examples/Stack RNN for formal sequence modelling.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "\n",
    "#theano imports\n",
    "import lasagne\n",
    "import theano\n",
    "import theano.tensor as T\n",
    "import sys\n",
    "sys.setrecursionlimit(100000)\n",
    "floatX = theano.config.floatX\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Stack-augmented RNN\n",
    "![caption](https://usercontent1.hubstatic.com/6172838_f260.jpg)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This notebook reproduces experiment from article http://arxiv.org/abs/1503.01007"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Task and architecture\n",
    "\n",
    "__The task__ is to train an RNN to generate sequences from expression `|`$ a^n b^m c^{n+m} $\n",
    " * n and m are integers\n",
    " * Performance metrics:\n",
    "  * Correct format - `|`, some a, some b, some c\n",
    "     * ||aaacbbcba would be an example of incorrect one\n",
    "  * Number of C letters to be as close to sum of A band B letters as possible\n",
    "     * Ideally, #c = #a + #b\n",
    "     \n",
    "\n",
    "Model we use:\n",
    " * Stack-augmented RNN\n",
    " \n",
    "Trained as a language model (predicting next symbols given preceding history)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Sequence generator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def generate_sequence(batch_size = 10,crop_length = 100 ):\n",
    "    \"\"\"\n",
    "    Generates sequence from pattern [0  1*n 2*m 3*(n+m)]\n",
    "    \"\"\"\n",
    "    sequences=[]\n",
    "    for i in range(batch_size):\n",
    "        seq = [0]\n",
    "        while len(seq) < crop_length:\n",
    "            n,m = np.random.randint(1,10,2)\n",
    "        \n",
    "            seq += [0]+ [1]*n+[2]*m+[3]*(n+m)\n",
    "        seq = seq[:crop_length]\n",
    "        sequences.append(seq)\n",
    "    return np.array(sequences,dtype='int32')\n",
    "\n",
    "alphabet = np.array(list('|abc'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#quality evaluation\n",
    "from operator import add\n",
    "import re\n",
    "def get_metrics(gen_sequences,alphabet):    \n",
    "    strings = map(lambda v:reduce(add,v), map(alphabet.__getitem__,gen_sequences))\n",
    "    \n",
    "    #at least one complete string\n",
    "    strings = filter(lambda v: '|' in v,strings)\n",
    "    \n",
    "    #cut off last unfinished string, if any\n",
    "    \n",
    "    strings = map(lambda v: v[:-1-v[::-1].index('|')],strings)\n",
    "    \n",
    "    all_strings = '|'.join(strings)\n",
    "    \n",
    "    \n",
    "    \n",
    "    seqs = filter(len,all_strings.split('|'))\n",
    "    \n",
    "    matches = map(lambda seq:re.match(r\"^a+b+c+\",seq),seqs)\n",
    "    \n",
    "    is_correct = map( lambda seq,m: m is not None and m.pos==0 and m.endpos == len(seq), \n",
    "                        seqs, matches)\n",
    "    \n",
    "    \n",
    "    \n",
    "    \n",
    "    correct_seqs = np.array(seqs)[np.array(is_correct,dtype='bool')]\n",
    "    \n",
    "    seqs_error = map(lambda s: s.count('a')+ s.count('b') - s.count('c'),correct_seqs)\n",
    "    \n",
    "    return np.mean(is_correct), np.mean(np.abs(seqs_error))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### samples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 0 ns, sys: 0 ns, total: 0 ns\n",
      "Wall time: 937 µs\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "['||aaaaabbbbccccccccc|aaaaaaaabbbccccccccccc|aabbcccc|aaaaaaaaabbbbbbbbccccccccccccccccc|aaaaaabbbbcc',\n",
       " '||aaaabbbbbbbccccccccccc|abbbbccccc|aaaabbbbbbbbcccccccccccc|aaaabbbbbccccccccc|aabbbbcccccc|aabbbbb',\n",
       " '||aaaaabbbbbbbbccccccccccccc|aaaabbbccccccc|aaabbbbccccccc|aaabbbbbcccccccc|aaabbbbccccccc|aaaabbbbc',\n",
       " '||aaabbbcccccc|abbbbbbccccccc|aaaaabbbbbbbcccccccccccc|aaaaaaabbbbbbbcccccccccccccc|aaabbbcccccc|aaa',\n",
       " '||aaaabccccc|aaaaaaaaabbbbbcccccccccccccc|aaaaaaaaabbbbbbbcccccccccccccccc|aaaabccccc|aaaaaaaaabbbbb']"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "map(''.join,map(alphabet.__getitem__,generate_sequence(5,100)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Constants"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# length of training/generation sequence\n",
    "SEQ_LENGTH = 50\n",
    "\n",
    "# Sequences in a minibatch\n",
    "BATCH_SIZE = 100\n",
    "\n",
    "# Total epochs\n",
    "N_EPOCHS = 5000\n",
    "\n",
    "#how often to print metrics\n",
    "REPORT_RATE = 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Reference sequences for training\n",
    "sequences_batch = T.matrix(dtype=\"int32\",name=\"reference_sequences\")\n",
    "\n",
    "#Batch size (theano expression)\n",
    "batch_size = sequences_batch.shape[0]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Define RNN\n",
    "\n",
    "A \"simple\" scheme of what was there in the paper, followed exactly by the code below\n",
    "\n",
    "* Time goes from left to right\n",
    "* Layers start at the bottom and go up\n",
    "* memory states are green, memory updates are purple\n",
    "* inputs and outputs are blue traingles\n",
    "* transformations are in pale yellow\n",
    "\n",
    "![canvas](http://s32.postimg.org/vefrk7vqt/stack_gru.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import lasagne\n",
    "from lasagne.layers import DenseLayer, ConcatLayer, InputLayer, EmbeddingLayer\n",
    "\n",
    "import agentnet\n",
    "from agentnet.resolver import ProbabilisticResolver\n",
    "\n",
    "from agentnet.memory import GRUMemoryLayer, StackAugmentation\n",
    "\n",
    "\n",
    "from agentnet.agent import Recurrence\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "#input letter goes here\n",
    "output_shape = (None,)\n",
    "observation_layer = InputLayer(output_shape,name=\"obs_input\")\n",
    "\n",
    "\n",
    "# embedding the input layer\n",
    "n_tokens = len(alphabet)\n",
    "obs_embedding = EmbeddingLayer(observation_layer,\n",
    "                                              input_size=n_tokens,\n",
    "                                              output_size=n_tokens,\n",
    "                                              name = \"input_embedding\")\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "#GRU n units\n",
    "n_hid_1 = 32\n",
    "\n",
    "#previous GRU state goes here\n",
    "prev_gru_layer = InputLayer((None,n_hid_1),name=\"prev_rnn_state\")\n",
    "\n",
    "\n",
    "# previous stack goes here\n",
    "stack_width = 3\n",
    "stack_depth = 50\n",
    "prev_stack_layer = InputLayer((None,stack_depth,stack_width))\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# Stack controls - push, pop and no-op\n",
    "stack_controls_layer = lasagne.layers.DenseLayer(prev_gru_layer,3,\n",
    "                                                nonlinearity= lasagne.nonlinearities.softmax,\n",
    "                                                name = \"stack controls\")\n",
    "\n",
    "# stack input\n",
    "stack_input_layer = lasagne.layers.DenseLayer(prev_gru_layer,stack_width,\n",
    "                                             nonlinearity = lasagne.nonlinearities.tanh,\n",
    "                                             name = \"stack input\")\n",
    "    \n",
    "    \n",
    "#new stack state  \n",
    "next_stack = StackAugmentation(stack_input_layer,\n",
    "                              prev_stack_layer,\n",
    "                              stack_controls_layer)\n",
    "\n",
    "\n",
    "#stack top (used for RNN)\n",
    "stack_top = lasagne.layers.SliceLayer(next_stack,0,1)\n",
    "\n",
    "\n",
    "\n",
    "#new GRU state\n",
    "gru = GRUMemoryLayer(n_hid_1,\n",
    "                     ConcatLayer([obs_embedding,stack_top]),\n",
    "                     prev_gru_layer)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "##Outputs\n",
    "\n",
    "# next letter probabilities\n",
    "\n",
    "probability_layer = lasagne.layers.DenseLayer(gru,\n",
    "                                         num_units = n_tokens,\n",
    "                                         nonlinearity=  lasagne.nonlinearities.softmax,\n",
    "                                         name=\"policy_original\")\n",
    "\n",
    "#resolver picks a particular letter in generation mode\n",
    "\n",
    "resolver = ProbabilisticResolver(probability_layer,\n",
    "                                assume_normalized=True,\n",
    "                                name=\"resolver\")\n",
    "\n",
    "\n",
    "#verify that letter shape matches\n",
    "assert tuple(lasagne.layers.get_output_shape(resolver)) == tuple(output_shape)\n",
    "\n",
    "\n",
    "#define a dictionary that maps agent's \"next memory\" output to its previous memory input\n",
    "from collections import OrderedDict\n",
    "memory_dict = OrderedDict([\n",
    "            (gru,prev_gru_layer),\n",
    "            (next_stack, prev_stack_layer),\n",
    "    ])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training mode\n",
    "\n",
    "* Here we define an in-training version of out generator.\n",
    "* To speed up training, we will feed agent with a reference sequences.\n",
    "* So, agent pretends to generate letters, but actually is fed with the reference ones.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "#define an input layer that stores sequences\n",
    "sequences_input_layer = InputLayer((None,SEQ_LENGTH),\n",
    "                                   input_var= sequences_batch,\n",
    "                                   name = \"reference sequences\"\n",
    "                                  )\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "#and another one that only pretends to, while actually uses the reference letters\n",
    "# we will use it for training.\n",
    "generator_passive = Recurrence(\n",
    "    \n",
    "    #we use out previously defined dictionary to update recurrent network parameters\n",
    "    state_variables = memory_dict,\n",
    "    \n",
    "    #we feed in reference sequence into \"prev letter\" input layer, tick by tick along the axis=1.\n",
    "    input_sequences = {observation_layer:sequences_input_layer},\n",
    "    \n",
    "    #we track agent would-be actions and probabilities \n",
    "    tracked_outputs = [resolver,probability_layer],\n",
    "    \n",
    "    n_steps = SEQ_LENGTH,\n",
    "    \n",
    "    #finally, we define an optional batch size param \n",
    "    #(if omitted, it will be inferred from inputs or initial value providers if there are any)\n",
    "    batch_size = batch_size,\n",
    "    \n",
    "    unroll_scan=False,\n",
    "    \n",
    "    )\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Unroll the recurrence\n",
    "\n",
    "* returns sessions of shape [batch_size, sequence_length, variable_size] for each variable\n",
    "\n",
    "* this is still happening inside a lasagne networks\n",
    "  * which allows you to, for example, create a recurrence inside another recurrence"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#get lasagne layers for sequences\n",
    "\n",
    "#sequence of predicted probabilities\n",
    "probas_seq_layer = generator_passive[probability_layer]\n",
    "\n",
    "# gru state sequence (can extract stack likewise)\n",
    "gru_seq_layer = generator_passive[gru]\n",
    "\n",
    "\n",
    "\n",
    "# for this task we only need probabilities :)\n",
    "# probas_seq_layer\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/hedgedir/agentnet/agentnet/agent/recurrence.py:595: UserWarning: Warning: recurrent loop without unroll_scan got nonempty random state updates list. That happened because there is some source of randomness (e.g. dropout) inside recurrent step graph. To compile such graph, one must either call .get_automatic_updates() right after .get_output and pass these updates to a function, or use no_defalt_updates=True when compiling theano.function.\n",
      "  warn(\"Warning: recurrent loop without unroll_scan got nonempty random state updates list. That happened\"\n"
     ]
    }
   ],
   "source": [
    "probas_seq = lasagne.layers.get_output(probas_seq_layer)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The rest is a regular lasagne"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[input_embedding.W,\n",
       " stack input.W,\n",
       " stack input.b,\n",
       " stack controls.W,\n",
       " stack controls.b,\n",
       " YetAnotherGRUMemoryLayer.W_in_to_updategate,\n",
       " YetAnotherGRUMemoryLayer.W_hid_to_updategate,\n",
       " YetAnotherGRUMemoryLayer.b_updategate,\n",
       " YetAnotherGRUMemoryLayer.W_in_to_resetgate,\n",
       " YetAnotherGRUMemoryLayer.W_hid_to_resetgate,\n",
       " YetAnotherGRUMemoryLayer.b_resetgate,\n",
       " YetAnotherGRUMemoryLayer.W_in_to_hidden_update,\n",
       " YetAnotherGRUMemoryLayer.W_hid_to_hidden_update,\n",
       " YetAnotherGRUMemoryLayer.b_hidden_update,\n",
       " policy_original.W,\n",
       " policy_original.b]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Obtain them\n",
    "weights = lasagne.layers.get_all_params(probas_seq_layer,\n",
    "                                        trainable=True,\n",
    "                                       )\n",
    "weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total weights: 4186\n"
     ]
    }
   ],
   "source": [
    "total_weights = int(T.sum([T.prod(w.shape) for w in weights]).eval())\n",
    "print \"Total weights:\", total_weights\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Loss function\n",
    "\n",
    "simple crossentropy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "\n",
    "# take all predicitons but for last one (we don't have it's 'next')\n",
    "predicted_probas = probas_seq[:,:-1]\n",
    "\n",
    "# crop probabilities to avoid -Inf logarithms\n",
    "predicted_probas = T.maximum(predicted_probas,1e-10)\n",
    "\n",
    "# correct answers\n",
    "references = sequences_batch[:,1:]\n",
    "\n",
    "\n",
    "\n",
    "# familiar lasagne crossentropy\n",
    "model_loss = lasagne.objectives.categorical_crossentropy(\n",
    "    predicted_probas.reshape([-1,n_tokens]),\n",
    "    references.ravel()\n",
    ").mean()\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Regularizer for kicks\n",
    "from lasagne.regularization import regularize_network_params, l2\n",
    "reg_l2 = regularize_network_params(resolver,l2)*10**-5\n",
    "loss = model_loss + reg_l2\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "updates = lasagne.updates.adadelta(loss,weights,learning_rate=0.1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Compile the training and evaluation steps\n",
    "\n",
    "* Since we unroll scans, the first compile may take approx. 3.5 cups of tea\n",
    "* If you want it to just compile something, consider either of the two:\n",
    " * setting unroll_scan=False in recurrence and adding generator_passive.get_automatic_updates()\n",
    " * reducing sequence length to, say, 20"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "rng_updates = generator_passive.get_automatic_updates()\n",
    "\n",
    "train_fun = theano.function([sequences_batch],[loss],updates=rng_updates + updates)\n",
    "    \n",
    "evaluation_fun = theano.function([sequences_batch],[loss,model_loss,reg_l2],updates=rng_updates)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Generating letters recurrently\n",
    "* In this mode, generator's i+1'st input is it's i-th output\n",
    "* the only difference in terms of code is that instead of reading letters from sequence,\n",
    "  * we feed bach the previously generated ones "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Recurrent state dictionary: maps state outputs to prev_state inputs\n",
    "feedback_dict = OrderedDict([\n",
    "            (gru,prev_gru_layer),\n",
    "            (next_stack, prev_stack_layer),\n",
    "            (resolver,observation_layer) #here we feed actions back as next observations\n",
    "    ])\n",
    "\n",
    "\n",
    "# generation batch size\n",
    "gen_batch_size = T.scalar('generated batch size','int32')\n",
    "\n",
    "\n",
    "\n",
    "#define a lasagne recurrence that actually generates letters\n",
    "# it will be used to generate sample sequences.\n",
    "\n",
    "generator_active = Recurrence(\n",
    "    state_variables = feedback_dict,\n",
    "    tracked_outputs = [resolver,probability_layer],\n",
    "    \n",
    "    #note that now we do need to provide batch size since generator does not have any inputs\n",
    "    batch_size = gen_batch_size,\n",
    "    n_steps = SEQ_LENGTH\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#get sequences as before\n",
    "generated_probas_seq,generated_action_seq = lasagne.layers.get_output([generator_active[probability_layer],\n",
    "                                                                      generator_active[resolver]])\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# ~0.5 tea cups\n",
    "get_sequences = theano.function([gen_batch_size],generated_action_seq ,updates=generator_active.get_automatic_updates())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array(['b', 'a', 'a', 'b', 'a', 'a', 'a', 'c', '|', '|', 'c', '|', 'c',\n",
       "        'c', 'a', '|', 'a', 'b', 'b', 'a', 'a', 'b', 'b', 'a', '|', 'c',\n",
       "        'a', '|', 'a', 'a', 'b', 'a', '|', 'a', 'b', 'b', 'b', 'c', 'c',\n",
       "        'a', 'c', 'b', 'a', 'b', 'b', 'a', 'a', 'c', 'a', 'c'], \n",
       "       dtype='|S1'),\n",
       " array(['b', 'a', 'a', 'a', 'c', '|', 'c', 'c', 'b', 'c', 'c', 'c', 'b',\n",
       "        'b', 'c', 'c', 'a', 'a', 'c', 'a', 'b', 'a', 'c', 'c', 'a', 'a',\n",
       "        'c', 'b', '|', '|', 'b', 'a', '|', 'c', 'c', 'b', 'a', 'c', '|',\n",
       "        '|', 'c', 'a', 'a', 'c', 'a', 'a', 'b', 'b', 'a', 'a'], \n",
       "       dtype='|S1'),\n",
       " array(['c', 'c', 'a', '|', 'a', 'b', '|', '|', 'c', 'b', '|', 'a', 'c',\n",
       "        'c', '|', 'a', 'a', 'a', 'a', 'c', 'b', 'c', '|', 'c', '|', 'c',\n",
       "        'c', '|', 'b', 'a', 'b', '|', 'c', 'a', 'a', 'b', 'c', 'a', 'b',\n",
       "        'a', 'a', 'a', 'b', 'a', 'a', 'a', 'a', 'a', 'c', 'a'], \n",
       "       dtype='|S1')]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#sample output\n",
    "map(alphabet.__getitem__,get_sequences(3))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training loop:\n",
    "\n",
    "* Generating sequences\n",
    "* Training on them\n",
    "* Printing metrics every once in a while"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from collections import defaultdict\n",
    "\n",
    "metrics = defaultdict(dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iter:0\tfull:1.38606\tllh:1.38550\treg:0.00056\n",
      "Correct sequences ratio: 0.03651\n",
      "C-s count mean abs error: 2.77083\n",
      "a c c c b | a c a c | c c a c a a b | a b | c b c a a | | c a | a c b c a c b a a | c b b | | c b a\n",
      "c b a b | | a a b c | | b a b | a c | | a | a | b b b | | c b | | c | b c | b | c | c c b | c c c a\n",
      "| b a | b b c a a c a b | b c b c c c a b c b | a | a b a | b c a | c b b | b b c | a c b a c a c c\n",
      "iter:100\tfull:1.17679\tllh:1.17621\treg:0.00057\n",
      "Correct sequences ratio: 0.04174\n",
      "C-s count mean abs error: 2.44633\n",
      "b c a b b a c | c c c c b a b b a | b b c a a b b b a b c c c a c a c a | b a a c c c c a c c b c b\n",
      "b c | a a | a c c a c b c a c | a c a c a c | b c b c a | a c b c c | | a c c a a c a c a b b c c c\n",
      "b c | a c a b a | c a | b a c c | c | b | a a | c b c c b a b b a c c | c c c b c c b c c | c c b c\n",
      "iter:200\tfull:0.70853\tllh:0.70789\treg:0.00064\n",
      "Correct sequences ratio: 0.09359\n",
      "C-s count mean abs error: 5.29710\n",
      "b b a c a a a a b a a b c c c c c a b a b c a c c c b c c c c c c c c c c c c c c c | c c c c c c c\n",
      "a b | b b b a c c c c c a c b c c c c c c c c c c c c | a a b a c a a a b b a c c c c c c c c b c a\n",
      "c b c c c c | a a b b b c c c b c c | a c b c c c c c | | a b a b a c b c b a a b b b c b | a a b b\n",
      "iter:300\tfull:0.47940\tllh:0.47868\treg:0.00072\n",
      "Correct sequences ratio: 0.37048\n",
      "C-s count mean abs error: 3.87628\n",
      "a a a b b | b b b b c c c c c c c c c c | a a a a a a a b b b c c c c c c c c | a a a a b c b c c c\n",
      "a a a a a a a a b b b c c c | b c c c c c c c c c c | a a a a a a a b b b c c c c c c c c c c c c c\n",
      "a a b b a a b b b b b c c c c c c c c c c | a a a a b b | c b c c b b b c c c c c c c c c c c c c |\n",
      "iter:400\tfull:0.43196\tllh:0.43120\treg:0.00075\n",
      "Correct sequences ratio: 0.55009\n",
      "C-s count mean abs error: 3.44371\n",
      "a a b b b b b b b b b b b b b c c c c c c c c c a a a a a a a a a b b a c b c c c c c c | a a a b b\n",
      "b a a a | a a a a b a b b b c c c c c c c a a a a a a a b b b b c c c c c c c c c c c c c c b c c c\n",
      "c b b | a b a a b b b b b b b b c c c c c c c | a a a a a a b b b b b c c c c c c c c c c c c a a a\n",
      "iter:500\tfull:0.39694\tllh:0.39616\treg:0.00078\n",
      "Correct sequences ratio: 0.61560\n",
      "C-s count mean abs error: 2.98710\n",
      "a a a a a b b b b b b b b b b b b c c c c c c c c c c c c | a a b b c c c c c c c c c c c c | a a a\n",
      "c | | a a a a a a a a b b b c c c c | b a b b b b b b c c c c c c c c c c c c c | a a b b b b b b b\n",
      "a a a a b b b c c c c c c c c | a a a a b b b b b c c c c c c c c c | | a a a a a b b b b b b b b b\n",
      "iter:600\tfull:0.39211\tllh:0.39131\treg:0.00081\n",
      "Correct sequences ratio: 0.67982\n",
      "C-s count mean abs error: 2.44467\n",
      "a a a b b c c c c c c | a a a a b | a a a b b b c c c c c c c c c c c | a | a a a a a a a b b b c c\n",
      "a a b b b b c c c c c c c c | a | a a a a a a a b b b c c c c c c | a a a a b a b b c c c c c c c c\n",
      "a b b b c c c | b a a a a b b b b b b b b b b b b b b c c c c c c c c c c c c c | a a a a a b b b b\n",
      "iter:700\tfull:0.36507\tllh:0.36425\treg:0.00083\n",
      "Correct sequences ratio: 0.72361\n",
      "C-s count mean abs error: 2.25464\n",
      "a a a a a a b c c c c c c | | a a a a a b b b b c b c c c c c c c c c | | a a a a a a a b b b b b c\n",
      "| b a a b b b b b b c c c c c c c c c c | a a a b b b c c c c c | a a a a a a a b c c c c c c c c c\n",
      "a a a a a b b b b b b c c c c c c c c c | | a a a a b b c b c c c c c c c | a a a a a b b b b b b b\n",
      "iter:800\tfull:0.37251\tllh:0.37167\treg:0.00084\n",
      "Correct sequences ratio: 0.75122\n",
      "C-s count mean abs error: 1.88419\n",
      "a a a b b c c c c c | a | a a a a a a a b b b b b b b c c c c c c c c c c c c c c | a a a a a a a b\n",
      "a a a a a b b b b b b b c c c c c c c c c c c c a a a a a a a a b b b b b b b c c c c c c c c c c c\n",
      "| a a a a a a b b c b b b b b b b b b b b c c b c c c c c c c c c c c c c c | a a a b b b b c c c c\n",
      "iter:900\tfull:0.37239\tllh:0.37153\treg:0.00086\n",
      "Correct sequences ratio: 0.76047\n",
      "C-s count mean abs error: 1.63364\n",
      "a a b b b b b b b b c c c c c c c c c | a a a a a b b b c c c c c c c | a a a b c | a a b b b b b b\n",
      "a b b b b c c c c c c c | a a a a b b c c | c c c c | a a a a a a a a b b b b c c c c c c c c c c |\n",
      "| a a a c c | | | a a c | | a a a a a b c c c c | | a a a b b b b b b c c c c c c | a a a a a a b b\n",
      "iter:1000\tfull:0.34506\tllh:0.34418\treg:0.00088\n",
      "Correct sequences ratio: 0.79151\n",
      "C-s count mean abs error: 1.43780\n",
      "a a b b b b b b c c b b b c c c c c c c c c a a a a a b b c c c c c c c c c | a a a a a a a a a b c\n",
      "a a a a a b b b b c c c c c c c c c c | | a a a b b b b b b c c c c c c c c c | a a a a a a a a b b\n",
      "a a a a a a b b b b b c c c c c c c c c c | | a a b c c c c | a a a a b b b b b b c c c c c c c c c\n",
      "iter:1100\tfull:0.33705\tllh:0.33615\treg:0.00090\n",
      "Correct sequences ratio: 0.80829\n",
      "C-s count mean abs error: 1.29976\n",
      "b b b b b b b b c c c c c c c c | | a a a a a a b b b b b c c c c c c c | a a a a b b b c c c c c c\n",
      "a a a a a a a a c c c c c c c c c c | a a a a a b b b b b c c c c c c c c | a a a a b b b b b b b b\n",
      "a a a a a a b b b b c c c c c c c c c | a a a a a a a a a b b b c c c c c c c c c c c c | a a a a a\n",
      "iter:1200\tfull:0.32928\tllh:0.32836\treg:0.00092\n",
      "Correct sequences ratio: 0.81774\n",
      "C-s count mean abs error: 1.16297\n",
      "| | | a a a a a a b b b b b c c c c c c c c c c | a a a a a a b b b b b b c c c c c c c c c c | a a\n",
      "a a a a a a b b b b b b b b c c c c c c c c c c c c c c | a a a a a a b b b b b b b b b c c c c c c\n",
      "| a a a a a a a b b b b b c c c c c c c c c c | a a a a a a a b b b b b b c c c c c c c c c c c c c\n",
      "iter:1300\tfull:0.31686\tllh:0.31592\treg:0.00094\n",
      "Correct sequences ratio: 0.84897\n",
      "C-s count mean abs error: 1.02476\n",
      "| a a a a a a b c c c c c c c c | a a a a a a a b b b b b b b c c c c c c c c c c c c c c | | a a b\n",
      "a a b b b b b c c c c c c c c | | a a a b b b b b b b c c c c c c c c c c | | a a a | a a a b b b b\n",
      "a a a a a a a b b b b b b b b c c c c c c c c c c c c c c c c | | a a a a a b b c c c c c c | a a b\n",
      "iter:1400\tfull:0.32037\tllh:0.31942\treg:0.00096\n",
      "Correct sequences ratio: 0.84334\n",
      "C-s count mean abs error: 0.90972\n",
      "| a a a a a a a a a b b b b c c c c c c c c c c c c | a a b b b c c c c c c c | a a a a a b c c c c\n",
      "a a a a a a a a a b b b b c c c c c c c c c c c c c | a a a a a b b b b c c c c c c c | a a a a a a\n",
      "a a a a b b b b b c c c c c c c c c | a a a a a a b c c c c c c c | a a a a b b b b b c c c c c c c\n",
      "iter:1500\tfull:0.31796\tllh:0.31699\treg:0.00098\n",
      "Correct sequences ratio: 0.85771\n",
      "C-s count mean abs error: 0.75000\n",
      "a a a a a b b b b c c c c c c c c c c | a a a a a a b b b c c c c c c c c c | a a a b b c c c c | a\n",
      "a a a a a a b b b b b c c c c c c c c c c c c | a a a a a a b b c c c c c c c c | a a a a a a a b b\n",
      "| | a a a a a b a b b b c c c c c c c c c a a a a b b b b b b b b c c c c c c c c c c | a a a a a a\n",
      "iter:1600\tfull:0.29895\tllh:0.29796\treg:0.00100\n",
      "Correct sequences ratio: 0.87024\n",
      "C-s count mean abs error: 0.69775\n",
      "a b b b b b b b c c c c c c c c c | a a a a a b b b b c c c c c c c | a a a a b b b b c c c c c c c\n",
      "a a a a a a a a a b b b b b c c c c c c c c c c c c c c c | a a a a a a a a a b b b b b b c c c c c\n",
      "a a b b b b b c c c c c c c c | a a a a a a b b b b b c c c c c c c c c c | a a a a a b c c c c c c\n",
      "iter:1700\tfull:0.29498\tllh:0.29396\treg:0.00102\n",
      "Correct sequences ratio: 0.89027\n",
      "C-s count mean abs error: 0.60896\n",
      "| a a a a a a b b b b b b b b c c c c c c c c c c c c c c | a a a a a a a b b c c c c c c c c | a a\n",
      "a a a a a a a a b b b b c c c c c c c c c c c | a a a a a b b b b b b c c c c c c c c c c c c c | a\n",
      "a a a b b b b c c c c c c c | a a a a a a c c c c c c | | a a a a a a b b b b b c c c c c c c c c c\n",
      "iter:1800\tfull:0.28787\tllh:0.28683\treg:0.00104\n",
      "Correct sequences ratio: 0.88979\n",
      "C-s count mean abs error: 0.58920\n",
      "| a a a a a a a a b b b c c c c c c c c c c c | a a a a a a a a a b b b b b b b c c c c c c c c c c\n",
      "a a a a a a b b b b b b c c c c c c c c c c c c | a a a a b b b b c c c c c c c c | a a a a a a b b\n",
      "a a a a a b b b b b b b c c c c c c c c c c c | a a a a a a a a b b b b b b c c c c c c c c c c c c\n",
      "iter:1900\tfull:0.28745\tllh:0.28640\treg:0.00105\n",
      "Correct sequences ratio: 0.89888\n",
      "C-s count mean abs error: 0.50114\n",
      "a b c c | | a a a a b a a a b b c c c c c c c c c c | a a a a a a a b b b b c c c c c c c c c c c |\n",
      "a a a a a a a a a a b b b b b b c c c c c c c c c c c c c c c c | | a a a a a b b b b c c c c c c c\n",
      "a a a a a a a a a b b b b b b c c c c c c c c c c c c c c c | a a a a a a b b c c c c c c a a a b b\n",
      "iter:2000\tfull:0.28881\tllh:0.28774\treg:0.00107\n",
      "Correct sequences ratio: 0.90311\n",
      "C-s count mean abs error: 0.43413\n",
      "a a a a a a a a a b c c c c c c c c c c | a a a a a a a a a a b b b b b c c c c c c c c c c c c c c\n",
      "| a a a a a a a b b b b b b b c c c c c c c c c c c c c c | a a a a b b b c c c c c c c | a a a a a\n",
      "| a a a a a a b b b b b b b b b c c c c c c c c c c c c c c | a a a a a a a b b b b b b b b b b b c\n",
      "iter:2100\tfull:0.27489\tllh:0.27380\treg:0.00109\n",
      "Correct sequences ratio: 0.91234\n",
      "C-s count mean abs error: 0.35837\n",
      "a b b b b b c c c c c c | a a a a a a a b b b b b c c c c c c c c c c c | c c | a a a a a b b b b c\n",
      "| a a a a b b c c c c c c c | a a a a a b b b c c c c c c c | a a a a a a a a a b b b b c c c c c c\n",
      "| a b b b b b c c c c c c | a a a a b b b b b b c c c c c c c c c | a a a a a b b b b b b b b c c c\n",
      "iter:2200\tfull:0.28200\tllh:0.28089\treg:0.00111\n",
      "Correct sequences ratio: 0.92665\n",
      "C-s count mean abs error: 0.34303\n",
      "| a a a a a b b b c c c c c c c c c | a a a a a a a a b b b b c c c c c c c c c c c | a a a a a b b\n",
      "| a a a b b b b b b c c c c c c c c | a b b c c c | a a a a b b b c c c c c c c | a b b b b b b c c\n",
      "| a a a a a a b b b c c c c c c c c c | a a b b b b b b b b b c c c c c c c c c c | a a a a a b b b\n",
      "iter:2300\tfull:0.26470\tllh:0.26357\treg:0.00113\n",
      "Correct sequences ratio: 0.92717\n",
      "C-s count mean abs error: 0.31798\n",
      "| a a a a a a a b b c c c c c c c c c | a a a b b b b b b b b c c c c c c c c c c c | a a a a a a b\n",
      "| a a a a a a a a b b b c c c c c c c c c c c | a a a a a a a a a a a a b c c c c c c c c c c c c c\n",
      "| a a a a b b b b b c c c c c c c c c | a a a a a a a b b b c c c c c c c c c c | a a a a a a a a a\n",
      "iter:2400\tfull:0.27150\tllh:0.27036\treg:0.00115\n",
      "Correct sequences ratio: 0.95388\n",
      "C-s count mean abs error: 0.30373\n",
      "a a a a a a b b b b b c c c c c c c c c c c | a b b b b c c c c c | a a b b b b b b b b b c c c c c\n",
      "c | | a a a a a b b b b b c c c c c c c c c c | a a a a a a a b b c c c c c c c c | a a a a b b b b\n",
      "| a a b c c c c | a a a a a a a a b b b b b b b c c c c c c c c c c c c c c | a a a b b b b c c c c\n",
      "iter:2500\tfull:0.28001\tllh:0.27885\treg:0.00116\n",
      "Correct sequences ratio: 0.95205\n",
      "C-s count mean abs error: 0.26076\n",
      "| | a a a a a b c c c c c c | a b b b b c c c c | a a a a a b b b b b b c c c c c c c c c c c | a a\n",
      "a a a a a b b b c c c c c c c c c | a a a a a a a a b b b c c c c c c c c c c c | a a a a b b b b b\n",
      "a a a a a a a a b b b b b b c c c c c c c c c c c c c c c | a a a a a a a a b b c c c c c c c c c c\n",
      "iter:2600\tfull:0.26287\tllh:0.26169\treg:0.00118\n",
      "Correct sequences ratio: 0.94433\n",
      "C-s count mean abs error: 0.24672\n",
      "a b b c c c | a b b b b b c c c c c c | a a a a a b b b b b b c c c c c c c c c c c | a a b b b c c\n",
      "| a a a a a a a b b b c c c c c c c c c c | a b b b b b b b c c c c c c c c | a b b b b b b c c c c\n",
      "| a a a a b b b b b b b b c c c c c c c c c c c c | a a a a a a a a a b b c c c c c c c c c c c | a\n",
      "iter:2700\tfull:0.25490\tllh:0.25370\treg:0.00119\n",
      "Correct sequences ratio: 0.95506\n",
      "C-s count mean abs error: 0.20065\n",
      "| a a a b b c c c c c | a a a a a a a c c c c c c c c | a a a a a b b b b b b b c c c c c c c c c c\n",
      "| a a a a a a b b b b b c c c c c c c c c c | a a a b c c c c | a a b c c c c | a a a a a a a a a a\n",
      "| | a a a a a a a a b c c c c c c c c c | a a a b b b b b b b b b b b c c c c c c c c c c c c c c |\n",
      "iter:2800\tfull:0.25837\tllh:0.25716\treg:0.00121\n",
      "Correct sequences ratio: 0.95643\n",
      "C-s count mean abs error: 0.20065\n",
      "a a a b b c c c c c | a a a a b b b b b b b b b c c c c c c c c c c c c | a a a a a a b b b b b b c\n",
      "| | a a a a a a a a a a a b b b b b b b b c c c c c c c c c c c c c c c c c c c | a a a a a b c c c\n",
      "a b b b b c c c c c | a a a b b c c c c c | a a a b b b b b b b b b c c c c c c c c c c c c | a a a\n",
      "iter:2900\tfull:0.24691\tllh:0.24568\treg:0.00123\n",
      "Correct sequences ratio: 0.96071\n",
      "C-s count mean abs error: 0.17666\n",
      "| a a a a b b b b b b c c c c c c c c c c | a a b c c c | | a a a a a a a a a a b b b c c c c c c c\n",
      "| a a a a a a a a a b b b b b b c c c c c c c c c c c c c c c | a a a a a a a a b b b c c c c c c c\n",
      "| a a b b b b c c c c c c | a a a a a b b b b b b b c c c c c c c c c c c c | a a a a a a b b b b b\n",
      "iter:3000\tfull:0.24248\tllh:0.24124\treg:0.00124\n",
      "Correct sequences ratio: 0.96319\n",
      "C-s count mean abs error: 0.15877\n",
      "| a a a a a a b b c c c c c c c c | a a a a a a b c c c c c c c | a a a a a a a a b b b b b b c c c\n",
      "| a a a a a a b b b c c c c c c c c c | a a a a a a a a a a a a b c c c c c c c c c c c c c | a b b\n",
      "| a a a a a a a b b b b b b c c c c c c c c c c c c c | a a a a a b b b b c c c c c c c c c | a b b\n",
      "iter:3100\tfull:0.24585\tllh:0.24460\treg:0.00125\n",
      "Correct sequences ratio: 0.96399\n",
      "C-s count mean abs error: 0.13698\n",
      "| a a a a b b b b c c c c c c c c | a a a a a a a a b b b b b b b b b b c c c c c c c c c c c c c c\n",
      "| a a a a a b b b c c c c c c c c | a a a a a b b c c c c c c c | a a a a a a a a b b b b b c c c c\n",
      "| a a a a a b b c c c c c c c | a a b b b b b b c c c c c c c c c | a a a a a a a a a a b b b c c c\n",
      "iter:3200\tfull:0.25275\tllh:0.25148\treg:0.00127\n",
      "Correct sequences ratio: 0.96365\n",
      "C-s count mean abs error: 0.12221\n",
      "a a a a a b b b b b b c c c c c c c c c c c | a a b b b b b b b b c c c c c c c c c c | a a a a a a\n",
      "| a a a a a a a a b b b b b b c c c c c c c c c c c c c c | a a a a a a a b b b b c c c c c c c c c\n",
      "a b b c c c | a a b b b b b b c c c c c c c c | a a a a a b b b b b c c c c c c c c c c | a a a a a\n",
      "iter:3300\tfull:0.25576\tllh:0.25448\treg:0.00128\n",
      "Correct sequences ratio: 0.97895\n",
      "C-s count mean abs error: 0.10645\n",
      "| a a a a a a a a b b b b b b c c c c c c c c c c c c c c | a a a a a a a b b b b c c c c c c c c c\n",
      "a a a b c c c c | a a a a a a b b c c c c c c c c | a a a a b b b b b b b c c c c c c c c c c c | a\n",
      "| | a a a a a a a a b b b b b b b b c c c c c c c c c c c c c c c c | a b b b b b b c c c c c c c |\n",
      "iter:3400\tfull:0.25367\tllh:0.25238\treg:0.00130\n",
      "Correct sequences ratio: 0.97754\n",
      "C-s count mean abs error: 0.09737\n",
      "| a a a a a a a b b b b b b b b b b b c c c c c c c c c c c c c c c c c c | a a a a b b b b b c c c\n",
      "a a a a a b b b b b c c c c c c c c c c | a a a a a a a b b b b b c c c c c c c c c c c c | a a a a\n",
      "| a a a a a a a a a b b b b b c c c c c c c c c c c c c c | a a a a a a b b c c c c c c c c | a a a\n",
      "iter:3500\tfull:0.25402\tllh:0.25271\treg:0.00131\n",
      "Correct sequences ratio: 0.97780\n",
      "C-s count mean abs error: 0.10876\n",
      "| a a a a a a a a a a b b b b b b b b c c c c c c c c c c c c c c c c c c | a a a a a a b b c c c c\n",
      "| a a a a a a a a a b b b b b c c c c c c c c c c c c c c | a a b b b b b c c c c c c c | a a a a a\n",
      "| a b c c | a a a a a a a b b b b b b b b c c c c c c c c c c c c c c c | a a a a a a b b b c c c c\n",
      "iter:3600\tfull:0.24165\tllh:0.24033\treg:0.00132\n",
      "Correct sequences ratio: 0.97600\n",
      "C-s count mean abs error: 0.07910\n",
      "| | a a a a a a a b b c c c c c c c c c | a a b b b b b b b c c c c c c c c c | a a a a b b b b b b\n",
      "| a a a a a a a b b b b b b b b b c c c c c c c c c c c c c c c c | a a a a a a b b b b c c c c c c\n",
      "a a a a b b b b b b b b b b c c c c c c c c c c c c c c | a a a a a a a a b b b b b b b c c c c c c\n",
      "iter:3700\tfull:0.24184\tllh:0.24051\treg:0.00133\n",
      "Correct sequences ratio: 0.98462\n",
      "C-s count mean abs error: 0.08998\n",
      "| a a b b b b c c c c c c | a a a a a a a b b b b c c c c c c c c c c c | a a a a b b b c c c c c c\n",
      "| a a b b b b c c c c c c | a a a a a a a a a a b b b b b b c c c c c c c c c c c c c c c c | a a b\n",
      "| a a a a a a a a b b b b b b b b b b c c c c c c c c c c c c c c c c c c | a a a a a a a b b b b c\n",
      "iter:3800\tfull:0.23620\tllh:0.23486\treg:0.00134\n",
      "Correct sequences ratio: 0.98566\n",
      "C-s count mean abs error: 0.08459\n",
      "| a a b b b c c c c c | a b b b b c c c c c | a a b b b b c c c c c c | a a a a a a a b b b b b b c\n",
      "| a a a a a a a a a b b c c c c c c c c c c c | a b b b b c c c c | a a b b b b b b b c c c c c c c\n",
      "| | a a a a a a a b b b b b b c c c c c c c c c c c c c | a a b b b b b b b c c c c c c c c c | a a\n",
      "iter:3900\tfull:0.23592\tllh:0.23457\treg:0.00135\n",
      "Correct sequences ratio: 0.98016\n",
      "C-s count mean abs error: 0.08098\n",
      "| a a a a a a a b b b c c c c c c c c c c | a a b b b b b b b b c c c c c c c c c c | a b b b b b c\n",
      "| a a a a a a a a a a b b b b b b c c c c c c c c c c c c c c c c | a a a a b b b c c c c c c c | a\n",
      "| a a a a a a a a b b b b b b b b c c c c c c c c c c c c c c c c | a a a a a b b b b b b c c c c c\n",
      "iter:4000\tfull:0.24218\tllh:0.24082\treg:0.00136\n",
      "Correct sequences ratio: 0.98945\n",
      "C-s count mean abs error: 0.06983\n",
      "| a a a a a b b b b b b b c c c c c c c c c c c c | a a a b b b c c c c c c | a a a a a a a a b b b\n",
      "| a a a b b b b b b b c c c c c c c c c c | a a b b b b b b c c c c c c c c | a a a a a a a b c c c\n",
      "| a a b b b b b b b b b c c c c c c c c c c c | a a a a a a a a a a b b b c c c c c c c c c c c c c\n",
      "iter:4100\tfull:0.23584\tllh:0.23447\treg:0.00137\n",
      "Correct sequences ratio: 0.98112\n",
      "C-s count mean abs error: 0.06841\n",
      "| a a a a a a a a b b b b b b c c c c c c c c c c c c c c | a a a a a a b b b b b b c c c c c c c c\n",
      "a a a a a a b b c c c c c c c c | a a a a a a b b b b b b c c c c c c c c c c c c | a a a a a a a b\n",
      "| a a a a a b b b c c c c c c c c | a a a a a a a a a a b c c c c c c c c c c c | a a a a a a a a a\n",
      "iter:4200\tfull:0.23984\tllh:0.23846\treg:0.00138\n",
      "Correct sequences ratio: 0.98287\n",
      "C-s count mean abs error: 0.07181\n",
      "a a a a a a a b c c c c c c c c | a a a a a a a b b b b b b c c c c c c c c c c c c c | a a a a b b\n",
      "| a a a a b b b c c c c c c c | a a a a a a b b b b b b b b c c c c c c c c c c c c c c | a a b c c\n",
      "| a a a a a a b b b b b b c c c c c c c c c c c c | a a a a a b b b b b b b b c c c c c c c c c c c\n",
      "iter:4300\tfull:0.22443\tllh:0.22304\treg:0.00139\n",
      "Correct sequences ratio: 0.98687\n",
      "C-s count mean abs error: 0.05961\n",
      "| a a a a a b b b b b b b b b c c c c c c c c c c c c c c | a a a a b b b b b b b c c c c c c c c c\n",
      "| a a a a a a a a a a b b b c c c c c c c c c c c c c | a a a a a a a b b b b b b b b b c c c c c c\n",
      "| a a a b b b b b b b c c c c c c c c c c | a a a a a a b b b b b b b c c c c c c c c c c c c c | a\n",
      "iter:4400\tfull:0.23611\tllh:0.23472\treg:0.00140\n",
      "Correct sequences ratio: 0.98325\n",
      "C-s count mean abs error: 0.04899\n",
      "| a a a a a a b b b b b b b c c c c c c c c c c c c c | a a a a a a b b b b c c c c c c c c c c | a\n",
      "a a a a a b b b b b b b b c c c c c c c c c c c c c | a a b b b b b b b b c c c c c c c c c c | a a\n",
      "| | a b c c | a b b b b b b b c c c c c c c c | a a b b b b c c c c c c | a a a a a a b b b b b b c\n",
      "iter:4500\tfull:0.23632\tllh:0.23492\treg:0.00140\n",
      "Correct sequences ratio: 0.98642\n",
      "C-s count mean abs error: 0.05929\n",
      "| a b b b b b c c c c c c a a a a b b b b b b b b c c c c c c c c c c c c | a a a a a a a b b c c c\n",
      "| a a a a a a b b b b b b b b b b b c c c c c c c c c c c c c c c c c | a a a a a a b b b c c c c c\n",
      "| a a a a a a b c c c c c c c | a a a a b b b b b b b b b c c c c c c c c c c c c c | a a b b b b c\n",
      "iter:4600\tfull:0.23629\tllh:0.23488\treg:0.00141\n",
      "Correct sequences ratio: 0.98528\n",
      "C-s count mean abs error: 0.04803\n",
      "| a a b b b c c c c c | a a a a a a a a b b b b b b b b c c c c c c c c c c c c c c c c | a a a b b\n",
      "| a a a a a a a b b b b b b b b c c c c c c c c c c c c c c c | a a a a a a a b b b b b b b b c c c\n",
      "| a a a a a a a b b b b b b b b b b b c c c c c c c c c c c c c c c c c c | a a a a b b b b b b b b\n",
      "iter:4700\tfull:0.23422\tllh:0.23280\treg:0.00142\n",
      "Correct sequences ratio: 0.98688\n",
      "C-s count mean abs error: 0.04574\n",
      "| a a a b b b b b b b b b b b c c c c c c c c c c c c c c | a a a a b a b b b b b b b b b b c c c c\n",
      "| a a a a a a a b b b b b b b b b c c c c c c c c c c c c c c c | a a a a b b b b b c c c c c c c c\n",
      "| a a a a a a a b b b b b b b c c c c c c c c c c c c c c | a a a a a a a b b b b b b b b c c c c c\n",
      "iter:4800\tfull:0.23460\tllh:0.23318\treg:0.00142\n",
      "Correct sequences ratio: 0.99000\n",
      "C-s count mean abs error: 0.04147\n",
      "| a a a a a a a a b b b b b b b b c c c c c c c c c c c c c c c c | a a a a a a a a b b b b b b b b\n",
      "| a a a a a a a a b b b b b c c c c c c c c c c c c c | a a a a a a a a b b b b b b b c c c c c c c\n",
      "| a a a a a a a a a a b b b b b b c c c c c c c c c c c c c c c c | a a a a b c c c c c | a a a a a\n",
      "iter:4900\tfull:0.23520\tllh:0.23377\treg:0.00143\n",
      "Correct sequences ratio: 0.99223\n",
      "C-s count mean abs error: 0.04021\n",
      "| a a a a a a b b b b b b b c c c c c c c c c c c c c | a a a a a a a a b b b b c c c c c c c c c c\n",
      "| | a a a b b b b c c c c c c c | a a a a a a a a a b b b b b b c c c c c c c c c c c c c c c | a a\n",
      "| a a a a a b b b b b b c c c c c c c c c c c | a a a a a a a b b b b b c c c c c c c c c c c c | a\n"
     ]
    }
   ],
   "source": [
    "\n",
    "for i in range(N_EPOCHS):\n",
    "    \n",
    "    #generate reference\n",
    "    new_batch = generate_sequence(BATCH_SIZE,SEQ_LENGTH)\n",
    "    #feed the monster\n",
    "    train_fun(new_batch)\n",
    "    \n",
    "    #Display metrics once in a while\n",
    "    if i % REPORT_RATE==0:\n",
    "        \n",
    "        loss_components = evaluation_fun(new_batch)\n",
    "        print \"iter:%i\\tfull:%.5f\\tllh:%.5f\\treg:%.5f\"%tuple([i]+map(float,loss_components))        \n",
    "        \n",
    "        metrics['crossentropy'][i]=float(loss_components[1])\n",
    "        \n",
    "\n",
    "        examples = get_sequences(1000)\n",
    "        \n",
    "        correctness_ratio,c_count_mae = get_metrics(examples,alphabet)\n",
    "        \n",
    "\n",
    "        metrics[\"correctness_rates\"][i] = correctness_ratio\n",
    "        metrics[\"c_count_errors\"][i]=c_count_mae\n",
    "        \n",
    "        print \"Correct sequences ratio: %.5f\"%(correctness_ratio)\n",
    "        print \"C-s count mean abs error: %.5f\"%(c_count_mae)\n",
    "        \n",
    "        for tid_line in examples[:3]:\n",
    "            print ' '.join(map(alphabet.__getitem__,tid_line))\n",
    "\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABIYAAAKvCAYAAAAfnMTEAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fab66085a10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJMAAAKvCAYAAADeG45oAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nk/d1BbBwcw4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fab5fd67c50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJMAAAKvCAYAAADeG45oAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nA5swohUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fab5fcec210>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for metric in metrics:\n",
    "    plt.figure(figsize=[14,8])\n",
    "    plt.plot(*zip(*sorted(metrics[metric].items(),key=lambda (k,v):k)),label=metric)\n",
    "    plt.legend(loc='best')\n",
    "    plt.ylabel(\"popugai\")\n",
    "    plt.xlabel(\"epoch\")\n",
    "    plt.grid()"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda env:py27]",
   "language": "python",
   "name": "conda-env-py27-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}