11_kmeans_clustering_mnist.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:516: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint8 = np.dtype([(\"qint8\", np.int8, 1)])\n",
      "/Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:517: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint8 = np.dtype([(\"quint8\", np.uint8, 1)])\n",
      "/Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:518: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint16 = np.dtype([(\"qint16\", np.int16, 1)])\n",
      "/Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:519: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint16 = np.dtype([(\"quint16\", np.uint16, 1)])\n",
      "/Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:520: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint32 = np.dtype([(\"qint32\", np.int32, 1)])\n",
      "/Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  np_resource = np.dtype([(\"resource\", np.ubyte, 1)])\n",
      "/Users/i550012/.pyenv/versions/3.7.17/lib/python3.7/importlib/_bootstrap.py:219: RuntimeWarning: compiletime version 3.6 of module 'tensorflow.python.framework.fast_tensor_util' does not match runtime version 3.7\n",
      "  return f(*args, **kwds)\n",
      "/Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:541: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint8 = np.dtype([(\"qint8\", np.int8, 1)])\n",
      "/Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:542: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint8 = np.dtype([(\"quint8\", np.uint8, 1)])\n",
      "/Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:543: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint16 = np.dtype([(\"qint16\", np.int16, 1)])\n",
      "/Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:544: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_quint16 = np.dtype([(\"quint16\", np.uint16, 1)])\n",
      "/Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:545: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  _np_qint32 = np.dtype([(\"qint32\", np.int32, 1)])\n",
      "/Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorboard/compat/tensorflow_stub/dtypes.py:550: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.\n",
      "  np_resource = np.dtype([(\"resource\", np.ubyte, 1)])\n"
     ]
    }
   ],
   "source": [
    "from __future__ import absolute_import, division, print_function\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "%matplotlib inline\n",
    "import matplotlib \n",
    "import matplotlib.pyplot as plt\n",
    "from tensorflow.examples.tutorials.mnist import input_data\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /var/folders/nl/yn0d4tg1107g2jk38nj1hc740000gn/T/ipykernel_37720/1723679289.py:1: read_data_sets (from tensorflow.contrib.learn.python.learn.datasets.mnist) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Please use alternatives such as official/mnist/dataset.py from tensorflow/models.\n",
      "WARNING:tensorflow:From /Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorflow/contrib/learn/python/learn/datasets/mnist.py:260: maybe_download (from tensorflow.contrib.learn.python.learn.datasets.base) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Please write your own downloading logic.\n",
      "WARNING:tensorflow:From /Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorflow/contrib/learn/python/learn/datasets/mnist.py:262: extract_images (from tensorflow.contrib.learn.python.learn.datasets.mnist) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Please use tf.data to implement this functionality.\n",
      "Extracting MNIST_data/train-images-idx3-ubyte.gz\n",
      "WARNING:tensorflow:From /Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorflow/contrib/learn/python/learn/datasets/mnist.py:267: extract_labels (from tensorflow.contrib.learn.python.learn.datasets.mnist) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Please use tf.data to implement this functionality.\n",
      "Extracting MNIST_data/train-labels-idx1-ubyte.gz\n",
      "Extracting MNIST_data/t10k-images-idx3-ubyte.gz\n",
      "Extracting MNIST_data/t10k-labels-idx1-ubyte.gz\n",
      "WARNING:tensorflow:From /Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorflow/contrib/learn/python/learn/datasets/mnist.py:290: DataSet.__init__ (from tensorflow.contrib.learn.python.learn.datasets.mnist) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Please use alternatives such as official/mnist/dataset.py from tensorflow/models.\n"
     ]
    }
   ],
   "source": [
    "mnist = input_data.read_data_sets(\"MNIST_data\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def display_digit(digit):\n",
    "    plt.imshow(digit.reshape(28,28), cmap=\"Greys\", interpolation='nearest')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "tr_digits, tr_labels = mnist.train.next_batch(20000)\n",
    "tst_digits, tst_labels = mnist.test.next_batch(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display_digit(tr_digits[42])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def group_by_cluster(predicted_ids, original_entities):\n",
    "    cluster_groups=[[] for i in range(10)]\n",
    "    for i in range(len(predicted_ids)):\n",
    "        cluster_id = predicted_ids[i]\n",
    "        cluster_groups[cluster_id].append(original_entities[i])\n",
    "    return cluster_groups"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.contrib.learn.python.learn.estimators import kmeans\n",
    "from tensorflow.contrib.factorization.python.ops import clustering_ops"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def input_fn(digits):\n",
    "    input_t = tf.convert_to_tensor(digits, dtype=tf.float32)\n",
    "    return (input_t, None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /var/folders/nl/yn0d4tg1107g2jk38nj1hc740000gn/T/ipykernel_37720/2468173247.py:1: KMeansClustering.__init__ (from tensorflow.contrib.learn.python.learn.estimators.kmeans) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Please use tf.contrib.factorization.KMeansClustering instead of tf.contrib.learn.KMeansClustering. It has a similar interface, but uses the tf.estimator.Estimator API instead of tf.contrib.learn.Estimator.\n",
      "WARNING:tensorflow:From /Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorflow/contrib/learn/python/learn/estimators/estimator.py:1179: BaseEstimator.__init__ (from tensorflow.contrib.learn.python.learn.estimators.estimator) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Please replace uses of any Estimator from tf.contrib.learn with an Estimator from tf.estimator.*\n",
      "WARNING:tensorflow:From /Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorflow/contrib/learn/python/learn/estimators/estimator.py:427: RunConfig.__init__ (from tensorflow.contrib.learn.python.learn.estimators.run_config) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "When switching to tf.estimator.Estimator, use tf.estimator.RunConfig instead.\n",
      "INFO:tensorflow:Using default config.\n",
      "WARNING:tensorflow:Using temporary folder as model directory: /var/folders/nl/yn0d4tg1107g2jk38nj1hc740000gn/T/tmphtks48y_\n",
      "INFO:tensorflow:Using config: {'_task_type': None, '_task_id': 0, '_cluster_spec': <tensorflow.python.training.server_lib.ClusterSpec object at 0x143a7ecd0>, '_master': '', '_num_ps_replicas': 0, '_num_worker_replicas': 0, '_environment': 'local', '_is_chief': True, '_evaluation_master': '', '_train_distribute': None, '_eval_distribute': None, '_experimental_max_worker_delay_secs': None, '_device_fn': None, '_tf_config': gpu_options {\n",
      "  per_process_gpu_memory_fraction: 1.0\n",
      "}\n",
      ", '_tf_random_seed': None, '_save_summary_steps': 100, '_save_checkpoints_secs': 600, '_log_step_count_steps': 100, '_protocol': None, '_session_config': None, '_save_checkpoints_steps': None, '_keep_checkpoint_max': 5, '_keep_checkpoint_every_n_hours': 10000, '_model_dir': '/var/folders/nl/yn0d4tg1107g2jk38nj1hc740000gn/T/tmphtks48y_'}\n",
      "WARNING:tensorflow:From /Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorflow/python/training/training_util.py:236: Variable.initialized_value (from tensorflow.python.ops.variables) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use Variable.read_value. Variables in 2.X are initialized automatically both in eager and graph (inside tf.defun) contexts.\n",
      "WARNING:tensorflow:From /Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorflow/contrib/learn/python/learn/estimators/kmeans.py:151: ModelFnOps.__new__ (from tensorflow.contrib.learn.python.learn.estimators.model_fn) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "When switching to tf.estimator.Estimator, use tf.estimator.EstimatorSpec. You can use the `estimator_spec` method to create an equivalent one.\n",
      "INFO:tensorflow:Create CheckpointSaverHook.\n",
      "WARNING:tensorflow:From /Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorflow/python/ops/array_ops.py:1354: add_dispatch_support.<locals>.wrapper (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.where in 2.0, which has the same broadcast rule as np.where\n",
      "INFO:tensorflow:Graph was finalized.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2024-04-11 05:33:10.616436: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Running local_init_op.\n",
      "INFO:tensorflow:Done running local_init_op.\n",
      "INFO:tensorflow:Saving checkpoints for 0 into /var/folders/nl/yn0d4tg1107g2jk38nj1hc740000gn/T/tmphtks48y_/model.ckpt.\n",
      "INFO:tensorflow:loss = 1384859.2, step = 1\n",
      "INFO:tensorflow:global_step/sec: 15.6549\n",
      "INFO:tensorflow:loss = 794629.0, step = 101 (6.388 sec)\n",
      "WARNING:tensorflow:It seems that global step (tf.train.get_global_step) has not been increased. Current value (could be stable): 122 vs previous value: 122. You could increase the global step by passing tf.train.get_global_step() to Optimizer.apply_gradients or Optimizer.minimize.\n",
      "WARNING:tensorflow:It seems that global step (tf.train.get_global_step) has not been increased. Current value (could be stable): 141 vs previous value: 141. You could increase the global step by passing tf.train.get_global_step() to Optimizer.apply_gradients or Optimizer.minimize.\n",
      "WARNING:tensorflow:It seems that global step (tf.train.get_global_step) has not been increased. Current value (could be stable): 143 vs previous value: 143. You could increase the global step by passing tf.train.get_global_step() to Optimizer.apply_gradients or Optimizer.minimize.\n",
      "WARNING:tensorflow:It seems that global step (tf.train.get_global_step) has not been increased. Current value (could be stable): 146 vs previous value: 146. You could increase the global step by passing tf.train.get_global_step() to Optimizer.apply_gradients or Optimizer.minimize.\n",
      "WARNING:tensorflow:It seems that global step (tf.train.get_global_step) has not been increased. Current value (could be stable): 169 vs previous value: 169. You could increase the global step by passing tf.train.get_global_step() to Optimizer.apply_gradients or Optimizer.minimize.\n",
      "INFO:tensorflow:global_step/sec: 66.0985\n",
      "INFO:tensorflow:loss = 793683.44, step = 201 (1.513 sec)\n",
      "INFO:tensorflow:global_step/sec: 64.7025\n",
      "INFO:tensorflow:loss = 793284.25, step = 301 (1.546 sec)\n",
      "INFO:tensorflow:global_step/sec: 66.0295\n",
      "INFO:tensorflow:loss = 793052.4, step = 401 (1.514 sec)\n",
      "INFO:tensorflow:global_step/sec: 63.3858\n",
      "INFO:tensorflow:loss = 792901.94, step = 501 (1.578 sec)\n",
      "INFO:tensorflow:global_step/sec: 64.1172\n",
      "INFO:tensorflow:loss = 792792.44, step = 601 (1.560 sec)\n",
      "INFO:tensorflow:global_step/sec: 63.55\n",
      "INFO:tensorflow:loss = 792708.7, step = 701 (1.573 sec)\n",
      "INFO:tensorflow:global_step/sec: 64.7138\n",
      "INFO:tensorflow:loss = 792643.25, step = 801 (1.545 sec)\n",
      "INFO:tensorflow:global_step/sec: 66.2656\n",
      "INFO:tensorflow:loss = 792588.0, step = 901 (1.509 sec)\n",
      "INFO:tensorflow:Saving checkpoints for 1000 into /var/folders/nl/yn0d4tg1107g2jk38nj1hc740000gn/T/tmphtks48y_/model.ckpt.\n",
      "INFO:tensorflow:Loss for final step: 792539.4.\n"
     ]
    }
   ],
   "source": [
    "k_mean_estimator = kmeans.KMeansClustering(num_clusters=10)\n",
    "fit = k_mean_estimator.fit(input_fn=lambda: input_fn(tr_digits), steps=1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /var/folders/nl/yn0d4tg1107g2jk38nj1hc740000gn/T/ipykernel_37720/3167643738.py:1: KMeansClustering.clusters (from tensorflow.contrib.learn.python.learn.estimators.kmeans) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Please use tf.contrib.factorization.KMeansClustering instead of tf.contrib.learn.KMeansClustering. It has a similar interface, but uses the tf.estimator.Estimator API instead of tf.contrib.learn.Estimator.\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "clusters = k_mean_estimator.clusters()\n",
    "for i in range(10):\n",
    "    plt.subplot(2,5,i+1)\n",
    "    display_digit(clusters[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "cluster_labels = [0,8,3,0,1,2,9,9,3,6]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in range(5):\n",
    "    plt.subplot(1,5,i+1)\n",
    "    display_digit(tst_digits[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /var/folders/nl/yn0d4tg1107g2jk38nj1hc740000gn/T/ipykernel_37720/887728804.py:1: calling BaseEstimator.predict (from tensorflow.contrib.learn.python.learn.estimators.estimator) with as_iterable is deprecated and will be removed after 2016-12-01.\n",
      "Instructions for updating:\n",
      "Estimator is decoupled from Scikit Learn interface by moving into\n",
      "separate class SKCompat. Arguments x, y and batch_size are only\n",
      "available in the SKCompat class, Estimator will only accept input_fn.\n",
      "Example conversion:\n",
      "  est = Estimator(...) -> est = SKCompat(Estimator(...))\n",
      "INFO:tensorflow:Graph was finalized.\n",
      "WARNING:tensorflow:From /Users/i550012/virtualenv/tensorflow1/lib/python3.7/site-packages/tensorflow/python/training/saver.py:1276: checkpoint_exists (from tensorflow.python.training.checkpoint_management) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use standard file APIs to check for files with this prefix.\n",
      "INFO:tensorflow:Restoring parameters from /var/folders/nl/yn0d4tg1107g2jk38nj1hc740000gn/T/tmphtks48y_/model.ckpt-1000\n",
      "INFO:tensorflow:Running local_init_op.\n",
      "INFO:tensorflow:Done running local_init_op.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'all_scores': array([[ 68.50983 ,  89.67316 ,  42.999874,  71.33064 ,  62.11988 ,\n",
       "          62.97288 ,  51.332886, 100.29912 ,  83.107574,  69.86878 ],\n",
       "        [ 49.377396,  54.341824,  65.51222 ,  74.29921 ,  72.6859  ,\n",
       "          68.370834,  61.9375  ,  90.05995 ,  59.434105,  37.143543],\n",
       "        [ 71.47589 ,  76.05458 ,  74.931564,  33.64155 ,  63.00096 ,\n",
       "          47.96951 ,  66.993675,  88.94453 ,  56.0177  ,  57.97996 ],\n",
       "        [ 56.91543 ,  38.473225,  57.16477 ,  59.37478 ,  49.522186,\n",
       "          77.063614,  62.54617 ,  68.00519 ,  56.876644,  60.951614],\n",
       "        [ 57.2503  ,  67.48408 ,  50.248756,  55.329388,  53.833214,\n",
       "          43.90102 ,  47.71221 ,  90.02969 ,  47.205994,  27.558659]],\n",
       "       dtype=float32),\n",
       " 'cluster_idx': array([2, 9, 3, 1, 9])}"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predict = k_mean_estimator.predict(input_fn=lambda: input_fn(tst_digits), as_iterable=False)\n",
    "predict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[3, 6, 0, 8, 6]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print([cluster_labels[i] for i in predict['cluster_idx']])\n",
    "for i in range(5):\n",
    "    plt.subplot(1,5, i+1)\n",
    "    display_digit(tst_digits[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([7, 0, 0, 2, 1], dtype=uint8)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tr_labels[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Graph was finalized.\n",
      "INFO:tensorflow:Restoring parameters from /var/folders/nl/yn0d4tg1107g2jk38nj1hc740000gn/T/tmphtks48y_/model.ckpt-1000\n",
      "INFO:tensorflow:Running local_init_op.\n",
      "INFO:tensorflow:Done running local_init_op.\n"
     ]
    }
   ],
   "source": [
    "predict_train = k_mean_estimator.predict(input_fn=lambda: input_fn(tr_digits), as_iterable=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def display_accuracy(cluster_labels, cluster_idx, actual_labels):\n",
    "    predict_labels = [cluster_labels[i] in cluster_idx]\n",
    "    num_accurate_predictions = (list(predict_labels==actual_labels)).count(True)\n",
    "    print('nAccuratePredictions: ',num_accurate_predictions)\n",
    "    pctAccuracy=float(num_accurate_predictions)/ float(len(actual_labels))\n",
    "    print('%AccuracyPredictions: ', pctAccuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "nAccuratePredictions:  2216\n",
      "%AccuracyPredictions:  0.1108\n"
     ]
    }
   ],
   "source": [
    "display_accuracy(cluster_labels, predict_train['cluster_idx'], tr_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def display_cluster_mean_std_dev(tr_cl_id, tr_labels):\n",
    "    cluster_group_labels = group_by_cluster(tr_cl_id,tr_labels)\n",
    "    mean = [[] for i in range(10)]\n",
    "    std = [[] for i in range(10)]\n",
    "    for i in range(10):\n",
    "        mean[i] = np.mean(cluster_group_labels[i])\n",
    "        std[i] = np.std(cluster_group_labels[i])\n",
    "    print('mean: ', mean)\n",
    "    print('std: ', std)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mean:  [2.287425149700599, 5.66507747318236, 4.456910569105691, 6.928955866523143, 6.3469026548672565, 6.690104166666667, 3.7536865101037686, 0.361328125, 4.0131736526946105, 2.337618403247632]\n",
      "std:  [1.1332048580841225, 1.248067287715437, 2.2422192343709715, 1.8089532604968575, 2.287483093277052, 2.4123285754258212, 1.9633330002115574, 1.474713710436928, 2.2826749435669664, 2.407915512543506]\n"
     ]
    }
   ],
   "source": [
    "display_cluster_mean_std_dev(predict_train['cluster_idx'], tr_labels )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy import stats\n",
    "def get_cluster_mode(trining_cluster_ids, training_labels):\n",
    "    cluster_group = group_by_cluster(trining_cluster_ids, training_labels)\n",
    "    mode = [[] for i in range(10)]\n",
    "    for i in range(10):\n",
    "        mode[i] = stats.mode(cluster_group[i])[0][0]\n",
    "    return mode"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0, 8, 3, 0, 1, 2, 9, 9, 3, 6]\n",
      "[2, 6, 3, 7, 4, 8, 3, 0, 5, 1]\n"
     ]
    }
   ],
   "source": [
    "mode_labels=get_cluster_mode(predict_train['cluster_idx'], tr_labels)\n",
    "print(cluster_labels)\n",
    "print(mode_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "nAccuratePredictions:  2216\n",
      "%AccuracyPredictions:  0.1108\n"
     ]
    }
   ],
   "source": [
    "display_accuracy(mode_labels, predict_train['cluster_idx'], tr_labels)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "tensorflow1",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.17"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}