when testing the (just now approved) package ShapleyCont.gfn I noticed
that in the GUI package window I get the string "Dependent variable"
translated into my interface language German, namely "Abhängige
Variable". The package itself has nothing German in it of course. Other
strings such as "Independent variables" stay in English (the language of
Is this intended? I cannot really imagine it is.
I don't really understand why the last line doesn't work (example should
run with any dataset that has the "index" generic variable):
list checkl = index
series st = index^2
checkl += st
checkl += log(index)
checkl += index^2 # fails
The on-the-fly transformation doesn't seem to be a problem per se, or
else log(index) would also fail.
Could it be that the "^" operation is mistaken by gretl to be the
special interaction thingy? (Which should only kick in if the right-hand
argument is a series or list, not a number like the 2 here.)
This is with the June 29th snapshot on Windows.
it would be nice if the error which is triggered here:
setobs 1 1900 --time-series
could get a message like "argument / index out of range" instead of the
generic "data error".
this is an interface annoyance that I've experienced for a while,
possibly even years. (On Windows, don't know right now if it happens on
Linux as well.) In the built-in script editor if you're typing at the
bottom of the screen and then hit return, often it happens that the new
bottom line isn't shifted up so that it's "in full sight". Either it's
fully off-window or it might be only partly visible.
A second variation of this problem is that when I do a text search (or
search&replace), the highlighted text that was found is also in the line
that is just below the bottom visible window area, if you understand
what I mean.
The workaround is to manually scroll down by the equivalent of one line
height, which gets annoying after a while.
Anybody else seeing this?
not sure if this is a bug:
matrix a = zeros(3,1)
matrix b = ones(2,3)
matrix hu = a' * b'
hu = a'(b')
hu = a'b' # fails
Apparently gretl interprets the last line as "(a'b)'" which kind of
implies that the implicit multiplication with the transpose sign has
higher precedence than the actual transposition of b.