SC_AI/cand_predict.py at main · 2MeterScale/SC_AI · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
# materials projectからNNに突っ込んで予測するための
# 情報を取得するためのソースコード
# ひとまずホウ化物？とかになるのかなという感じ
# 例えばホウ化物で構成元素が3つ以上とか

# %%
import os

import pandas as pd
import periodictable
from chempy import Substance
from mp_api.client import MPRester
# import ast

pd.options.display.max_columns = 100
pd.options.display.max_rows = 100

MY_API_KEY = "09vnTU6R8jWfPhuZ6bFs3niTwe1F03x4"

# Output directory
path_output_dir = "/workspaces/codespace_practice/candidate_material_list_final/"
if not os.path.exists(path_output_dir):
    os.makedirs(path_output_dir)


# %%
with MPRester(MY_API_KEY) as mpr:
    res = mpr.materials.summary.search(formula = "SrTiO3", is_stable=True)
    #print(res)

if len(res[0].database_IDs) != 0:
    ID_bool = True
else:
    ID_bool = False
print(ID_bool)
print(res[0].database_IDs)
keys = res[0].database_IDs.keys()
for key in res[0].database_IDs.keys():
    print(res[0].database_IDs[key])
#print(res[0].database_IDs[keys][0])


# %%
# define physical properties/infos you want to obtain
properties = ["material_id", "formula_pretty", "band_gap", "efermi", "energy_above_hull", "database_IDs"]
# nameは*の数で構成する元素数を指定できる？
name = "**"
_is_stable = True

# 他にはBa, Sr, Niとか?
# 上原先生の助言の元,C,Al,Si,Li,B,Niで半導体周りでやってみる
# 酸素を除外したデータでもやってみる
# 酸素除外について、あまり意味がないというのと
# 窒化物を研究室で作ることが出来ないのでこれを除外する

element_filter = ["C", "Al", "Si", "Li", "B", "Ni", "Mg", "Ca", "K"]


with MPRester(MY_API_KEY) as mpr:
    for atm in element_filter:
        results = mpr.materials.summary.search(fields=properties, elements=[atm], is_stable=_is_stable)

        cand_df = pd.DataFrame(columns=["formula", "material_id", "band_gap", "efermi", "energy_above_hull", "database_IDs"])

        for mat in results:
            if len(mat.database_IDs) != 0:
                database_keys = mat.database_IDs.keys()
                for key in database_keys:
                    database_ID = mat.database_IDs[key][0]
                    break

                subdata = {"formula": mat.formula_pretty, "material_id": mat.material_id, "band_gap": mat.band_gap, "efermi": mat.efermi, "energy_above_hull": mat.energy_above_hull, "database_IDs": database_ID}
                subdf = pd.DataFrame(subdata, index=[0])

                cand_df = pd.concat([cand_df, subdf], axis=0)

        cand_df = cand_df.reset_index(drop=True)
        cand_df["comp_atoms"] = cand_df["formula"].apply(lambda x: Substance.from_formula(x).composition)
        cand_df["sorted_keys_comp_atoms"] = cand_df["formula"].apply(lambda x: sorted(Substance.from_formula(x).composition.keys()))
        cand_df["comp_number"] = cand_df.apply(lambda row: "".join(f'{k}{row["comp_atoms"][k]}' for k in row["sorted_keys_comp_atoms"]), axis=1)

        chosen_material_list = []

        for i, subdf in cand_df.groupby("comp_number"):
            if len(subdf) > 1:
                subdf = subdf.sort_values("energy_above_hull")
                chosen_material_list.append(subdf.index[0])

            else:
                chosen_material_list.append(subdf.index[0])

        chosen_cand_df = cand_df.loc[chosen_material_list]

        #chosen_cand_df = chosen_cand_df.dropna(subset=["efermi"])

        chosen_cand_df = chosen_cand_df.reset_index(drop=True)

        chosen_cand_df["new_formula"] = None

        for i in range(len(chosen_cand_df)):
            dict_keys = list(chosen_cand_df.loc[i, "comp_atoms"].keys())

            comp = ""

            for j in range(len(dict_keys)):
                element = str(periodictable.elements[dict_keys[j]])
                comp += element + str(chosen_cand_df.loc[i, "comp_atoms"][dict_keys[j]])

            chosen_cand_df.loc[i, "new_formula"] = comp

        #display(chosen_cand_df)
        #print(chosen_cand_df.isnull().sum())

        #break

        chosen_cand_df.to_csv(path_output_dir + atm + "_is_stable_candidate_material_list.csv", index=False)


# %%