def test_single_timestep(self):
chunks, next_chars = chunk_names(self.test_df, 'name', 1)
self.assertEqual(['d', 'a', 'r', 't', 'h', ' ', 'v', 'a', 'd', 'e', 'r'], chunks)
self.assertEqual(['a', 'r', 't', 'h', ' ', 'v', 'a', 'd', 'e', 'r', '\n'], next_chars)
def test_three_timesteps(self):
chunks, next_chars = chunk_names(self.test_df, 'name', 3)
# 'darth vader'
self.assertEqual(['dar', 'art', 'rth', 'th ', 'h v', ' va', 'vad', 'ade', 'der'], chunks)
self.assertEqual(['t', 'h', ' ', 'v', 'a', 'd', 'e', 'r', '\n'], next_chars)
vocab_size = len(chars)
vocab_size
# In[178]:
timesteps = 1
# In[179]:
chunks, next_char = chunk_names(combined, 'name', timesteps)
# In[180]:
X = numpy.zeros((len(chunks), timesteps, vocab_size))
y = numpy.zeros((len(chunks), vocab_size))
# In[181]:
for i, chunk in enumerate(chunks):
for t, token in enumerate(chunk):
index = token_to_index[token]