 3
#MODEL 1
var 6 y --silent
matrix Ac = $compan[1:6,]
matrix U = $uhat. +$coeff[1,]
#Initial values
matrix y0 = {var1, var2, var3, var4, var5, var6}[1:6,]
#MODEL 2
var 6 y1; x --silent
matrix Ac1 = $compan[1:5,]
matrix U1 = $uhat.+coeff[1,]
You need to include the effect of var6 in your MODEL 2 simulation.
After $coeff[1,] (the intercept) you should add
$coeff[7,] .* {x}[7:]
--
Allin Cottrell
Department of Economics
Wake Forest University
2:28 p.m.
Dear Allin,
Thanks for your kind help. The results do not change too much.
Meanwhile, I have noticed that when estimating the var via menu, when I
specify the exogenous variables, it considers only the contemporaneous
value of the exogenous observable. And it is not possible to specify lags.
I have checked for the Svar function and I guess it takes only the
contemporaneous value as well.
Is it possible to estimate a var with lagged exogenous variables?
Thanks again for your time and attention.
Gabriela
On 18/06/2013 3:34 PM, "Cottrell, Allin" <cottrell(a)wfu.edu> wrote:
On Mon, Jun 17, 2013 at 7:30 AM, Gabriela Nodari
<gabriela.nodari(a)gmail.com> wrote:
> Yes, the code works! but I'm not sure the result is correct. [...]
> here is the complete code:
>
> D=max(abs(IP))
> tcount=0
> y = 8 13 12 1 2 3
> y1 = 8 13 12 1 2
> x = 3
>
> #MODEL 1
> var 6 y --silent
> matrix Ac = $compan[1:6,]
> matrix U = $uhat. +$coeff[1,]
> #Initial values
> matrix y0 = {var1, var2, var3, var4, var5, var6}[1:6,]
>
> #MODEL 2
> var 6 y1; x --silent
> matrix Ac1 = $compan[1:5,]
> matrix U1 = $uhat.+coeff[1,]
You need to include the effect of var6 in your MODEL 2 simulation.
After $coeff[1,] (the intercept) you should add
$coeff[7,] .* {x}[7:]
--
Allin Cottrell
Department of Economics
Wake Forest University
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
3:01 a.m.
On Tue, Jun 18, 2013 at 10:28 AM, Gabriela Nodari
<gabriela.nodari(a)gmail.com> wrote:
Dear Allin,
Thanks for your kind help. The results do not change too much.
Meanwhile, I have noticed that when estimating the var via menu, when I
specify the exogenous variables, it considers only the contemporaneous
value of the exogenous observable. And it is not possible to specify lags.
You need to look at the dialog box more carefully. Consider what the "Lags"
button does.
Allin Cottrell
2:18 p.m.
Dear all/Allin,
I am trying to follow Allin's suggestion while resampling data and
estimating the var with lagged exogenous variables. However, I am getting
an error "ambiguos matrix index".
I am not able to include the effect of the exogenous varibles.. the code I
am running now is.
D=max(abs(IP))
tcount=0
y = 8 13 12 1 2 3
y1 = 8 13 12 1 2
x = 21 22 23 24 25 26 *#exogenous lagged variables*
#MODEL 1
var 6 y --silent
matrix Ac = $compan[1:6,]
matrix U = $uhat. +$coeff[1,]
#Initial values
matrix y0 = {var1, var2, var3, var4, var5, var6}[1:6,]
#MODEL 2
var 6 y1; x --silent
matrix Ac1 = $compan[1:5,]
matrix U1 = $uhat.+coeff[1,] + $coeff[7,] .* {x}[7:] + $coeff[8,] .*
{x}[8:] +
$coeff[9,] .* {x}[9:] + $coeff[10,] .* {x}[10:] + $coeff11,] .*
{x}[11:] + $coeff[12,] .* {x}[12:]
...
The error is appearing with respect to the last command above...
I would be very grateful for any suggestion!
Thanks in advance,
Gabriela
2013/6/18 Cottrell, Allin <cottrell(a)wfu.edu>
On Mon, Jun 17, 2013 at 7:30 AM, Gabriela Nodari
<gabriela.nodari(a)gmail.com> wrote:
> Yes, the code works! but I'm not sure the result is correct. [...]
> here is the complete code:
>
> D=max(abs(IP))
> tcount=0
> y = 8 13 12 1 2 3
> y1 = 8 13 12 1 2
> x = 3
>
> #MODEL 1
> var 6 y --silent
> matrix Ac = $compan[1:6,]
> matrix U = $uhat. +$coeff[1,]
> #Initial values
> matrix y0 = {var1, var2, var3, var4, var5, var6}[1:6,]
>
> #MODEL 2
> var 6 y1; x --silent
> matrix Ac1 = $compan[1:5,]
> matrix U1 = $uhat.+coeff[1,]
You need to include the effect of var6 in your MODEL 2 simulation.
After $coeff[1,] (the intercept) you should add
$coeff[7,] .* {x}[7:]
--
Allin Cottrell
Department of Economics
Wake Forest University
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
12:06 p.m.
On 17-06-2013, at 13:45, Gabriela Nodari <gabriela.nodari(a)gmail.com> wrote:
right!! thanks..
Now the problem is with
set horizon = 60
error message: syntax error in command line...
The Gretl Command Reference A4 English version for 1.9.12 (March 2013) on page 57 gives
an explicit example for set horizon.
Berend
4203
days inactive
4210
days old
15 comments
5 participants
participants (5)
-
Allin Cottrell -
Artur Tarassow -
Berend Hasselman -
Cottrell, Allin
-
Gabriela Nodari