Put your vectors onto tensorboard!

vecvis.py

# encoding: utf-8
"""
Original credits
@author: BrikerMan 
@contact: [email protected]
@blog: https://eliyar.biz
@version: 1.0
@license: Apache Licence
@file: w2v_visualizer.py
@time: 2017/7/30 上午9:37
"""
import sys
import os
# from gensim.models import Word2Vec
import tensorflow as tf
import numpy as np
import pickle
from tensorflow.contrib.tensorboard.plugins import projector

def visualize(arr, labels, output_path):
    meta_file = "vis_metadata.tsv"
    placeholder = arr
    print placeholder.shape
    with open(os.path.join(output_path, meta_file), 'wb') as file_metadata:
        for i in xrange(arr.shape[0]):
            # temporary solution for https://github.com/tensorflow/tensorflow/issues/9094
            word = labels[i]
            if word == '':
                print("Emply Line, should replecaed by any thing else, or will cause a bug of tensorboard")
                file_metadata.write("{0}".format('<Empty Line>').encode('utf-8') + b'\n')
            else:
                file_metadata.write("{0}".format(word).encode('utf-8') + b'\n')

    # define the model without training
    sess = tf.InteractiveSession()

    embedding = tf.Variable(placeholder, trainable=False, name='vis_metadata')
    tf.global_variables_initializer().run()

    saver = tf.train.Saver()
    writer = tf.summary.FileWriter(output_path, sess.graph)

    # adding into projector
    config = projector.ProjectorConfig()
    embed = config.embeddings.add()
    embed.tensor_name = 'vis_metadata'
    embed.metadata_path = meta_file

    # Specify the width and height of a single thumbnail.
    projector.visualize_embeddings(writer, config)
    saver.save(sess, os.path.join(output_path, 'vec_metadata.ckpt'))
    print('Run `tensorboard --logdir={0}` to run visualize result on tensorboard'.format(output_path))

if __name__ == "__main__":
    """
    Just run `python w2v_visualizer.py word2vec.model visualize_result`
    """
    try:
        model_path = sys.argv[1]
        output_path  = sys.argv[2]
    except:
        print("Please provice model path and output path")
    model = np.load(model_path + '.npy')
    labels = pickle.load(open(model_path + '_labels.pkl', 'rb'))
    visualize(model, labels, output_path)

Prerequisites:
2D numpy array and pickle file of a list of corresponding labels

Usage:
python vecvis.py <numpy_array_filename_without_extension> <log_folder>

if 2D numpy array is called mymodel.npy, labels pickle file must be named mymodel_labels.pkl
Run python vecvis.py mymodel tensorboard_logs