Hi,
I am writing large number of data-sets in a HDF5 file. they are one
dimensional compound data-sets.
The compound data type is as follows...
typedef struct tick_t {
int64_t bid_trade_value;
int64_t ask_value;
int32_t cap_sec;
int32_t cap_usec;
int32_t exg_sec;
uint32_t exg_seq_id;
uint32_t tfp_seq_id;
uint32_t lrt_id;
uint32_t bid_trade_size;
uint32_t ask_size;
uint32_t flags2;
uint16_t flags;
uint8_t bid_trade_base;
uint8_t ask_base;
uint8_t bid_trade_exg;
uint8_t ask_exg;
} Tick;
There are large number of data-sets, so I am opening data-set before writing
and closing after the write.
Is type conversion happens for this type of data-sets?
Providing application specified buffers using *H5Pset_buffer(hid_t plist,
hsize_t size, void *tconv, void *bkg )* improves overall writing
performance?
--
View this message in context:
http://hdf-forum.184993.n3.nabble.com/User-provided-Conversion-and-background-buffers-tp4025532.html
Sent from the hdf-forum mailing list archive at Nabble.com.
_______________________________________________
Hdf-forum is for HDF software users discussion.
[email protected]
http://mail.hdfgroup.org/mailman/listinfo/hdf-forum_hdfgroup.org
