zombie missings when constructing new lists
by Sven Schreiber
Hi,
I have a mysterious error occurring which has to do with magically
appearing missing values in the course of constructing lists. I'm
putting together a series (no missings) and an existing list (no
missings), but the resulting list suddenly has 95 obs (out of 140) with
missings! How is that possible?
Here's the exact debug code to show the problem (ripped out of a bigger
script, so it's not self-contained, sorry):
<excerpt>
loop foreach i endolist
# debug
list debuglist = endolist.$i laglist
series anymiss_i = missing(endolist.$i)
series anymiss_l = missing(laglist)
series anymiss = missing(debuglist)
numofmiss_i = sum(anymiss_i)
numofmiss_l = sum(anymiss_l)
numofmiss = sum(anymiss)
printf "Num of missings here lhs: %d\n", numofmiss_i
printf "Num of missings here laglist: %d\n", numofmiss_i
printf "Num of missings here: %d\n", numofmiss
# end debug
</excerpt>
Output:
Num of missings here lhs: 0
Num of missings here laglist: 0
Num of missings here: 95
I should add that this happens inside a function, and while 'endolist'
is not a function argument itself, it contains series and lists that
were passed as arguments. I tried both the 'endolist.$i' variant and the
plain '$i', with the same result.
My feeling is that this is a bug, but as always, I may be missing (pun
not intended) something...
thanks,
sven
12 years, 7 months
Cointegration change backwards incompatible
by Sven Schreiber
Hi,
I just looked at the changelogs (regular and incompatible changes) for
1.9.7, and it seems to me that the adoption of the PcGive convention for
restricted exogenous terms should also appear in the log of backwards
incompatible changes, no?
cheers,
sven
12 years, 7 months
follow-up on system estimation in a script
by Sven Schreiber
Hi,
Allin wrote last month:
> Here's what's now in CVS (I'm not documenting it yet because it's still open to discussion):
>
> A. The simple case with common regressors:
>
> <hansl>
> open denmark.gdt
> list Y = LRM LRY IBO IDE
> var 3 Y
>
> list X = const lags(3, Y)
> system method=SUR
> equations Y X
> end system
> </hansl>
>
> B. With differing regressors:
>
> <hansl>
> # subset of the Grunfeld investment data (attached)
> open grunfeld.txt
>
> list Y = ige iwh
> list X1 = const fge kge
> list X2 = const fwh kwh
> # join lists with separator
> list X = X1 ; X2
>
> system method=SUR
> equations Y X
> end system
> </hansl>
>
> In form B the composite X-list must contain as many sub-lists
> (joined via ";") as there are series in the Y-list.
Now I have tried to test and use this, but I don't know how to construct
the semicolon-separated list in a loop (since in general I don't know
the number of equations in advance). Here's what I tried but that didn't
work (on current cvs):
<hansl>
open denmark
list lhs = IBO IDE LRM LRY
list allrhs = null
loop foreach i lhs
list rhs_$i = LRM(-1) # trivial example
list allrhs = allrhs ; rhs_$i # does this work?
endloop
system method=SUR
equations lhs allrhs
end system
</hansl>
I get the following error:
<error>
equations lhs allrhs
Command has insufficient arguments
Error executing script: halting
</error>
Any ideas?
BTW, I noticed that one year ago there was already a discussion about
how to construct (SUR) systems in a script, see
http://lists.wfu.edu/pipermail/gretl-devel/2010-December/003042.html
We somehow seem to have forgotten about it in the meantime. Would that
approach still work, I haven't tried it so far?
Thanks,
sven
12 years, 7 months
Problem Installing/updating function file on Mac OS X Lion/Snow Leopard
by Berend Hasselman
I'm testing the latest gretl snapshot on Max OS X Lion 10.7.2 and Max OS X Snow Leopard 10.6.8.
On Lion I had no function files installed.
1.
I try to install the gig package from the server without realising that gig is part of Gretl.app.
Gretl downloads the package (so I assume since I'm seeing a progress bar) but then gives the message "Unspecified error".
But now comes the weird thing: menu Help, Check addons says that gig 2.1 is installed.
So does File, Function files, On local machine
(But shows version 2.0 for Function files, On server)
This also the case on Snow Leopard.
Inspecting Gretl.app/Contents/Resources/share/gretl/functions/gig reveals that some files have been updated (dated Yesterday) but not all (they have a date (mm-dd-yyyy) 13-10-2011).
Before trying the install of gig from the server in Gretl.app (regardless of Lion or Snow Leopard) all these files are dated Yesterday.
Initially I thought the problem was Lion specific but it isn't.
Gretl is trying to install an older version of gig on top of a newer version.
2.
Trying to install a not installed function file gives the error message "Couldn't find <package name> for gretl 1.9.7"
(this is regardless of Snow Leopard or Lion).
Gretl doesn't seem to take the <minver> field into account.
3.
The pdf for gig mentions version 2.2! But gig.gfn has the version set to 2.1
Berend
12 years, 7 months
time for 1.9.7?
by Allin Cottrell
Looking at the ChangeLog, it seems to me the 1.9.7 release
should go out fairly soon: a good number of bugs have been
fixed -- some of them crashers -- and some useful new
functionality has been added. And at this point we haven't
made any serious systemic changes that could be harboring a
lot of new bugs. (Which is not to claim that there are zero
new bugs, of course.)
I've recently updated the translatable strings, so if anyone
has time to update their translations, please do so.
As with the last release, the more testing people are able to
do beforehand, the better. However, I think this is not quite
such a high priority for 1.9.7 since the focus lately has
mostly been on bug-fixing.
Allin
12 years, 7 months
about error while applying quantile regression on a dataset
by Kalam narendar reddy
hello sir,
i have used the ramanathan data4-1 file which is about prices of single
family home.
using Data>sample file>ramanathan>data4-1 which contain as the independent
vriables and price as the dependent variable.
so i have imported the data and then i selected the
model.robust estimation>quantile regression
and
added price to dependent variable panel
and
sqft,
bedrms,
baths
as the independent variables.
then i have specified 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
as the tau value
and then
i clicked on ok
but i encountered a error saying
*invalid quantile specification.*
*why it is.pls reply me as early as possible.*
*thanks in advance*
*with regards*
*Kalam narendar reddy*
,
12 years, 7 months
