Table data

Read table data as following format, and organize by composition.

Data

F0

F1

…

H

V

V

…

He

V

V

…

Li

V

V

…

Be

V

V

…

…

…

…

…

  • Then run the code.

>>> from featurebox.featurizers.atom.mapper import AtomTableMap
>>> tmps = AtomTableMap(search_tp="name",tablename=your_pd_dataframe)
>>> com = [{"H": 2, }, {"Po": 1}, {"C": 2}]
>>> a = tmps.convert(com)
../_images/2_1.png

  • In default, the proportion would be multiplied in data, also you can neglect weight.

>>> tmps = AtomTableMap(search_tp="name", weight=False, tablename=your_pd_dataframe)
>>> com = [{"H": 2, }, {"Po": 1}, {"C": 2}]
>>> a2 = tmps.convert(com)
../_images/2_2.png

  • Index by structure

>>> tmps = AtomTableMap(search_tp="number",tablename="oe.csv")
>>> a = tmps.convert(structure)

Note

>>> com = [i.species.as_dict() for i in structure.sites]

or

>>> com =  [{str(i.symbol): 1} for i in structure.species]