bandwidth default for nadarwat()
by Sven Schreiber
Hi,
the doc for nadarwat says: "a popular choice is n^-0.2". Why not make
that the default value officially, such that the third/final arg becomes
optional?
thanks,
sven
6 years, 10 months
text search result off-window (bottom)
by Sven Schreiber
Hi,
this is not a new bug / odd behavior, but perhaps it's not on the radar
anymore, because it might be a Windows-only thing. It still happens
"often" (not always), that when you search for a word in the script
editor, the window content is moved such that the next match is just in
the next line below the bottom of the window, so not visible without
scrolling further.
This is with recent snapshots and releases, and I think both on Windows
7 and 10.
Maybe tricky to fix, and of course not release-critical.
thanks,
sven
7 years, 1 month
corr --triangle option not listed
by Artur Tarassow
Dear all,
the --triangle option is missing among the set of possible options in
the command reference, even though it's mentioned in the description.
Best,
Artur
7 years, 2 months
an issue with Octave export/import
by oleg_komashko＠ukr.net
Dear all,
Below is the script to illustrate the problem:
<hansl>
#### run in 2 chunks
open bjg.gdt
store "bjg.m" --gnu-octave
ma = mread("bjg.m")
eval ma[1:3,]
## why only g?
## this crushes gretl on my Ubuntu 17.10, 64
open bjg.m
<hansl>
Open GUI has '.m' option
Trying open bjg.m via GUI also crushes
Oleh
7 years, 2 months
'for' not reserved
by Sven Schreiber
Hi,
another question on whether it's intended (or wise...) or not:
<hansl>
clear
for = 1 # works
print for # prints out 1
function void for(void) # gretl doesn't complain
print "ouf"
end function
for() # error "unexpected ("
</hansl>
First, I'm wondering whether 'for' shouldn't simply be made a reserved
word. One fallout from the present situation is that this:
'loop for(i=0;......'
--i.e. without a blank after 'for'-- doesn't work. No big thing, of course.
Secondly, a totally different thing -- and this applies to every
identifier, not just 'for' as above: If 'for' is already a variable, you
can define a function with the same name, but then you cannot call it.
Wouldn't it be better if gretl already complained when it encounters the
function definition?
thanks,
sven
7 years, 2 months
Problems downloading help-docs (pdf)
by Artur Tarassow
Hi,
I just tried to download the following document through gretl:
Help-->"Command reference" (pdf)
but the downloaded document is only a few Bytes big. The same applies to
the "Keyboard shortcuts" document.
Best,
Artur
7 years, 2 months
Re: [Gretl-devel] [Gretl-users] When Julia is faster than C and Gretl
by Sven Schreiber
[taking this to devel instead of users]
Am 13.02.2018 um 18:23 schrieb Allin Cottrell:
> On Tue, 13 Feb 2018, Sven Schreiber wrote:
>> Am 13.02.2018 um 03:43 schrieb Allin Cottrell:
>>> But given how fast Julia is at generating random floating-point
>>> values, it seems to me there should be a real live example not far off.
>>
>> Yes. Again, I suggest to tackle the SB.gfn package as a benchmark.
>
> Would make an interesting test case.
OK, since I still don't have much experience with Julia I have played
around with Python/Numba on the stationary bootstrap (SB) first. I'm
attaching a horserace between the default SB function from SB.gfn and a
JITted Python/Numpy/Numba implementation of the loop-based version SB_old2.
It turns out Numba is 5x faster for the pure calculations (well,
depending on the problem size due to the fixed overhead), but the
bottleneck is the disk-based transfer of the big random matrix back to
the gretl environment. Here are my results for 39999 replications:
SB 1.3, 2017-11-14 (F. Di Iorio, S. Fachin, A. Tarassow, Riccardo "Jack"
Lucchetti)
Gretl native (SB package)
This took = 6.067643 sec.
Python / Numba pure calculation
This took = 1.323467 sec.
Python / Numba with transfer
This took = 7.564688 sec.
I believe something similar would emerge for Julia. So unless the data
transfer can be organized in some other way I think it's only possible
to beat well-written hansl code when the data reduction step can already
happen on the Python (or Julia) side.
cheers,
sven
7 years, 2 months
string with integer offset: off-by-one bug or unclear doc?
by Sven Schreiber
Hi,
not sure if this is intended (probably), or just an oversight from
0-based to 1-based indexing:
string ss = "abc"
eval ss + 1 # gives "bc"
The doc (ch. 14) says: " ...starting at the given character offset."
Normally I would interpret "starting at" meaning "right there", so in
the above example starting from the 1st character ("a"). OTOH I can see
that an "offset" could mean "right after that".
I tend to think this isn't a bug, but perhaps the doc could be
reformulated as "starting at the character after the given offset".
thanks,
sven
7 years, 2 months
Can't suppress BFGS message from inside a function
by oleg_komashko＠ukr.net
Dear all,
The scrip below illustrates the problem
Hopefully, the printing of the script is transparent
The problem is in that 'set verbose off' inside
a function does not suppress BFGS message
<hansl>
open greene5_1.gdt -q
set verbose off
catch arima 2 1 0; 0 1 1; pop --nc -q
print "caught: arima 2 1 0; 0 1 1; pop --nc -q"
err = $error
msg = errmsg(err)
printf "This is error message: %s\n\n", msg
printf "f() simply runs 'catch arima 2 1 0; 0 1 1; y --nc -q'\n"
printf "inside a loop, preceded by 'set verbose off'\n"
printf "failed to suppress: f(pop, 3)\n\n"
function void f(series y, int n[1])
loop i=1..n -q
set verbose off
catch arima 2 1 0; 0 1 1; y --nc -q
set verbose on
endloop
end function
f(pop,3)
<hansl>
Oleh
P.S.
this model gives excellent Q-statistics
After the estimation, it is clear, AR(2) term is redundant
But it is not clear before estimation
The default init values are not outside the unit circle
Since 2018a is coming it is difficult to find time to
translate my method of transforming init values
into c and try
May be it would worth considering to send me
hansl code of the Rissanen init values for me
to experiment with
7 years, 2 months
git issue
by Artur T.
Dear all,
I just wanted to compile a fresh version of gretl but I've obtained the
following error after "git pull"
<error>
at@ati7:~/git/gretl-git$ git pull
Aktualisiere 53f234b8a..74367f589
error: Ihre lokalen Änderungen in den folgenden Dateien würden durch den
Merge überschrieben werden:
lib/src/geneval.c
lib/src/gretl_foreign.c
Bitte committen oder stashen Sie Ihre Änderungen, bevor sie mergen.
Abbruch
</error>
It's saying that the 2 listed files would be overwritten by the merge
command, and I should commit or stash my changes (I didn't do any!)
before merging.
Any suggestions what to do?
Thanks,
Artur
7 years, 2 months
