9:36 p.m.
Dear Gretl Community,
In a "Principal Component Analysis" context, is there an accessor that can
be used for retrieve the component loading coefficients into a matrix?
Best regards,
Henrique Andrade*
*
11:18 p.m.
On Mon, 5 Mar 2012, Henrique Andrade wrote:
Dear Gretl Community,
In a "Principal Component Analysis" context, is there an accessor that can
be used for retrieve the component loading coefficients into a matrix?
No, but you can easily get such a matrix by using eigensym() on the
correlation (or covariance) matrix of the list of series (converted
to a matrix). Example:
<hansl>
open data7-12
list L = wbase length width height weight cyl liters
matrix C = mcorr({L})
matrix V
eigensym(C, &V)
print V # eigenvectors = loadings
# compare the PCA printout
pca L
</hansl>
Note that eigensym() orders things by ascending value, so the
loadings associated with the first principal component are in the
last column of V.
Allin Cottrell
9:59 a.m.
Am 06.03.2012 00:18, schrieb Allin Cottrell:
On Mon, 5 Mar 2012, Henrique Andrade wrote:
> Dear Gretl Community,
>
> In a "Principal Component Analysis" context, is there an accessor that can
> be used for retrieve the component loading coefficients into a matrix?
No, but you can easily get such a matrix by using eigensym() on the
correlation (or covariance) matrix of the list of series (converted
to a matrix). Example:
<hansl>
open data7-12
list L = wbase length width height weight cyl liters
matrix C = mcorr({L})
matrix V
eigensym(C,&V)
print V # eigenvectors = loadings
# compare the PCA printout
pca L
</hansl>
Note that eigensym() orders things by ascending value, so the
loadings associated with the first principal component are in the
last column of V.
I'm wondering whether there may be a case to extend the princomp()
function somewhat to get a direct access to that (optional third matrix
parameter in pointer form for example). Sure it would be sort of
redundant from a syntactic point of view, but since those eigenvectors
have to be computed anyway it could be an efficient way to get those.
cheers,
sven
10:08 a.m.
Am 06.03.2012 10:59, schrieb Sven Schreiber:
I'm wondering whether there may be a case to extend the princomp()
function somewhat to get a direct access to that (optional third matrix
parameter in pointer form for example). Sure it would be sort of
redundant from a syntactic point of view, but since those eigenvectors
have to be computed anyway it could be an efficient way to get those.
[p.s. to gretl-devel only:]
A general syntactic question about those pointer-style arguments: Given
that hansl is usually very concise, is it really always necessary to
define/initialize the object in a separate line?
What I mean is, instead of:
<hansl_snippet>
matrix argument_to_be
eigensym(foo,&argument_to_be)
</hansl_snippet>
would it be possible to do something like:
<hansl_snippet>
eigensym(foo,&[matrix myarg])
</hansl_snippet>
Note the added '[matrix ...]' -- I am not proposing to use the old
syntax and simply create the argument on the fly and guess its type,
which would not catch typos for example (if really previously defined
objects are meant).
This is really just about saving the extra line and hopefully make hansl
even more readable. I know tastes will differ here (especially those of
the C people...).
thanks,
sven
1:01 p.m.
Dear Allin,
Thanks for your help. It solved my problem :-)
By the way... I liked Sven's suggestion.
Best regards,
Henrique
2012/3/6 Sven Schreiber <svetosch(a)gmx.net>
Am 06.03.2012 00:18, schrieb Allin Cottrell:
> On Mon, 5 Mar 2012, Henrique Andrade wrote:
>
>> Dear Gretl Community,
>>
>> In a "Principal Component Analysis" context, is there an accessor
that
can
>> be used for retrieve the component loading coefficients into a matrix?
>
> No, but you can easily get such a matrix by using eigensym() on the
> correlation (or covariance) matrix of the list of series (converted
> to a matrix). Example:
>
> <hansl>
> open data7-12
> list L = wbase length width height weight cyl liters
> matrix C = mcorr({L})
> matrix V
> eigensym(C,&V)
> print V # eigenvectors = loadings
> # compare the PCA printout
> pca L
> </hansl>
>
> Note that eigensym() orders things by ascending value, so the
> loadings associated with the first principal component are in the
> last column of V.
>
I'm wondering whether there may be a case to extend the princomp()
function somewhat to get a direct access to that (optional third matrix
parameter in pointer form for example). Sure it would be sort of
redundant from a syntactic point of view, but since those eigenvectors
have to be computed anyway it could be an efficient way to get those.
cheers,
sven
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
--
Henrique Andrade*
*
4453
days inactive
4454
days old
5 comments
3 participants
participants (3)
-
Allin Cottrell -
Henrique Andrade -
Sven Schreiber