![](/img/trans.png)
[英]TypeError: unhashable type: 'list' for nltk.FreqDist even though I've converted my list into a tuple
[英]Showing “TypeError: unhashable type: 'list'” while using it in nltk.FreqDist() function
我正在嘗試使用nltk根據短語的程度獲取單詞中單詞的頻率分布。 它雖然顯示“ TypeError:不可哈希類型:'列表'”。 不明白是什么問題。 請幫忙。
PS:該代碼有很多錯誤,所以不要介意。 我正在嘗試使用許多程序的代碼片段來構建關鍵字提取器,因為我對python不熟悉。 如果有人想闡明其他錯誤,也歡迎您。
碼:
from __future__ import division
import operator
import nltk
import string
def isPunct(word):
return len(word) == 1 and word in string.punctuation
def isNumeric(word):
try:
float(word) if '.' in word else int(word)
return True
except ValueError:
return False
class KeyExt:
def __init__(self):
self.stopwords = set(nltk.corpus.stopwords.words())
self.top_fraction = 1
def _generate_candidate_keywords(self, sentences):
phrase_list = []
for sentence in sentences:
words = map(lambda x: "|" if x in self.stopwords else x, nltk.word_tokenize(sentence.lower()))
phrase = []
for word in words:
if word == "|" or isPunct(word):
if len(phrase) > 0:
phrase_list.append(phrase)
phrase = []
else:
phrase.append(word)
return phrase_list
def _calculate_word_scores(self, phrase_list):
word_freq = nltk.FreqDist()
word_degree = nltk.FreqDist()
for phrase in phrase_list:
degree = [x for x in phrase if not isNumeric(x)]
for word in phrase:
word_freq[word]=word_freq[word]+1
word_degree[word, degree]=word_degree[word, degree]+1
for word in word_freq.keys():
word_degree[word] = word_degree[word] + word_freq[word]
word_scores = {}
for word in word_freq.keys():
word_scores[word] = word_degree[word] / word_freq[word]
return word_scores
def _calculate_phrase_scores(self, phrase_list, word_scores):
phrase_scores = {}
for phrase in phrase_list:
phrase_score = 0
for word in phrase:
phrase_score += word_scores[word]
phrase_scores[" ".join(phrase)] = phrase_score
return phrase_scores
def extract(self, text, incl_scores=False):
sentences = nltk.sent_tokenize(text)
phrase_list = self._generate_candidate_keywords(sentences)
word_scores = self._calculate_word_scores(phrase_list)
phrase_scores = self._calculate_phrase_scores(phrase_list, word_scores)
sorted_phrase_scores = sorted(phrase_scores.items(), key=operator.itemgetter(1), reverse=True)
n_phrases = len(sorted_phrase_scores)
if incl_scores:
return sorted_phrase_scores[0:int(n_phrases/self.top_fraction)]
else:
return map(lambda x: x[0],
sorted_phrase_scores[0:int(n_phrases/self.top_fraction)])
def test():
search=input("Enter Text: ")
ke = KeyExt()
keywords = ke.extract(search, incl_scores=True)
print (keywords)
if __name__ == "__main__":
test()
完整回溯:
Traceback (most recent call last): File "C:\Users\SAURAV
DAS\AppData\Local\Programs\Python\Python35\projects\nlpproj.py", line
81, in <module>
test() File "C:\Users\SAURAV DAS\AppData\Local\Programs\Python\Python35\projects\nlpproj.py", line
77, in test
keywords = ke.extract(search, incl_scores=True) File "C:\Users\SAURAV
DAS\AppData\Local\Programs\Python\Python35\projects\nlpproj.py", line
64, in extract
word_scores = self._calculate_word_scores(phrase_list) File "C:\Users\SAURAV
DAS\AppData\Local\Programs\Python\Python35\projects\nlpproj.py", line
44, in _calculate_word_scores
word_degree[word, degree]=word_degree[word, degree]+1 TypeError: unhashable type: 'list'
當您嘗試使用list
作為dict
鍵時,會發生此錯誤。 list
不可散列,不能用作鍵(例如,可以使用tuple
)。
在這種情況下
word_degree[word, degree]=word_degree[word, degree]+1
您正在使用度作為word_degree
的索引,但沒有意義,因為度本身就是一個list
。
聲明:本站的技術帖子網頁,遵循CC BY-SA 4.0協議,如果您需要轉載,請注明本站網址或者原文地址。任何問題請咨詢:yoyou2525@163.com.