I'm currently trying to install a poetry generator on my virtualmachine - running ubunutu. When I run the server, and enter the text, and hit generate, I get the below error:
Exception Type: IOError
Exception Value:
[Errno 2] No such file or directory: 'poetry_generator/resources/bigram.pickle'
Exception Location: /home/lee/Downloads/PoEmo-master/poetry_generator/architecture/experts/generating_experts/collocation_expert.py in train, line 31
There is no /resources folder within the poetry_generator, and no bigram.pickles file within it either. It identifies linke 31 of the follow script as the issue:
import nltk
import os
import pickle
from pattern import en
from poetry_generator.structures.word import Word
from poetry_generator.architecture.experts.generating_experts.word_generating_expert import WordGeneratingExpert
from poetry_generator.settings import resources_dir
class CollocationExpert(WordGeneratingExpert):
"""Generating most common contexts for words for words"""
def __init__(self, blackboard):
super(
CollocationExpert,
self).__init__(
blackboard,
"Collocation Expert")
self.word_tag_pairs = []
def train(self):
bigram_pickle_file = os.path.join(resources_dir, 'bigram.pickle')
try:
with open(bigram_pickle_file,'rb') as f:
self.word_tag_pairs = pickle.load(f)
except IOError:
tagged_words = nltk.corpus.brown.tagged_words(tagset='universal')
self.word_tag_pairs = list(nltk.bigrams(tagged_words))
with open(bigram_pickle_file,'w') as f:
pickle.dump(self.word_tag_pairs,f)
'''Finding verbs for noun '''
def _find_verbs(self, word):
word_bigrams = [(a[0], b[0]) for a, b in self.word_tag_pairs
if a[0] == word.name and a[1] == 'NOUN' and b[1] == 'VERB'
and en.conjugate(b[0], "inf") not in ('be', 'have')]
return self.__get_best_collocations(word, word_bigrams)
'''Finding adjectives for noun'''
def _find_epithets(self, word):
word_bigrams = [(b[0], a[0]) for (a, b) in self.word_tag_pairs
if b[0] == word.name and b[1] == 'NOUN' and a[1] == 'ADJ']
epithets = self.__get_best_collocations(word, word_bigrams)
return epithets
'''Finding nouns described by adjective'''
def _find_comparisons(self, adjective):
word_bigrams = [(a[0], b[0]) for (a, b) in self.word_tag_pairs
if a[0] == adjective.name and b[1] == 'NOUN' and a[1] == 'ADJ']
comparisons = self.__get_best_collocations(adjective, word_bigrams)
return comparisons
'''Adding epithets for noun to pool'''
def _add_epithets(self, word):
epithets = set([Word(w, "JJ") for w in self._find_epithets(word)])
if word not in self.blackboard.pool.epithets:
self.blackboard.pool.epithets[word] = []
self.blackboard.pool.epithets[word] += list(epithets)
return epithets
def _add_verbs(self, word):
verbs = set([Word(w, "V") for w in self._find_verbs(word)])
self.blackboard.pool.verbs[word] = list(verbs)
return verbs
'''Adding nouns for adjectives to pool'''
def _add_comparisons(self, adj):
comparisons = set([Word(w, "N") for w in self._find_comparisons(adj)])
self.blackboard.pool.comparisons[adj] = comparisons
return comparisons
def __get_best_collocations(self, word, word_bigrams, n=20):
words = nltk.ConditionalFreqDist(word_bigrams)[word.name]
best_bigrams = sorted(words.items(), key=lambda (k, v): v, reverse=False)[:n]
return dict(best_bigrams).keys()
def generate_words(self):
super(CollocationExpert, self).generate_words()
counter = 0
for w in self.blackboard.pool.nouns:
eps = self._add_epithets(w)
vs = self._add_verbs(w)
le = len(eps)
lv = len(vs)
counter += le + lv
for adj in self.blackboard.pool.adjectives:
comps = self._add_comparisons(adj)
counter += len(comps)
return counter
any ideas? I'm new to python so I'm not really clued up. Thanks!
