Re: RFC: method for storing data

[Jaroslav Hajek]

On Tue, Jun 29, 2010 at 7:15 PM, CdeMills <address@hidden> wrote:
>
> A new implementation, should be much more robust. More polymorphism, more
> ways to manipulate data. Test and enjoy.
>
> Pascal
>
> http://octave.1599824.n4.nabble.com/file/n2272479/dataframe.zip
> dataframe.zip


I need to go through this in more detail, but seems like a good start.
I have a few suggestions off hand:

1. helper funcs (df_name2idx) can reside in a private/ subdirectory.
2. when working with fopen et al, always ensure proper closing in
library functions. Use unwind_protect or the newly supported onCleanup
3. To properly support the element deletion syntax A(I) = [], use
isnull() instead of isempty() to test the rhs. This correctly
distinguishes between A(I) = [] and A(I) = B where B just happens to
be 0x0. (Apparently, in Matlab this is not possible).


-- 
RNDr. Jaroslav Hajek, PhD
computing expert & GNU Octave developer
Aeronautical Research and Test Institute (VZLU)
Prague, Czech Republic
url: www.highegg.matfyz.cz