Complex Types

[ax3l]

Check how the Python dtypes numpy.complex64 and numpy.complex128 look in HDF5 and implement as new types.

[ax3l]

import h5py as h5
import numpy as np

f = h5.File("complex.h5")
f["data64"] = np.ones((10,20,), dtype=np.complex64)
f["data128"] = np.ones((10,20,), dtype=np.complex128)
f.close()

$ h5ls -v complex.h5                                                     
Opened "complex.h5" with sec2 driver.
data128                  Dataset {10/10, 20/20}
    Location:  1:1400
    Links:     1
    Storage:   3200 logical bytes, 3200 allocated bytes, 100.00% utilization
    Type:      struct {
                   "r"                +0    native double
                   "i"                +8    native double
               } 16 bytes
data64                   Dataset {10/10, 20/20}
    Location:  1:800
    Links:     1
    Storage:   1600 logical bytes, 1600 allocated bytes, 100.00% utilization
    Type:      struct {
                   "r"                +0    native float
                   "i"                +4    native float
               } 8 bytes

$ h5dump complex.h5
HDF5 "complex.h5" {
GROUP "/" {
   DATASET "data128" {
      DATATYPE  H5T_COMPOUND {
         H5T_IEEE_F64LE "r";
         H5T_IEEE_F64LE "i";
      }
      DATASPACE  SIMPLE { ( 10, 20 ) / ( 10, 20 ) }
      DATA { ... }
   }
   DATASET "data64" {
      DATATYPE  H5T_COMPOUND {
         H5T_IEEE_F32LE "r";
         H5T_IEEE_F32LE "i";
      }
      DATASPACE  SIMPLE { ( 10, 20 ) / ( 10, 20 ) }
      DATA { ... }
   }
}
}