# -- Code Cell --
import pandas as pd
from gensim.models import KeyedVectors

train_data = pd.read_csv('train_data.csv')
test_data = pd.read_csv('test_data.csv')

# -- Code Cell --
# Path to your .vec files
model_path_en = 'fasttext/en.vec'
model_path_ro = 'fasttext/ro.vec'

model_en = KeyedVectors.load_word2vec_format(model_path_en, binary=False, limit=50000)
model_ro = KeyedVectors.load_word2vec_format(model_path_ro, binary=False, limit=50000)

# -- Code Cell --
# Place the answer for subtask one here
subtask_one_ans = 0
# Replace this with the list of correctly arranged translated words for subtask two here
subtask_two_ans = test_data['Similar']

# -- Code Cell --
# Submission generation code
task_one_df = pd.DataFrame([{
    'subtaskID': 1,
    'datapointID': 0,
    'answer': subtask_one_ans
}])
task_two_df = pd.DataFrame({
    'subtaskID': [2]*len(subtask_two_ans),
    'datapointID': range(1, len(subtask_two_ans)+1),
    'answer': subtask_two_ans
})
submission = pd.concat([task_one_df, task_two_df], axis=0)
submission.to_csv('submission.csv', index=False)