dim-6.py

import pandas as pd
import umap
import sys
import numpy as np
import os


def load_data(file_path, dimension=3):
    # If file path extension is numpy extension, load it as numpy array
    if file_path.endswith('.npy'):
        data = np.load(file_path)

    # If file path is csv file load with pandas
    if file_path.endswith('.csv'):
        data = pd.read_csv(file_path)
        data = data.dropna()
        print('Data has been read!')
    return data

def dimension_reduction(data):
    
    reducer = umap.UMAP(n_neighbors=10, n_components=6, min_dist=0.1, metric='sokalsneath', spread=18)
    embedding = reducer.fit_transform(data)
    print('Embeddings are created!')
    return embedding

def write_csv(data, embedding, output_path):
    dim6_df = pd.DataFrame(embedding, columns=['v1', 'v2', 'v3', 'v4', 'v5', 'v6'])
    # dim6_df['label'] = data['words']
    print('Dataframe is saving...')
    dim6_df.to_csv(output_path, index=False)


# data = pd.read_csv(FILE_PATH)
#fit = umap.UMAP(n_neighbors=10, min_dist=0.1, n_components=6, metric='cosine', spread=10)
# u = fit.fit_transform(data.iloc[:, :-1])
# dim_red = pd.DataFrame(u, columns=['v1', 'v2', 'v3', 'v4', 'v5', 'v6'])



# if __name__ == '__main__':
#     for file in os.listdir('./rumi-text-umaps'):
#         if file.endswith('.csv'):
#             data = load_data('./rumi-text-umaps/' + file)
#             embedding = dimension_reduction(data)
#             write_csv(data, embedding, file.replace('v2', 'v3'))

if __name__ == '__main__':
    input_file = sys.argv[1]
    output_file = sys.argv[2]

    data = load_data(input_file)
    embedding = dimension_reduction(data)
    write_csv(data, embedding, output_file)