Do we need a "Simple NILM Metadata" schema as well as "NILM Metadata" schema?

[Jack Kelly]

NILM Metadata tries to make it possible to capture pretty much any conceivable scenario. But, as more datasets become available, it appears that a large proportion of datasets could be described using a simpler metadata schema. It would be great to discuss the design of a "Simple NILM Metadata" schema which exists along side "NILM Metadata". Perhaps CSV is even easier to read than YAML in Matlab, Java etc so it might be nice if we can use CSV. We'd have a check-list to help people decide whether they require the full expressive power of "NILM Metadata" or if they can get by with "Simple NILM Metadata".

The full proposal is here: https://github.com/nilmtk/nilm_metadata/issues/16   (and I'd suggest we discuss the proposal on the issue queue rather than here).

[Liudmila Belenki]

Hi Jack,

could you help me to convert the .yaml files to the hdf5 format ? 

I am trying to use nilm_medata.convert_yaml_tohdf5(dir_yaml, hdf5_filename) but with no success.

I get error: 

 ----

 File "tables\hdf5extension.pyx", line 755, in tables.hdf5extension.AttributeSet._g_getattr (tables\hdf5extension.c:8149)

tables.exceptions.HDF5ExtError: HDF5 error back trace

  File "C:\aroot\work\hdf5-1.8.15-patch1\src\H5A.c", line 476, in H5Aopen_by_name

    can't open attribute

  File "C:\aroot\work\hdf5-1.8.15-patch1\src\H5Aint.c", line 451, in H5A_open_by_name

    unable to load attribute info from object header

  File "C:\aroot\work\hdf5-1.8.15-patch1\src\H5Oattribute.c", line 537, in H5O_attr_open_by_name

    can't locate attribute: 'pandas_type'

End of HDF5 error back trace

Can't get type info on attribute pandas_type in node /.

Closing remaining open files:C:\NILMTK\nilmtk\data\test_2015-11-30_2015-12-01_csv\test_2015-11-30_2015-12-01_yaml_.h5...done

()

['TITLE', 'CLASS', 'VERSION']

TITLE

['TITLE', 'CLASS', 'VERSION']

CLASS

['TITLE', 'CLASS', 'VERSION']

VERSION

['CLASS', 'TITLE', 'VERSION']

pandas_type

Process finished with exit code 1

 

Thanks.

Regards

Liudmila