Here's a simpler code snippet to reproduce the error. It appears there is a
maximum number of columns in a table, and it depends on the data type in an
unusual way (at least to me). All floats have one limit and all integers
have another limit, regardless of the bit size. I didn't test strings or
booleans.
1092 for floats (both 32 bit and 64 bit)
1260 for ints and unsigned ints (both 32 bit and 64 bit)
I did some more searching of the email list archives, and I found ticket
#211 which describes a similar problem. It was caused by a limitation in
the HDF5 library. Could this be the same issue?
##formats=['f8'] * 1092 # does not cause an error
formats=['f8'] * 1093
import os.path
import numpy as np
import tables as tb
# change this to a temporary folder where test database can be created
tempfolder = 'C:\\Users\\jayers\\Desktop\\New Folder'
dtype = np.format_parser(formats,[],[]).dtype
h = tb.openFile(os.path.join(tempfolder,'test.h5'),mode='w')
h.createTable('/','data',dtype)
h.close()
------------------------------------------------------------------------------
Start uncovering the many advantages of virtual appliances
and start using them to simplify application deployment and
accelerate your shift to cloud computing.
http://p.sf.net/sfu/novell-sfdev2dev
_______________________________________________
Pytables-users mailing list
Pytables-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/pytables-users
