#!/usr/bin/env python
# coding: utf-8

# # 1. Skip-gram with naiive softmax 

# I recommend you take a look at these material first.

# * http://web.stanford.edu/class/cs224n/lectures/cs224n-2017-lecture2.pdf
# * https://arxiv.org/abs/1301.3781
# * http://mccormickml.com/2016/04/19/word2vec-tutorial-the-skip-gram-model/

# In[1]:


import torch
import torch.nn as nn
from torch.autograd import Variable
import torch.optim as optim
import torch.nn.functional as F
import nltk
import random
import numpy as np
from collections import Counter
flatten = lambda l: [item for sublist in l for item in sublist]
random.seed(1024)


# In[2]:


print(torch.__version__)
print(nltk.__version__)


# In[3]:


USE_CUDA = torch.cuda.is_available()
gpus = [0]
torch.cuda.set_device(gpus[0])

FloatTensor = torch.cuda.FloatTensor if USE_CUDA else torch.FloatTensor
LongTensor = torch.cuda.LongTensor if USE_CUDA else torch.LongTensor
ByteTensor = torch.cuda.ByteTensor if USE_CUDA else torch.ByteTensor


# In[4]:


def getBatch(batch_size, train_data):
    random.shuffle(train_data)
    sindex = 0
    eindex = batch_size
    while eindex < len(train_data):
        batch = train_data[sindex: eindex]
        temp = eindex
        eindex = eindex + batch_size
        sindex = temp
        yield batch
    
    if eindex >= len(train_data):
        batch = train_data[sindex:]
        yield batch


# In[5]:


def prepare_sequence(seq, word2index):
    idxs = list(map(lambda w: word2index[w] if word2index.get(w) is not None else word2index["<UNK>"], seq))
    return Variable(LongTensor(idxs))

def prepare_word(word, word2index):
    return Variable(LongTensor([word2index[word]]) if word2index.get(word) is not None else LongTensor([word2index["<UNK>"]]))


# ## Data load and Preprocessing 

# ### Load corpus : Gutenberg corpus

# If you don't have gutenberg corpus, you can download it first using nltk.download()

# In[6]:


nltk.corpus.gutenberg.fileids()


# In[7]:


corpus = list(nltk.corpus.gutenberg.sents('melville-moby_dick.txt'))[:100] # sampling sentences for test
corpus = [[word.lower() for word in sent] for sent in corpus]


# ### Extract Stopwords from unigram distribution's tails

# In[8]:


word_count = Counter(flatten(corpus))
border = int(len(word_count) * 0.01) 


# In[9]:


stopwords = word_count.most_common()[:border] + list(reversed(word_count.most_common()))[:border]


# In[10]:


stopwords = [s[0] for s in stopwords]


# In[11]:


stopwords


# ### Build vocab

# In[12]:


vocab = list(set(flatten(corpus)) - set(stopwords))
vocab.append('<UNK>')


# In[13]:


print(len(set(flatten(corpus))), len(vocab))


# In[14]:


word2index = {'<UNK>' : 0} 

for vo in vocab:
    if word2index.get(vo) is None:
        word2index[vo] = len(word2index)

index2word = {v:k for k, v in word2index.items()} 


# ### Prepare train data 

# window data example

# <img src="../images/01.skipgram-prepare-data.png">
# <center>borrowed image from http://mccormickml.com/2016/04/19/word2vec-tutorial-the-skip-gram-model/</center>

# In[15]:


WINDOW_SIZE = 3
windows = flatten([list(nltk.ngrams(['<DUMMY>'] * WINDOW_SIZE + c + ['<DUMMY>'] * WINDOW_SIZE, WINDOW_SIZE * 2 + 1)) for c in corpus])


# In[16]:


windows[0]


# In[17]:


train_data = []

for window in windows:
    for i in range(WINDOW_SIZE * 2 + 1):
        if i == WINDOW_SIZE or window[i] == '<DUMMY>': 
            continue
        train_data.append((window[WINDOW_SIZE], window[i]))

print(train_data[:WINDOW_SIZE * 2])


# In[18]:


X_p = []
y_p = []


# In[19]:


train_data[0]


# In[20]:


for tr in train_data:
    X_p.append(prepare_word(tr[0], word2index).view(1, -1))
    y_p.append(prepare_word(tr[1], word2index).view(1, -1))


# In[21]:


train_data = list(zip(X_p, y_p))


# In[22]:


len(train_data)


# ## Modeling

# <img src="../images/01.skipgram-objective.png">
# <center>borrowed image from http://web.stanford.edu/class/cs224n/lectures/cs224n-2017-lecture2.pdf</center>

# In[59]:


class Skipgram(nn.Module):
    
    def __init__(self, vocab_size, projection_dim):
        super(Skipgram,self).__init__()
        self.embedding_v = nn.Embedding(vocab_size, projection_dim)
        self.embedding_u = nn.Embedding(vocab_size, projection_dim)

        self.embedding_v.weight.data.uniform_(-1, 1) # init
        self.embedding_u.weight.data.uniform_(0, 0) # init
        #self.out = nn.Linear(projection_dim,vocab_size)
    def forward(self, center_words,target_words, outer_words):
        center_embeds = self.embedding_v(center_words) # B x 1 x D
        target_embeds = self.embedding_u(target_words) # B x 1 x D
        outer_embeds = self.embedding_u(outer_words) # B x V x D
        
        scores = target_embeds.bmm(center_embeds.transpose(1, 2)).squeeze(2) # Bx1xD * BxDx1 => Bx1
        norm_scores = outer_embeds.bmm(center_embeds.transpose(1, 2)).squeeze(2) # BxVxD * BxDx1 => BxV
        
        nll = -torch.mean(torch.log(torch.exp(scores)/torch.sum(torch.exp(norm_scores), 1).unsqueeze(1))) # log-softmax
        
        return nll # negative log likelihood
    
    def prediction(self, inputs):
        embeds = self.embedding_v(inputs)
        
        return embeds 


# ## Train  

# In[60]:


EMBEDDING_SIZE = 30
BATCH_SIZE = 256
EPOCH = 100


# In[61]:


losses = []
model = Skipgram(len(word2index), EMBEDDING_SIZE)
if USE_CUDA:
    model = model.cuda()
optimizer = optim.Adam(model.parameters(), lr=0.01)


# In[62]:


for epoch in range(EPOCH):
    for i, batch in enumerate(getBatch(BATCH_SIZE, train_data)):
        
        inputs, targets = zip(*batch)
        
        inputs = torch.cat(inputs) # B x 1
        targets = torch.cat(targets) # B x 1
        vocabs = prepare_sequence(list(vocab), word2index).expand(inputs.size(0), len(vocab))  # B x V
        model.zero_grad()

        loss = model(inputs, targets, vocabs)
        
        loss.backward()
        optimizer.step()
   
        losses.append(loss.data.tolist()[0])

    if epoch % 10 == 0:
        print("Epoch : %d, mean_loss : %.02f" % (epoch,np.mean(losses)))
        losses = []


# ## Test

# In[63]:


def word_similarity(target, vocab):
    if USE_CUDA:
        target_V = model.prediction(prepare_word(target, word2index))
    else:
        target_V = model.prediction(prepare_word(target, word2index))
    similarities = []
    for i in range(len(vocab)):
        if vocab[i] == target: continue
        
        if USE_CUDA:
            vector = model.prediction(prepare_word(list(vocab)[i], word2index))
        else:
            vector = model.prediction(prepare_word(list(vocab)[i], word2index))
        cosine_sim = F.cosine_similarity(target_V, vector).data.tolist()[0] 
        similarities.append([vocab[i], cosine_sim])
    return sorted(similarities, key=lambda x: x[1], reverse=True)[:10] # sort by similarity


# In[64]:


test = random.choice(list(vocab))
test


# In[65]:


word_similarity(test, vocab)


# In[ ]: