Re: [SQLObject] storing objects that have matrices as attributes

[Ryan K. <rya...@co...>]

I thought I had replied to this, but don't see it any where in my sent 
folder.  So, please forgive me if this is the second time I have sent this.

The PickleCol works great.  It is a neat idea and it does exactly what I 
need in this situation.  But the files are very large.  SciPy has a 
function called shelve and using it to store the same data seems to 
result in files about 1/10th the size of using PickleCol and matrices.  
I think shelve uses some sort of compression or something.  Is there 
away to use compression with PickleCol?

Thanks,

Ryan

Ian Bicking wrote:

> Ryan Krauss wrote:
>
>> Maybe an example would make this clearer.  What if I want to store a 
>> collection of objects with the following properties:
>> entry1.input='a1'
>> entry1.output='j2'
>> entry1.datafile=['file1','file2','file3']
>> entry1.data1=[1.2,1.47,3.56,....]
>> entry1.data2=[-1.5,-19.5,35.5,....]
>>
>> And what if data1 and data2 are Numeric arrays/matrices (or from 
>> numarray)?  Can SQLObject store a collection of such objects?
>
>
> You can use PickleCol for this, so long as its okay that the data is 
> opaque to the database and to queries.  Or you could create a 
> _set_data1() method that inserts the values into another table, and 
> _get_data1() that retrieves them.
>