[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
st: RE: xtfevd
Nick Cox <email@example.com>
st: RE: xtfevd
Thu, 16 Feb 2012 10:57:04 +0000
-xtfevd- is a user-written program. It is not even found by -findit-. Either way, you are asked to explain where user-written programs you refer to come from. I don't think the authors of -xtfevd- are members of Statalist, so I doubt that you will get good support on it through Statalist.
Incidentally, a Google for fixed effects vector decomposition will yield references to discussion papers maintaining that this method has been oversold, to put it fairly politely. I can't adjudicate on this -- I've not even read the papers -- but my impression is that the critics are weightier authorities than the proponents. (I are aware from argument from authority is often regarded as a fallacy, but it's what I have here.)
That aside, Mata can't find a way to allocate a matrix that is [6168, 6168]. Deeper diagnosis of that problem may depend on some combination of (1) the details of -xtfevd- (2) the details of your dataset and what this command implies (3) the memory available on your computer (4) the flavour of your Stata.
I am getting the following error message when I try to do an estimation by using xtfevd. I've never seen this message before, so I don't know what to do. Can anyone suggest something?
The error message I get is:
. xtfevd lrealexp lmig ldist lgdpcons_o lgdpcons_d comlang_off contig colony com
> lang_ethno gatt_d rta comcur lmigmpc lmigeec lmigasea, invariant( ldist colony co
> mlang_ethno comlang_off contig)
*: 3900 unable to allocate real <tmp>[6168,6168]
vc_eta(): - function returned error
<istmt>: - function returned error
* For searches and help try: