This is not necessarily true. Many people store this type of data in PyTables, which is the backing store for Pandas and uses HDF5 internally.
I'm travelling today, but I'll provide a real response to this in a day or two. Dana > -----Original Message----- > From: Hdf-forum [mailto:[email protected]] On Behalf Of > ?????? ????????? > Sent: Monday, October 27, 2014 8:58 AM > To: [email protected] > Subject: Re: [Hdf-forum] Best structure for an Hdf5 file to store > historical data and how to access quickly from python > > 26.10.2014 21:53, Dan E пишет: > > I have equity options historical data, now in csv files - one file for > > each day, that I'd like to store in one or multiple h5 files for > > efficient resource usage and fast access. One row of data is about a > > dozen columns (date, symbol name, options characteristics (maturiy, > > exercise price etc..), price (open/high/low/close), volume + some > > other infos). One day is about 700,000 rows accross 4,000 different > symbols. > > Hi Dan! > > It seems that indexed SQLite database would do the job for you. > > HDF5 is best for homogeneous array-like primitive-type data. However, you > are going to store tabular data. In any case, I recommend you to test > proposed HDF5 solutions vs SQLite solution. > > Best wishes, > Andrey Paramonov > > > -- > This message has been scanned for viruses and dangerous content by > MailScanner, and is believed to be clean. > > > _______________________________________________ > Hdf-forum is for HDF software users discussion. > [email protected] > http://mail.lists.hdfgroup.org/mailman/listinfo/hdf- > forum_lists.hdfgroup.org > Twitter: https://twitter.com/hdf5 _______________________________________________ Hdf-forum is for HDF software users discussion. [email protected] http://mail.lists.hdfgroup.org/mailman/listinfo/hdf-forum_lists.hdfgroup.org Twitter: https://twitter.com/hdf5
