-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathcard2vec.py
177 lines (158 loc) · 7.95 KB
/
card2vec.py
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
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
# -*- coding: utf-8 -*-
import json
#import MeCab
from gensim.models import doc2vec
import os
TARGET_GAME_NAME = "all"
def load_json(target_game_name,file_name):
# カード名とカードテキストの入力データ作成
names = []
text = ""
texts = []
# Mecabの出力を分かち書きに指定
#mecab = MeCab.Tagger("-Owakati")
json_path = target_game_name + ".json"
#import stanza
#nlp = stanza.Pipeline(lang='en', processors='tokenize,lemma')
# カードのテキストを形態素解析し、分かち書きしたものを改行区切りで一つのstringにする
with open(json_path, "r") as file:
card_dict = json.load(file)
card_name_list = []
for card_id in card_dict:
card_name_list.append(card_dict[card_id]["name_"])
for card_id in card_dict:
card = card_dict[card_id]
#print(card.keys())
if card["name_"] not in names:
names.append(card["name_"])
pp = str(card["pp_"]) #"pp: " +
card_type = card["type_"] #"type: " +
# pp = "pp: " + str(card["pp_"])
# craft = "craft: " + card["craft_"].replace("craft","")
# card_type = "type: " + card["type_"]
# base_atk = "baseAtk: -"
# base_def = "baseDef: -"
# evo_atk = "evoAtk : -"
# evo_def = "evoDef: -"
# if "Follower" in card_type:
# base_atk = "baseAtk: " + str(card["baseAtk_"])
# base_def = "baseDef: " + str(card["baseDef_"])
# evo_atk = "evoAtk: " + str(card["evoAtk_"])
# evo_def = "evoDef: " + str(card["evoDef_"])
base_effect = card["baseEffect_"] #"baseEffect: " +
base_effect = base_effect.replace("\n","$")
evo_effect = card["evoEffect_"] # "evoEffect: " +
evo_effect = evo_effect.replace("\n","$")
if "Same as the unevolved form" in evo_effect:
evo_effect = card["baseEffect_"]
if "----------" in evo_effect:
evo_effect = evo_effect.split("----------")[1]
except_words = ["Enhance","Accelerate","Crystallize","Fanfare","Invocation",\
"When this follower comes into play"]
for except_word in except_words:
if except_word in evo_effect:
evo_effect = evo_effect.split("$")
evo_effect = [sentence for sentence in evo_effect if except_word not in sentence]
if len(evo_effect) > 0:
evo_effect = "$".join(evo_effect)
else:
evo_effect = "-"
default_text = card["baseEffect_"]
if "Fanfare" in default_text and "Last Words" in default_text:
evo_effect += "Last Words:" + default_text.split("Last Words:")[-1]
evo_effect = "evoEffect: " + evo_effect
base_effect = base_effect.replace("$", " ")
evo_effect = evo_effect.replace("$"," ")
parse_result = base_effect + " " + evo_effect
parse_result = parse_result.replace(".", " . ")
parse_result = parse_result.replace("/", " / ")
parse_result = parse_result.replace(":", " : ")
parse_result = parse_result.lower()
for name in card_name_list:
if name.lower() in parse_result:
new_name = name.replace(" ","")
new_name = new_name.replace(", ",",")
parse_result = parse_result.replace(name.lower(),new_name)
parse_result = parse_result.replace(", ", " , ")
parse_result = pp + "\n" + card_type + "\n" + parse_result
#doc = nlp(parse_result)
#parse_result = " ".join([word.lemma for sent in doc.sentences for word in sent.words])
# parse_result = craft + " " + card_type + "\n"
# parse_result += pp + "\n"
# parse_result += base_atk + " " + base_def + "\n"
# parse_result += base_effect + "\n"
# parse_result += evo_atk + " " + evo_def + "\n"
# parse_result += evo_effect
#mecab_result = mecab.parse(card["text"])
if parse_result is False:
text += "\n"
texts.append("")
else:
#text += parse_result + "\n"
texts.append(parse_result)
text += parse_result.replace("\n"," ") + "\n"
#names.append(card["name_"]+"(EVO)")
#parse_result=card["evoEffect_"]
#parse_result = parse_result.replace("\n"," ")
#if parse_result is False:
# text += "\n"
# texts.append("")
#else:
# text += parse_result + "\n"
# texts.append(card["evoEffect_"])
TARGET = target_game_name +"_"+file_name+ ".txt"
if os.path.isfile(TARGET) is False:
with open(TARGET, "w") as file:
file.write(text)
return names, texts
def generate_doc2vec_model(target_game_name,size=300,window=8):
print("Training Start")
# カードテキスト読み込み
card_text = doc2vec.TaggedLineDocument(target_game_name + ".txt")
# 学習
model = doc2vec.Doc2Vec(card_text, size=size, window=window, min_count=1,
workers=4, iter=400, dbow_words=1, negative=10,dm=0)
# モデルの保存
model.save(target_game_name + ".model")
print("Training Finish")
return model
if __name__ == '__main__':
import argparse
parser = argparse.ArgumentParser(description='Card2Vec学習コード')
parser.add_argument('--vector_size', help='カードベクトルのサイズ', type=int, default=300)
parser.add_argument('--window_size', help='窓サイズ', type=int, default=8)
parser.add_argument('-visualize',help='ベクトルの可視化')
args = parser.parse_args()
names, texts = load_json(TARGET_GAME_NAME,"{}_{}".format(args.vector_size,args.window_size))
target_file_name = "{}_{}_{}".format(TARGET_GAME_NAME,
args.vector_size,args.window_size)
if os.path.isfile(target_file_name + ".model") is True:
model = doc2vec.Doc2Vec.load(TARGET_GAME_NAME + ".model")
else:
model = generate_doc2vec_model(target_file_name,size=args.vector_size,window=args.window_size)
if args.visualize is not None:
init_word = input("input initial word:")
positive_word = input("input positive word:")
negatice_word = ""#("input negative word:")
query = model.most_similar(positive=[init_word,positive_word])#,negative=[negatice_word])
print(query)
#print(names[card_index])
#print(texts[card_index])
#print(model.docvecs[card_index])
TARGET_CARD_NAME = input("input card_name:")
card_index = names.index(TARGET_CARD_NAME)
print(model.docvecs[card_index])
print(model.docvecs["Goblin"])
reverse_flg = input("reverse?(y/n):") == "y"
# 類似カードと類似度のタプル（類似度上位10件）のリストを受け取る
similar_docs = model.docvecs.most_similar(card_index,topn=len(names))
print("name_len:",len(names),"txt_len:",len(texts))
print(names[card_index])
print(texts[card_index])
print("--------------------is similar to--------------------")
if reverse_flg:
similar_docs = list(reversed(similar_docs))
for similar_doc in similar_docs[0:10]:
#print("similar_doc:",similar_doc)
print(names[similar_doc[0]] + " " + str(similar_doc[1]))
print(texts[similar_doc[0]], "\n")