final_project_3.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "%matplotlib inline \n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sb\n",
    "import random\n",
    "import math\n",
    "import itertools as it\n",
    "\n",
    "\n",
    "def len_seqs(sequences, number):\n",
    "    seq_len = []\n",
    "    max_len = 0\n",
    "    for i in range(number):\n",
    "        seq_len.append(len(sequences[i]))\n",
    "        if len(sequences[i]) > max_len:\n",
    "            max_len = len(sequences[i])\n",
    "    for j in range(number):\n",
    "        if seq_len[j] < max_len:\n",
    "            add_spaces = max_len - seq_len[j]\n",
    "            sequences[j]= sequences[j] + (add_spaces*\"-\")\n",
    "    return sequences, number\n",
    "\n",
    "def count_spaces(seq):\n",
    "    count = 0\n",
    "    seq_without_space = []\n",
    "    for i in range(len(seq)):\n",
    "        if i==\"-\":\n",
    "            count = count +1\n",
    "        else:\n",
    "            seq_without_space.append(seq[i])\n",
    "            \n",
    "    return count, seq_without_space\n",
    "\n",
    "def random_seq(seq):\n",
    "    num_space, seq_no_space = count_spaces(seq)\n",
    "    space_position = []\n",
    "    for i in range(num_space):\n",
    "        space_position.append(random.randint(0, len(seq)-1))\n",
    "    \n",
    "    ran_seq = []\n",
    "    for l in range(len(seq)):\n",
    "        ran_seq.append(\"0\")\n",
    "\n",
    "    for j in range(num_space):\n",
    "        ran_seq[space_position[j]] = \"-\"\n",
    "    \n",
    "    m = 0\n",
    "    for k in range(len(seq)):\n",
    "        if ran_seq[k] == \"0\":\n",
    "            ran_seq[k] = seq_no_space[m]\n",
    "            m = m + 1\n",
    "    \n",
    "    ran_seq = str(ran_seq)\n",
    "    return ran_seq\n",
    "\n",
    "#for initialize a random individual\n",
    "#pass a list of sequences and a number of sequence, for example seq = [-AA-,A--A, AA--], number = 3\n",
    "def generate_random_agent():\n",
    "    sequences = [\"AAACCC\",\"ATAAAAAAAAAAAAAAAAAAAAAAATC\" , \"CTCATTTTTT\"]\n",
    "    number = len(sequences)\n",
    "    align_seqs, num = len_seqs(sequences, number) \n",
    "    \n",
    "    new_random_agent = []\n",
    "    for i in range(num):\n",
    "        ran_seq = random_seq(align_seqs[i])\n",
    "        new_random_agent.append(ran_seq)\n",
    "        \n",
    "    return tuple(new_random_agent)\n",
    "\n",
    "#initialize a random population (generation zero)\n",
    "def initialize_random_population(population_size):\n",
    "    random_population = []\n",
    "    for agent in range(population_size):\n",
    "        random_population.append(generate_random_agent())\n",
    "    return random_population\n",
    "\n",
    "#evaluate the fitness of all initial individuals in population\n",
    "def cal_fitness(agent, match_score, mismatch_score, gap_penalty):\n",
    "    fitness_score = 0\n",
    "    num_seq = len(agent)\n",
    "    len_seq = len(agent[0])\n",
    "    \n",
    "    for each_column in range(len_seq):\n",
    "        temp_column = \"\"\n",
    "        column_score = 0\n",
    "        for each_seq in range(num_seq):\n",
    "            temp_column = temp_column + agent[each_seq][each_column]\n",
    "            combinations = it.combinations(temp_column, 2)\n",
    "            com = list(combinations)\n",
    "        for ec in range(len(com)):\n",
    "            if com[ec][0] == com[ec][1]:\n",
    "                column_score = column_score + match_score \n",
    "            else:\n",
    "                if com[ec][0]== \"-\" or com[ec][1] == \"-\":\n",
    "                    column_score = column_score + gap_penalty\n",
    "                else:\n",
    "                    column_score = column_score + mismatch_score\n",
    "            fitness_score = fitness_score + column_score \n",
    "    return fitness_score\n",
    "        \n",
    "\n",
    "#select two individuals, X and Y, with highest fitness value from current evaluation\n",
    "#def select_parents(agent):   \n",
    "\n",
    "#recombination(crossover, mating)\n",
    "#def recombinate_parent()\n",
    "\n",
    "#muate the parent to get the new offsprings\n",
    "def mutate_agent(agent_genome, max_mutations=3):\n",
    "    agent_genome = list(agent_genome)\n",
    "    for i in range(len(agent_genome)):\n",
    "        agent_genome[i] = list(agent_genome[i])\n",
    "    num_mutations = random.randint(1, max_mutations)\n",
    "    \n",
    "    for mutation in range(num_mutations):\n",
    "        swap_index1 = random.randint(0, len(agent_genome[0]) - 1)\n",
    "        swap_index2 = swap_index1\n",
    "\n",
    "        while swap_index1 == swap_index2:\n",
    "            swap_index2 = random.randint(0, len(agent_genome[0]) - 1)\n",
    "\n",
    "        agent_genome[0][swap_index1], agent_genome[1][swap_index2] = agent_genome[0][swap_index2], agent_genome[1][swap_index1]\n",
    "    agent_genome =str(agent_genome)        \n",
    "    return tuple(agent_genome)\n",
    "\n",
    "# def shuffle_mutation(agent_genome):\n",
    "#     agent_genome = list(agent_genome)\n",
    "    \n",
    "#     start_index = random.randint(0, len(agent_genome) - 1)\n",
    "#     length = random.randint(2, 20)\n",
    "    \n",
    "#     genome_subset = agent_genome[start_index:start_index + length]\n",
    "#     agent_genome = agent_genome[:start_index] + agent_genome[start_index + length:]\n",
    "    \n",
    "#     insert_index = random.randint(0, len(agent_genome) + len(genome_subset) - 1)\n",
    "#     agent_genome = agent_genome[:insert_index] + genome_subset + agent_genome[insert_index:]\n",
    "    \n",
    "#     return tuple(agent_genome)\n",
    "\n",
    "#select the survivors from actual fitness\n",
    "#select the ones with highest fitness among all the parents, offsprings and new offsprings\n",
    "#def select_survival()\n",
    "\n",
    "#principle for running genetic algorithm(GA)\n",
    "def run_GA_MSA(generations=10000, population_size=100):\n",
    "    #initialize a random population (generation zero)\n",
    "    population = initialize_random_population(population_size)\n",
    "    \n",
    "    population_subset_size = int(population_size / 10.)\n",
    "    generations_10pct = int(generations / 10.)\n",
    "    match_score, mismatch_score, gap_penalty = 1, -1, -4\n",
    "    \n",
    "    #test for termination criteria(fitness score)\n",
    "    for generation in range(int(generations)):\n",
    "        population_fitness = {}\n",
    "        \n",
    "        #evaluate the fitness of all initial individuals in population \n",
    "        for agent in population:\n",
    "            if agent in population_fitness:\n",
    "                continue\n",
    "            population_fitness[agent] = cal_fitness(agent,match_score, mismatch_score, gap_penalty)\n",
    "        \n",
    "        new_population = []\n",
    "        #select = sorted(population_fitness,key=population_fitness.get,reverse=True)[:population_subset_size]\n",
    "        #select and recombinate individuals with highest fitness value(as parents) from current evaluation\n",
    "        for rank, agent_genome in enumerate(sorted(population_fitness,\n",
    "                                                   key=population_fitness.get,reverse=True)[:population_subset_size]):\n",
    "\n",
    "            if (generation % generations_10pct == 0 or generation == (generations - 1)) and rank == 0:\n",
    "                print(\"Generation %d best: %f\" % (generation, population_fitness[agent_genome]))\n",
    "                print(agent_genome) \n",
    "        \n",
    "            # create 1 exact copy of each top path\n",
    "            new_population.append(agent_genome)\n",
    "\n",
    "            #create 4 offspring with 1-3 mutations\n",
    "            #for offspring in range(4):\n",
    "            #    new_population.append(mutate_agent(agent_genome, 3))\n",
    "                \n",
    "            # create 5 offspring with a single shuffle mutation\n",
    "#             for offspring in range(5):\n",
    "#                 new_population.append(shuffle_mutation(agent_genome))\n",
    "\n",
    "        # Replace the old population with the new population of offspring\n",
    "        for i in range(len(population))[::-1]:\n",
    "            del population[i]\n",
    "\n",
    "        population = new_population"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Generation 0 best: 73.000000\n",
      "(\"['A', 'A', 'A', '-']\", \"['A', 'T', 'A', 'A']\", \"['C', 'T', 'C', 'A']\")\n",
      "Generation 1000 best: 83.000000\n",
      "(\"['A', 'T', 'A', 'A']\", \"['C', 'T', 'C', 'A']\", \"['A', 'A', 'A', '-']\")\n",
      "Generation 2000 best: 83.000000\n",
      "(\"['A', 'T', 'A', 'A']\", \"['C', 'T', 'C', 'A']\", \"['A', 'A', 'A', '-']\")\n",
      "Generation 3000 best: 83.000000\n",
      "(\"['A', 'T', 'A', 'A']\", \"['C', 'T', 'C', 'A']\", \"['A', 'A', 'A', '-']\")\n",
      "Generation 4000 best: 83.000000\n",
      "(\"['A', 'T', 'A', 'A']\", \"['C', 'T', 'C', 'A']\", \"['A', 'A', 'A', '-']\")\n",
      "Generation 5000 best: 83.000000\n",
      "(\"['A', 'T', 'A', 'A']\", \"['C', 'T', 'C', 'A']\", \"['A', 'A', 'A', '-']\")\n",
      "Generation 6000 best: 83.000000\n",
      "(\"['A', 'T', 'A', 'A']\", \"['C', 'T', 'C', 'A']\", \"['A', 'A', 'A', '-']\")\n",
      "Generation 7000 best: 83.000000\n",
      "(\"['A', 'T', 'A', 'A']\", \"['C', 'T', 'C', 'A']\", \"['A', 'A', 'A', '-']\")\n",
      "Generation 8000 best: 83.000000\n",
      "(\"['A', 'T', 'A', 'A']\", \"['C', 'T', 'C', 'A']\", \"['A', 'A', 'A', '-']\")\n",
      "Generation 9000 best: 83.000000\n",
      "(\"['A', 'T', 'A', 'A']\", \"['C', 'T', 'C', 'A']\", \"['A', 'A', 'A', '-']\")\n",
      "Generation 9999 best: 83.000000\n",
      "(\"['A', 'T', 'A', 'A']\", \"['C', 'T', 'C', 'A']\", \"['A', 'A', 'A', '-']\")\n"
     ]
    }
   ],
   "source": [
    "run_GA_MSA(generations=10000, population_size=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Generation 0 best: 160.000000\n",
      "(\"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-']\", \"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', '-', '-']\", \"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T']\")\n",
      "Generation 1000 best: 170.000000\n",
      "(\"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-']\", \"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T']\")\n",
      "Generation 2000 best: 170.000000\n",
      "(\"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-']\", \"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T']\")\n",
      "Generation 3000 best: 170.000000\n",
      "(\"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-']\", \"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T']\")\n",
      "Generation 4000 best: 170.000000\n",
      "(\"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-']\", \"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T']\")\n",
      "Generation 5000 best: 170.000000\n",
      "(\"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-']\", \"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T']\")\n",
      "Generation 6000 best: 170.000000\n",
      "(\"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-']\", \"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T']\")\n",
      "Generation 7000 best: 170.000000\n",
      "(\"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-']\", \"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T']\")\n",
      "Generation 8000 best: 170.000000\n",
      "(\"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-']\", \"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T']\")\n",
      "Generation 9000 best: 170.000000\n",
      "(\"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-']\", \"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T']\")\n",
      "Generation 9999 best: 170.000000\n",
      "(\"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-']\", \"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T']\")\n"
     ]
    }
   ],
   "source": [
    "run_GA_MSA(generations=10000, population_size=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Generation 0 best: 306.000000\n",
      "(\"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'T', 'C']\", \"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\")\n",
      "Generation 1000 best: 399.000000\n",
      "(\"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'T', 'C']\")\n",
      "Generation 2000 best: 399.000000\n",
      "(\"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'T', 'C']\")\n",
      "Generation 3000 best: 399.000000\n",
      "(\"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'T', 'C']\")\n",
      "Generation 4000 best: 399.000000\n",
      "(\"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'T', 'C']\")\n",
      "Generation 5000 best: 399.000000\n",
      "(\"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'T', 'C']\")\n",
      "Generation 6000 best: 399.000000\n",
      "(\"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'T', 'C']\")\n",
      "Generation 7000 best: 399.000000\n",
      "(\"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'T', 'C']\")\n",
      "Generation 8000 best: 399.000000\n",
      "(\"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'T', 'C']\")\n",
      "Generation 9000 best: 399.000000\n",
      "(\"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'T', 'C']\")\n",
      "Generation 9999 best: 399.000000\n",
      "(\"['C', 'T', 'C', 'A', 'T', 'T', 'T', 'T', 'T', 'T', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'A', 'A', 'C', 'C', 'C', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-', '-']\", \"['A', 'T', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'T', 'C']\")\n"
     ]
    }
   ],
   "source": [
    "run_GA_MSA(generations=10000, population_size=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.4.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}