mbkumar / pycdt Goto Github PK

I have download the latest anaconda3(python 3.8) and pycdt(2.0.5), when I try to run the example code, there is a werid ERROR： Illegal instruction (core dumped)
I tried the code on both super computing(Linux AMD roma) and my workstation(Ubuntu18 Intel)

The code:
from pymatgen.ext.matproj import MPRester
with MPRester() as mpr:
mgo_struct = mpr.get_structure_by_material_id('mp-1265')
from pycdt.core.defectsmaker import DefectChargerInsulator
defect_charger = DefectChargerInsulator(mgo_struct)
mg_vac_charges = defect_charger.get_charges('vacancy',site_specie='Mg')

from pycdt.core.defectsmaker import ChargedDefectsStructures
charged_defects_structs = ChargedDefectsStructures(mgo_struct)
from pycdt.utils.vasp import DefectRelaxSet
defect_struct = charged_defects_structs.get_ith_supercell_of_defect_type(0, 'vacancies')
vasp_input = DefectRelaxSet(defect_struct) # defect_struct is defectsupercell
vasp_input.write_input(output_dir)
vasp_input = DefectRelaxSet(defect_struct, potcar_functional="PBE")
user_settings = {'potcar_functional': 'PBE','user_incar_settings': {'EDIFFG': -1e-3, 'EDIFF': 1e-8, 'LASPH': True}}
vasp_input = DefectRelaxSet(defect_struct,**user_settings)

>python setup.py install
>error in pycdt setup command: 'install_requires' must be a string or list of strings containing valid project/version requirement specifiers; Expected end or semicolon (after version specifier)
    matplotlib>=3.1monty>=3.0.2
              ~~~~~^

Seems that on line 16 in 'setup.py' a separator is missing between "matplotlib>=3.1" and "monty>=3.0.2". Adding one fixes the above error.

install_requires=["numpy>=1.18.15", "pymatgen>=2020.6.8,<=2022.7.8", "matplotlib>=3.1""monty>=3.0.2"]

With version 2.0, indices included in folder naming from ChargedDefectStructures is not consistent with the sublattice site anymore.

A fix for this can make use of the PointDefectComparator in pymatgen to assign sublattice indices.

make use of additional interstitial features. Several exist within pymatgen.analysis.defects.generators right now. Would not be hard to add them to the pycdt command line script.

Confirm that defects being used have the same metadata as that of the Materials Project. If a different U parameter is used or if a different set of pseudopotentials is used then chemical potentials from Materials Project should likely not be used.

Hi,

I'm getting error "cannot import name 'str2unicode' from 'monty.string'" when I try to generate the input using the PbTiO3.json example. I'm using PyCDT-2.0.2 with latest Pymatgen-2020 and python3.8.1

I can see that there have been no new commits to PyCDT for ~1 year. Is the project still actively maintained/will there be any updates?

These corrections are not in pymatgen right now - only in PyCDT. They should be documented better and possibly merged to pymatgen.

Add capabilities for setting up and running follow up calculations (larger supercells or perturbations) based on metadata that exists locally.

This can make use of pymatgen / atomate / emmet features.