from knock30 import load_morpheme_list
from knock36 import get_word_frequency
from typing import List, Dict, Tuple, Iterable
from matplotlib import pyplot as plt
from matplotlib.font_manager import FontProperties
from os.path import expanduser
# 日本語フォントの読み込み
fp = FontProperties(
fname="/".join([expanduser("~"), "Library", "Fonts", "NotoSansCJKjp-DemiLight.otf"])
)
def zipf(freq: Iterable[Tuple[str, int]]) -> None:
value = [v for _, v in freq] # type: List[int]
rank = range(1, len(value) + 1)
plt.scatter(rank, value, s=1)
plt.title("Zipfの法則", fontproperties=fp)
plt.xlabel("単語の出現頻度順位", fontproperties=fp)
plt.ylabel("単語の出現頻度", fontproperties=fp)
plt.xscale("log")
plt.yscale("log")
plt.show()
if __name__ == "__main__":
zipf(get_word_frequency(load_morpheme_list()))
# 31. 動詞
# 動詞の表層形をすべて抽出せよ．
from knock30 import load_morpheme_list
from typing import List, Dict
M = List[Dict[str, str]]
def get_verbs(morphemes: M) -> List[str]:
verbs = [] # type: List[str]
for morpheme in morphemes:
if morpheme["pos"] == "動詞":
verbs.append(morpheme["surface"])
return verbs
if __name__ == "__main__":
for verb in get_verbs(load_morpheme_list()):
print(verb)
# 33. サ変名詞
# サ変接続の名詞をすべて抽出せよ．
from knock30 import load_morpheme_list
from typing import List, Dict
M = List[Dict[str, str]]
def get_sahen_noun(morphemes: M) -> List[str]:
nouns = [] # type: List[str]
for morpheme in morphemes:
if morpheme["pos1"] == "サ変接続":
nouns.append(morpheme["surface"])
return nouns
if __name__ == "__main__":
for noun in get_sahen_noun(load_morpheme_list()):
print(noun)
# 34. 「AのB」
# 2つの名詞が「の」で連結されている名詞句を抽出せよ
from knock30 import load_morpheme_list
from typing import List, Dict
M = List[Dict[str, str]]
def get_noun_phrases(m: M) -> List[str]:
noun_phrases = []
for i in range(len(m) - 2):
if m[i]["pos"] == m[i + 2]["pos"] == "名詞" and m[i +
1]["surface"] == "の":
noun_phrases.append(m[i]["surface"] + m[i + 1]["surface"] +
m[i + 2]["surface"])
return noun_phrases
if __name__ == "__main__":
for noun_phrase in get_noun_phrases(load_morpheme_list()):
print(noun_phrase)
from knock30 import load_morpheme_list
from knock36 import get_word_frequency
from typing import Dict, List, Tuple, Iterable
from matplotlib import pyplot as plt
from matplotlib.font_manager import FontProperties
from os.path import expanduser
# 日本語フォントの読み込み
fp = FontProperties(fname="/".join(
[expanduser("~"), "Library", "Fonts", "NotoSansCJKjp-DemiLight.otf"]))
def plot_word_frequency(freq: List[Tuple[str, int]], num: int) -> None:
labels = [s[0] for s in freq[:num]]
counts = [s[1] for s in freq[:num]]
axis = [i for i in range(num)]
plt.bar(axis, counts, tick_label=labels, align="center")
plt.xticks(range(num), labels, fontproperties=fp)
plt.title(f"頻度上位 {num} 語", fontproperties=fp)
plt.xlabel(f"出現頻度が高い {num} 語", fontproperties=fp)
plt.ylabel("出現頻度", fontproperties=fp)
plt.show()
if __name__ == "__main__":
result = get_word_frequency(load_morpheme_list())
plot_word_frequency(result, 10)
# 32. 動詞の原形
# 動詞の原形をすべて抽出せよ．
from knock30 import load_morpheme_list
from typing import List, Dict
M = List[Dict[str, str]]
def get_base_form_verbs(morphemes: M) -> List[str]:
verbs = [] # type: List[str]
for morpheme in morphemes:
if morpheme["pos"] == "動詞":
verbs.append(morpheme["base"])
return verbs
if __name__ == "__main__":
for verb in get_base_form_verbs(load_morpheme_list()):
print(verb)