def get(self, program_id=None):
program = Program.get_by_id(int(program_id))
rendered = parse.flatten(parse.parse_workout(program.program_code))
template_values = {
'json': rendered
}
template = template_file('program_run.html')
self.response.out.write(template.render(template_values))
def batch_to_input(batch, word2index):
contexts, questions, answer_I, answer_II, gold_output = list(zip(*batch))
max_ctxt = max([len(ctxt) for ctxt in contexts])
max_len = max([ctxt.size(1) for ctxt in parse.flatten(contexts)])
max_q = max([qq.size(1) for qq in questions])
max_a_I = max([aa.size(1) for aa in answer_I])
max_a_II = max([aa.size(1) for aa in answer_II])
c_p, c_mask, q_p, a_I_p, a_II_p = [], [], [], [], []
for i in range(len(batch)):
temp_c_p = []
for j in range(len(contexts[i])):
if contexts[i][j].size(1) == max_len:
temp_c_p.append(contexts[i][j])
else:
temp_c_p.append(torch.cat([contexts[i][j], Variable(LongTensor([word2index['<PAD>']]*(max_len-contexts[i][j].size(1)))).view(1,-1)], 1))
while(len(temp_c_p) < max_ctxt):
temp_c_p.append(Variable(LongTensor([word2index['<PAD>']]*max_len)).view(1,-1))
temp_c_p = torch.cat(temp_c_p)
c_p.append(temp_c_p)
c_mask.append(torch.cat([Variable(ByteTensor(tuple(map(lambda s: s ==0, t.data))),volatile=False) for t in temp_c_p]).view(temp_c_p.size(0),-1))
if questions[i].size(1) == max_q:
q_p.append(questions[i])
else:
q_p.append(torch.cat([questions[i],Variable(LongTensor([word2index['<PAD>']]*(max_q-questions[i].size(1)))).view(1,-1)],1))
if answer_I[i].size(1) == max_a_I:
a_I_p.append(answer_I[i])
else:
a_I_p.append(torch.cat([answer_I[i],Variable(LongTensor([word2index['<PAD>']]*(max_a_I-answer_I[i].size(1)))).view(1,-1)],1))
if answer_II[i].size(1) == max_a_II:
a_II_p.append(answer_II[i])
else:
a_II_p.append(torch.cat([answer_II[i],Variable(LongTensor([word2index['<PAD>']]*(max_a_II-answer_II[i].size(1)))).view(1,-1)],1))
q_p = torch.cat(q_p)
a_I_p = torch.cat(a_I_p)
a_II_p = torch.cat(a_II_p)
q_mask = torch.cat([Variable(ByteTensor(tuple(map(lambda s: s ==0, t.data))),volatile=False) for t in q_p]).view(q_p.size(0),-1)
a_I_mask = torch.cat([Variable(ByteTensor(tuple(map(lambda s: s ==0, t.data))),volatile=False) for t in a_I_p]).view(a_I_p.size(0),-1)
a_II_mask = torch.cat([Variable(ByteTensor(tuple(map(lambda s: s ==0, t.data))),volatile=False) for t in a_II_p]).view(a_II_p.size(0),-1)
return c_p, c_mask, q_p, q_mask, a_I_p, a_I_mask, a_II_p, a_II_mask, Variable(LongTensor(gold_output))
import parse from functools import reduce import numpy import matplotlib.pyplot morpheme_data = parse.get_morpheme_from_neko() counter = parse.get_counter_from_words(parse.flatten(morpheme_data)).most_common(10) xticks = numpy.array(reduce(lambda result, count: result + [count[0][0]], counter, [])) x = numpy.array(list(range(1, 11))) y = numpy.array(reduce(lambda result, count: result + [count[1]], counter, [])) matplotlib.pyplot.bar(x, y, align = 'center') matplotlib.pyplot.xticks(x, xticks) matplotlib.pyplot.show()
def flatten(s): return p.flatten(s)
import parse
from functools import reduce
import numpy
import matplotlib.pyplot
from collections import Counter
morpheme_data = parse.get_morpheme_from_neko()
counter_by_word = parse.get_counter_from_words(
parse.flatten(morpheme_data)).items()
counter_by_number = list(
Counter(list(map(lambda count: count[1], counter_by_word))).items())
sorted_by_number = sorted(counter_by_number, key=lambda value: value[0])
x = numpy.array(
reduce(lambda result, count: result + [count[0]], sorted_by_number, []))
y = numpy.array(
reduce(lambda result, count: result + [count[1]], sorted_by_number, []))
matplotlib.pyplot.bar(x, y, align='center')
matplotlib.pyplot.show()
import parse
from functools import reduce
import numpy
import matplotlib.pyplot
morpheme_data = parse.get_morpheme_from_neko()
counter_by_word = list(parse.get_counter_from_words(parse.flatten(morpheme_data)).items())
sorted_by_number = sorted(list(map(lambda count: count[1], counter_by_word)), reverse = True)
matplotlib.pyplot.plot(sorted_by_number)
matplotlib.pyplot.xscale('log')
matplotlib.pyplot.yscale('log')
matplotlib.pyplot.show()
def test_parse_workout(self):
r = parse.parse_workout("""
- 10m easy warmup
- 5x:
- 1m gear up
- 1m gear even further up
- 2x:
- 15s insane
- 15s slow
- 10m cool down
""")
self.assertEquals(
r,
[{'d1': 'easy warmup', 'd2': '', 'dur': 600},
(('REPEAT', 5),
[{'d1': 'gear up', 'd2': '', 'dur': 60},
{'d1': 'gear even further up', 'd2': '', 'dur': 60},
(('REPEAT', 2),
[{'d1': 'insane', 'd2': '', 'dur': 15},
{'d1': 'slow', 'd2': '', 'dur': 15}])]),
{'d1': 'cool down', 'd2': '', 'dur': 600}])
flt = parse.flatten(r)
self.assertEquals(
flt,
[{'d1': 'easy warmup', 'd2': '', 'dur': 600, 'index': 0, 'level': 0},
{'d1': 'gear up', 'd2': '', 'dur': 60, 'index': 600, 'level': 1},
{'d1': 'gear even further up', 'd2': '', 'dur': 60, 'index': 660, 'level': 1},
{'d1': 'insane', 'd2': '', 'dur': 15, 'index': 720, 'level': 2},
{'d1': 'slow', 'd2': '', 'dur': 15, 'index': 735, 'level': 2},
{'d1': 'insane', 'd2': '', 'dur': 15, 'index': 750, 'level': 2},
{'d1': 'slow', 'd2': '', 'dur': 15, 'index': 765, 'level': 2},
{'d1': 'gear up', 'd2': '', 'dur': 60, 'index': 780, 'level': 1},
{'d1': 'gear even further up', 'd2': '', 'dur': 60, 'index': 840, 'level': 1},
{'d1': 'insane', 'd2': '', 'dur': 15, 'index': 900, 'level': 2},
{'d1': 'slow', 'd2': '', 'dur': 15, 'index': 915, 'level': 2},
{'d1': 'insane', 'd2': '', 'dur': 15, 'index': 930, 'level': 2},
{'d1': 'slow', 'd2': '', 'dur': 15, 'index': 945, 'level': 2},
{'d1': 'gear up', 'd2': '', 'dur': 60, 'index': 960, 'level': 1},
{'d1': 'gear even further up',
'd2': '',
'dur': 60,
'index': 1020,
'level': 1},
{'d1': 'insane', 'd2': '', 'dur': 15, 'index': 1080, 'level': 2},
{'d1': 'slow', 'd2': '', 'dur': 15, 'index': 1095, 'level': 2},
{'d1': 'insane', 'd2': '', 'dur': 15, 'index': 1110, 'level': 2},
{'d1': 'slow', 'd2': '', 'dur': 15, 'index': 1125, 'level': 2},
{'d1': 'gear up', 'd2': '', 'dur': 60, 'index': 1140, 'level': 1},
{'d1': 'gear even further up',
'd2': '',
'dur': 60,
'index': 1200,
'level': 1},
{'d1': 'insane', 'd2': '', 'dur': 15, 'index': 1260, 'level': 2},
{'d1': 'slow', 'd2': '', 'dur': 15, 'index': 1275, 'level': 2},
{'d1': 'insane', 'd2': '', 'dur': 15, 'index': 1290, 'level': 2},
{'d1': 'slow', 'd2': '', 'dur': 15, 'index': 1305, 'level': 2},
{'d1': 'gear up', 'd2': '', 'dur': 60, 'index': 1320, 'level': 1},
{'d1': 'gear even further up',
'd2': '',
'dur': 60,
'index': 1380,
'level': 1},
{'d1': 'insane', 'd2': '', 'dur': 15, 'index': 1440, 'level': 2},
{'d1': 'slow', 'd2': '', 'dur': 15, 'index': 1455, 'level': 2},
{'d1': 'insane', 'd2': '', 'dur': 15, 'index': 1470, 'level': 2},
{'d1': 'slow', 'd2': '', 'dur': 15, 'index': 1485, 'level': 2},
{'d1': 'cool down', 'd2': '', 'dur': 600, 'index': 1500, 'level': 0}])
import parse
from functools import reduce
import numpy
import matplotlib.pyplot
morpheme_data = parse.get_morpheme_from_neko()
counter_by_word = list(
parse.get_counter_from_words(parse.flatten(morpheme_data)).items())
sorted_by_number = sorted(list(map(lambda count: count[1], counter_by_word)),
reverse=True)
matplotlib.pyplot.plot(sorted_by_number)
matplotlib.pyplot.xscale('log')
matplotlib.pyplot.yscale('log')
matplotlib.pyplot.show()
import parse
def get_nouns_from_sentence(sentence):
result = []
nouns = ''
for i in range(0, len(sentence)):
if sentence[i]['品詞'] == '名詞':
nouns = nouns + sentence[i]['表層形']
elif len(nouns) != 0:
result.append(nouns)
nouns = ''
if len(nouns) != 0:
result.append(nouns)
return result
morpheme_data = parse.get_morpheme_from_neko()
print(parse.flatten(list(map(lambda sentence: get_nouns_from_sentence(sentence), morpheme_data))))
import parse morpheme_data = parse.get_morpheme_from_neko() print(parse.get_counter_from_words(parse.flatten(morpheme_data)))
import parse
from functools import reduce
import numpy
import matplotlib.pyplot
morpheme_data = parse.get_morpheme_from_neko()
counter = parse.get_counter_from_words(
parse.flatten(morpheme_data)).most_common(10)
xticks = numpy.array(
reduce(lambda result, count: result + [count[0][0]], counter, []))
x = numpy.array(list(range(1, 11)))
y = numpy.array(reduce(lambda result, count: result + [count[1]], counter, []))
matplotlib.pyplot.bar(x, y, align='center')
matplotlib.pyplot.xticks(x, xticks)
matplotlib.pyplot.show()
import parse
def get_noun_phrase_from_sentence(sentence):
result = []
for i in range(1, len(sentence) - 1):
word = sentence[i]
if word['表層形'] == 'の' and word['品詞'] == '助詞' and word[
'品詞細分類1'] == '連体化' and sentence[
i - 1]['品詞'] == '名詞' and sentence[i + 1]['品詞'] == '名詞':
result.append(sentence[i - 1]['表層形'] + 'の' +
sentence[i + 1]['表層形'])
return result
morpheme_data = parse.get_morpheme_from_neko()
print(
parse.flatten(
list(
map(lambda sentence: get_noun_phrase_from_sentence(sentence),
morpheme_data))))
