2010/7/21 Shalom Rav
> Hi Francesc,
>
> I did try previously to work with arrays. If I am not mistaken, I then ran
> into a limitation (of pytables?) not allowing say arrays of size
> 30,000 (where tables, on the other hand, are not limited in number of rows).
>
No. *Arrays should not have a li
2010/7/18 Shalom Rav
The problem I described got solved. As per the design, would an array of
> say 20,000 floats be better than a table (with one field), containing 20,000
> rows of floats? (suppose we define "better" in terms of writing time / read
> time).
>
Maybe the difference is not large
Hi,
Thank you so much for your reply.
The problem I described got solved. As per the design, would an array of
say 20,000 floats be better than a table (with one field), containing 20,000
rows of floats? (suppose we define "better" in terms of writing time / read
time).
Thank you.
On Thu, Jul 1
A Thursday 15 July 2010 19:00:08 Shalom Rav escrigué:
> Hi,
>
> I created a PyTables file containing circa 15000 groups, where in each of
> these groups, there is one single table, defined using the following class:
>
>
> class *SimpleScheme*(tables.IsDescription):
>
> value = tables.Floa
Hi,
I created a PyTables file containing circa 15000 groups, where in each of
these groups, there is one single table, defined using the following class:
class *SimpleScheme*(tables.IsDescription):
value = tables.Float32Col(pos = 0, dflt = numpy.NaN)
Each such a table has about 1500 rows
