Tuesday, 15 September 2015

How can I turn on HDF5 Dense Attribute Storage in PyTables/Python? -


Since 1.8.8 HDF 5 offers several HDF 5 nodes with many or large features for fast handling. To start dense collection, the threshold is set to the default ( H5Pset_attr_phase_change ) 8 attributes.

Is there any way to change it from <0 to Python or even better through 0 (i.e., always using dense storage)?


Edit: If it can not be changed via PyTables, but to be done externally, then it is important that this external change still Affect PyTables. Thus, all HDF 5 attributes to be added through PETABLES should be deposited in a dense quantity.

It seems that access to the appropriate attribute has not yet been written:

The process of wrapping new attributes appears to be well documented here:

With a practical example it can be 4 or 5 (relatively) There are simple steps ... If you are accustomed to compile, then you should take an hour or less.

Is this help? Edit: PETTLEDOX: "Instead, PETABLES have created an extra abstract layer on top of HDF 5 [1] and NMPE [3], where it implements the enhanced things ... "I have the impression that it is far more than the HDF 5 library, and therefore, these options can be more difficult to reach.

As PyTables does not use the h5py, in h5py No changes will help you ...


No comments:

Post a Comment