On Wed, 2011-10-26 at 21:28 +0200, Riccardo (Jack) Lucchetti wrote:
> On Wed, 26 Oct 2011, Giuseppe Vittucci wrote:
>
> > The following code (adapted from the manual):
> >
> > mle logl = check ? - ln(pstr_cssr(y,X,q,gamma,c,m,Z) : NA
> > scalar check = (gamma > zeros(r,1)) && (c >= c_min) && (c
On Wed, 26 Oct 2011, Giuseppe Vittucci wrote:
> The following code (adapted from the manual):
>
> mle logl = check ? - ln(pstr_cssr(y,X,q,gamma,c,m,Z) : NA
> scalar check = (gamma > zeros(r,1)) && (c >= c_min) && (c <=c_max)
> params gamma c
> end mle
>
> simply checks that the unconstrained
The following code (adapted from the manual):
mle logl = check ? - ln(pstr_cssr(y,X,q,gamma,c,m,Z) : NA
scalar check = (gamma > zeros(r,1)) && (c >= c_min) && (c <=c_max)
params gamma c
end mle
simply checks that the unconstrained maximum is in the parameter space
and returns an error if it is
On Wed, 26 Oct 2011, Simari wrote:
> May I also ask you which log of data Gretl makes?I mean Log 10,Natural log
> etc...
Ok, I've always wanted to keep this to myself, but... I'll share a secret
with you. Gretl has a hidden backdoor, which gives you acces to a whole
lot of classified
May I also ask you which log of data Gretl makes?
I mean Log 10,Natural log etc...
And if there's an important difference in using one instead of another.
Thanks
S.
On Wed, Oct 26, 2011 at 5:43 PM, Allin Cottrell wrote:
> On Wed, 26 Oct 2011, Simari wrote:
>
> > Ok,
> > can you please
Dear Gretl Developers,
A colleague here at my work is trying to perform some regressions
using .xlsx files but he is getting the following error messages:
"Invalid argument for function" in the GUI
"gretl_unzip_file: 'zip error: File not found or no read permission'"
in the script output window
Ok,
can you please tell me why?
I see the same result without the time trend as regressor.
Thanks
S.
On Wed, Oct 26, 2011 at 3:37 PM, Allin Cottrell wrote:
> On Wed, 26 Oct 2011, Simari wrote:
>
> > Importing data (set on 1,2,3 instead of 13/06/2011,
> > 14/06/2011 etc. on file Excel)
Yes Allin, I found the error thanks to Ignacio and asked a couple of things
in the previous email.
S.
On Wed, Oct 26, 2011 at 3:20 PM, Allin Cottrell wrote:
> On Wed, 26 Oct 2011, Simari wrote:
>
> > I have in excel:
> > 1. a column with date (13/06/10, 13/07/11 etc) daily basis
> > 2.
Ignacio,
thanks, I did it.
Importing data (set on 1,2,3 instead of 13/06/2011, 14/06/2011 etc. on file
Excel)
importing them as cross section instead of series and then not adding the
time trend
I have the same result.
Do you think was I right?
Simari
On Wed, Oct 26, 2011 at 11:45 AM,
Am 10/26/2011 11:17 AM, schrieb Ignacio Diaz-Emparanza:
> El 26/10/11 09:10, Sven Schreiber escribió:
...
>> would need to be available on the user's system for that). But my guess
>> is that Allin in one of his coding flashes solves the problem before that...
>>
>> cheers,
>> sven
>
> Sorry,
On Wed, 26 Oct 2011, Anutechia Asongu wrote:
> Hi Allin, please could you expatiate or paraphrase the term
> : "just-identified" expressed in your last comment?
The number of added instruments equals the number of
endogenous variables.
Allin
El 25/10/11 17:59, Allin Cottrell escribió:
> On Mon, 24 Oct 2011, Simari wrote:
>
>> I am trying to make a linear regression between X (days, 13/06/2011,
>> 14/06/2011, 15/06/2011 and so on) and Y (price data)
>> and look for the slope and R2.
>
> Unless you have special requirements not
On Wed, 26 Oct 2011, Simari wrote:
> Ok,
> can you please tell me why?
>
> I see the same result without the time trend as regressor.
Sorry, I'm now lost. If you're getting the results you were
hoping for, that's fine.
Allin Cottrell
> On Wed, Oct 26, 2011 at 3:37 PM, Allin Cottrell wrote:
>
On Wed, 26 Oct 2011, Annaert Jan wrote:
> I came across strange behaviour of the summary command used
> with the --by option. When I use it on a sample less than
> the full range and the byvar is a monthly dummy variable
> (generated using the GUI), I do get the summary statistics
> for each
El 26/10/11 09:10, Sven Schreiber escribió:
> Two remarks, one to point out a workaround, one for a potential future
> solution:
>
> The workaround: One could use LibreOffice (or OpenOffice) to read the
> xlsx file and then save as a format which gretl understands.
>
> A potential solution:
But robustness in terms of specification means to get *similar* results,
not identical.
You only want to get *identical* results if you want to check whether
the software has implemented it correctly.
-s
Am 10/26/2011 10:40 AM, schrieb Anutechia Asongu:
> Hi Sven, thanks for the time indeed. My
I have in excel:
1. a column with date (13/06/10, 13/07/11 etc) daily basis
2. A column with price
I make the log of data
I make the OLS from Model and I get this result:
^l_Serie_T = 1,62
(0,00107)
T = 88, R-quadro = 0,000
In Excel instead of:
R2 0,9306
Coeff. 0,0004
I don't understand -- if TSLS does what you want, why do it via GMM? If
it doesn't do what you want, why try to get identical results? Or is
this a way of testing/checking gretl?
-sven
Am 10/26/2011 10:35 AM, schrieb Anutechia Asongu:
> Yeah Sven, I hope to get numerically identical results.
So it does run it seems. What's your aim now? To get numerically
identical results, but why? Or are you worried that the results are
"too" different? (In which case presumably it's not a gretl problem, but
a matter of your application.)
cheers,
sven
Am 10/26/2011 10:13 AM, schrieb Anutechia
Well if it makes any sense in your context, maybe you could restrict the
sample "manually" (= --no-missing) and then apply GMM. Haven't tested
this though.
hth,
sven
Am 10/26/2011 10:00 AM, schrieb Anutechia Asongu:
> Hi All,
>Can't one-step GMM that is compatible with TSLS be
On Wed, 26 Oct 2011, Simari wrote:
> Importing data (set on 1,2,3 instead of 13/06/2011,
> 14/06/2011 etc. on file Excel) importing them as cross
> section instead of series and then not adding the time trend
> I have the same result.
You need to add the time trend and use it as a regressor.
On Wed, 26 Oct 2011, Sven Schreiber wrote:
> But robustness in terms of specification means to get *similar* results,
> not identical.
>
> You only want to get *identical* results if you want to check whether
> the software has implemented it correctly.
And in the context of tsls "versus"
On Wed, 26 Oct 2011, Sven Schreiber wrote:
> b is the coefficient -- if you have trouble finding it in the output, I
> predict some wonderful weeks ahead for you in which you will discover
> the beautiful world of econometrics.
:-D
> As for the different R2, you would need to post an example.
On Wed, 26 Oct 2011, Simari wrote:
> I have in excel:
> 1. a column with date (13/06/10, 13/07/11 etc) daily basis
> 2. A column with price
>
> I make the log of data
> I make the OLS from Model and I get this result:
>
> ^l_Serie_T = 1,62
>(0,00107)
>
> T = 88, R-quadro = 0,000
b is the coefficient -- if you have trouble finding it in the output, I
predict some wonderful weeks ahead for you in which you will discover
the beautiful world of econometrics.
As for the different R2, you would need to post an example. This stuff
is so standard that I'm willing to bet a large
Two remarks, one to point out a workaround, one for a potential future
solution:
The workaround: One could use LibreOffice (or OpenOffice) to read the
xlsx file and then save as a format which gretl understands.
A potential solution: Instead of having Allin code the xlsx import stuff
in C, one
Thanks for the answer Allin,
the problem that I have is that
making the OLS command I do not see the "b" of the Linear regression.
At same time I see different results, as R2 for the same regression made on
Excel.
I'd like to understand why this difference.
Regards,
Simari
On Tue, Oct 25,
Hi Allin, please could you expatiate or paraphrase the term : "just-identified"
expressed in your last comment?
From: Allin Cottrell
To: Gretl list
Sent: Wednesday, October 26, 2011 3:34 PM
Subject: Re: [Gretl-users] One Step GMM and TSLS
On Wed, 26 Oct
I came across strange behaviour of the summary command used with the --by
option. When I use it on a sample less than the full range and the byvar is a
monthly dummy variable (generated using the GUI), I do get the summary
statistics for each category of the dummy variable, but they are
Thanks Sven
From: Sven Schreiber
To: gretl-users(a)lists.wfu.edu
Sent: Wednesday, October 26, 2011 10:49 AM
Subject: Re: [Gretl-users] One Step GMM and TSLS
But robustness in terms of specification means to get *similar* results,
not identical.
You only want
Hi Sven, thanks for the time indeed. My purpose of using a one-step GMM is for
robustness test.
From: Sven Schreiber
To: gretl-users(a)lists.wfu.edu
Sent: Wednesday, October 26, 2011 10:37 AM
Subject: Re: [Gretl-users] One Step GMM and TSLS
I don't
Yeah Sven, I hope to get numerically identical results. Beyond this wish, I
also hope to get the Sargan OIR test results( which do not appear in my GMM
approach)
From: Sven Schreiber
To: gretl-users(a)lists.wfu.edu
Sent: Wednesday, October 26, 2011 10:29 AM
Thanks Sven, I'm tried the option but results are different.
From: Sven Schreiber
To: gretl-users(a)lists.wfu.edu
Sent: Wednesday, October 26, 2011 10:10 AM
Subject: Re: [Gretl-users] One Step GMM and TSLS
Well if it makes any sense in your context, maybe you
Hi All,
Can't one-step GMM that is compatible with TSLS be performed with
missing values?. Indeed I'm using TSLS and should like to use one-step GMM for
robustness test. Please is there a way one can turn-around this "missing values
encountered." spectre that keeps hunting me?
34 matches
Mail list logo