Re: [Gretl-users] any documentation on error message
Thank you, Allin.
I have a further question.
I have time series data for 50 variables. I want to calculated the mean for each variable
and store the data as a cross-section dataset that has two columns: one for 50 means and
one for their correpeponding varibles.
Could you tell me how to do that in gretl? Thank you!
Regards,
Rebecca
--- On Fri, 11/7/08, Allin Cottrell <cottrell(a)wfu.edu> wrote:
From: Allin Cottrell <cottrell(a)wfu.edu>
Subject: Re: [Gretl-users] any documentation on error message
To: zzh5460(a)yahoo.com, "Gretl list" <gretl-users(a)lists.wfu.edu>
Date: Friday, November 7, 2008, 9:11 AM
On Fri, 7 Nov 2008, Rebecca Zhang wrote:
> I got below error message when I tried to calculate
mean on a
> matrix (use meanc).
>
> "Data types not conformable for operation"
>
> Does anyone know what is reason?
Can you show the command in context? This function take a
matrix
as argument and returns a row vector, as in
matrix M = mnormal(4,2)
matrix v = meanc(M)
print M v
If I try, e.g.,
scalar s = meanc(M)
I get the error message you mentioned (unless M has just
one
column, in which case gretl will convert the result to a
scalar).
You can never assign the result of meanc() to a variable of
series type.
> Also is there any documentation I can find the
detailed help for
> such error messages?
You really just have to cross-reference these errors with
the
"function reference" to see why they arise.
Allin Cottrell