3:19 p.m.
Hi,
I'm working with GRETL 1.84 CVS.
I have noticed that when importing EViews file the sample range is set
incorrectly when first quarter in the year is the third.
I have attached an EViews 6 file ranging 1980Q3 - 1982Q3.
An import will set the Gretl range to 1980:1 - 1982:1 incorrectly.
With kind regards,
Ofer
7:28 p.m.
On Thu, 24 Sep 2009, Ofer Cornfeld wrote:
I have noticed that when importing EViews file the sample range is
set
incorrectly when first quarter in the year is the third.
I have attached an EViews 6 file ranging 1980Q3 - 1982Q3.
An import will set the Gretl range to 1980:1 - 1982:1 incorrectly.
Thanks. Unfortunately, getting date information out of Eviews
files is a guessing game, because they use an unpublished
proprietary binary format (which we have attempted to
reverse-engineer). I thought we had guessed how to find the
starting period correctly, but it seems not.
If you have access to Eviews, it might be helpful if you could
send me a few variant wf1 files: for example, a year's worth of
monthly data, in one case with the starting date set to 1990:01,
then with the start set to 1990:02, and so on. Hopefully this
would enable us to track the bytes where the starting period is
recorded.
Allin Cottrell
3:06 a.m.
Dear Allin
I created 3 files as attached.
All of these Eviews 6.x files contain only one variable named year
which has the same four samples. The only differences among these 3
files are the ranges, starting from 1990:1, 1990:2, and 1990:3 for
1990-1.wf1, 1990-2.wf1 and 1990-3.wf1, respectively.
Hopefully they are useful.
Yi-Nung Yang
If you have access to Eviews, it might be helpful if you could
send me a few variant wf1 files: for example, a year's worth of
monthly data, in one case with the starting date set to 1990:01,
then with the start set to 1990:02, and so on. Hopefully this
would enable us to track the bytes where the starting period is
recorded.
Allin Cottrell
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
3:16 a.m.
On Fri, 25 Sep 2009, yinung at Gmail wrote:
Dear Allin
I created 3 files as attached.
All of these Eviews 6.x files contain only one variable named year
which has the same four samples. The only differences among these 3
files are the ranges, starting from 1990:1, 1990:2, and 1990:3 for
1990-1.wf1, 1990-2.wf1 and 1990-3.wf1, respectively.
Hopefully they are useful.
Thanks very much; I'll see what I can discover from these files.
(I might mention, gretl has made the running in trying to
reverse-engineer Eviews data. Gretl has borrowed from R in some
respects but this is a case where R has borrowed from us.)
Allin.
9:40 a.m.
Dear Allin,
I have attached files with some examples of Eviews files, where the names
imply the date ranges.
I have taken a pick at the hex dump of the files and it seems to me that the
word (two bytes) located at offsets 84 and 85 Hex contain the starting
period within the year.
Hope this helps,
With kind regards,
Ofer Cornfeld
-----Original Message-----
From: gretl-users-bounces(a)lists.wfu.edu
[mailto:gretl-users-bounces@lists.wfu.edu] On Behalf Of Allin Cottrell
Sent: Thursday, September 24, 2009 10:28 PM
To: Gretl list
Subject: Re: [Gretl-users] Eviews Import - a bug report
On Thu, 24 Sep 2009, Ofer Cornfeld wrote:
I have noticed that when importing EViews file the sample range is
set
incorrectly when first quarter in the year is the third.
I have attached an EViews 6 file ranging 1980Q3 - 1982Q3.
An import will set the Gretl range to 1980:1 - 1982:1 incorrectly.
Thanks. Unfortunately, getting date information out of Eviews
files is a guessing game, because they use an unpublished
proprietary binary format (which we have attempted to
reverse-engineer). I thought we had guessed how to find the
starting period correctly, but it seems not.
If you have access to Eviews, it might be helpful if you could
send me a few variant wf1 files: for example, a year's worth of
monthly data, in one case with the starting date set to 1990:01,
then with the start set to 1990:02, and so on. Hopefully this
would enable us to track the bytes where the starting period is
recorded.
Allin Cottrell
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
2:37 p.m.
On Fri, 25 Sep 2009, Ofer Cornfeld wrote:
I have attached files with some examples of Eviews files, where
the names imply the date ranges.
I have taken a pick at the hex dump of the files and it seems to
me that the word (two bytes) located at offsets 84 and 85 Hex
contain the starting period within the year.
Very helpful. Yes, I think you're right. I've revised the Eviews
reader in CVS accordingly.
Allin.
5326
days inactive
5327
days old
5 comments
3 participants
participants (3)
-
Allin Cottrell -
Ofer Cornfeld -
yinung＠Gmail