[This message was meant to be posted here] ---------------------------------------------------------------------------- De: Ralf Juengling <ralf.juengl...@synopsys.com> A: "pytables-us...@lists.sf.net" <pytables-us...@lists.sf.net> Data: Avui 01:45:50 > Hello, > > I was just using PyTables for the first time (version 2.1.2), created an hf5 > file holding an EArray under root, appended a sequence of arrays to that > EArray, and let all the objects related to the hf5 file go out of scope > without explicitly calling a flush or close method. > > The file was empty. Do the PyTable's destructors not invoke 'close' to get > all pending data to disk when an object is destructed? > > Ralf
No, PyTables should close the file automatically for you when you exit the Python interpreter, and in fact it issues a message telling you so: """ Closing remaining open files: /tmp/test-seq.h5... done """ so, in order to avoid the message, it is always better to explicitly close the file by hand. The problem you were having was related with a bug in PyTables: it was not considering the shape of multidimensional atoms when computing automatic chunksizes. See: http://pytables.org/trac/ticket/273 for details. Cheers, -- Francesc Alted ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ Pytables-users mailing list Pytables-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/pytables-users
