mids_ml@s_hw_09_groupY.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Julian Phillips, Steven Schaefer, Amir Ziai**\n",
    "\n",
    "julian.phillips@gmail.com<br>\n",
    "steven.schaefer@ischool.berkeley.edu<br>\n",
    "amir@ischool.berkeley.edu<br>\n",
    "\n",
    "UC Berkeley MIDS<br>\n",
    "Machine Learning at Scale<br>\n",
    "Week 9 Assignment<br>\n",
    "November 3, 2015"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## HW 9.0\n",
    "  \n",
    "**What is PageRank and what is it used for in the context of web search?**  \n",
    "  \n",
    "*PageRank is an algorithm used to rank websites, and is one among many features used to improve the quality of search results. It looks at the number of incoming links, and how highly rated the pages are that link to a page to determine an overall score. The idea being important, useful websites also link to important, useful websites. Pages with few links may not be valuable.*\n",
    "\n",
    "**What modifications have to be made to the webgraph in order to leverage the machinery of Markov Chains to compute the steady stade distibuton?**  \n",
    "  \n",
    "*To ensure the algorithm converges, we must redistribute mass from dangling nodes, and factor in a constant teleportation factor, which allows the random surfer to teleport from its current page to any other page with a uniform distribution.*\n",
    "\n",
    "\n",
    "## Addendum questions from HW7\n",
    "\n",
    "**Can we utilize combiners in the HW 7 to perform the shortest path implementation?**  \n",
    "  \n",
    "*Yes, we can utilize combiners to reduce the number of Queued entries for a particular node. Only the shortest one is of value.*\n",
    "  \n",
    "**Does order inversion help with the HW 7 shortest path implementation?**  \n",
    "  \n",
    "*Order inversion is useful in situations where we need to use some aggregate statistic to do some computation such as normalization of sub-parts, e.g. probabilities of word co-occurance. This is of no use to weighted shortest paths. For the unweighted version, it can be useful to ensure the 'visited' record arrives first, so that subsequent queue records can be ignored. However for the weighted version we must make sure we haven't found a shorter path, therefor sorting visisted records first does not help.*"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## HW 9.1\n",
    "  \n",
    "**Write a basic MRJob implementation of the iterative PageRank algorithm that takes sparse adjacency lists as input (as explored in HW 7). Make sure that you implementation utilizes teleportation (1-damping/the number of nodes in the network), and further, distributes the mass of dangling nodes with each iteration so that the output of each iteration is correctly normalized (sums to 1).**  \n",
    "  \n",
    "*Step 0: Pre-process the index files to count the total number of nodes in the graph. This value will be input to the real page-rank MR job.*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Overwriting NodeCounter.py\n"
     ]
    }
   ],
   "source": [
    "%%writefile NodeCounter.py\n",
    "#!/usr/bin/python\n",
    "\n",
    "from mrjob.job import MRJob\n",
    "from mrjob.step import MRStep\n",
    "\n",
    "import ast\n",
    "import itertools\n",
    "\n",
    "class NodeCounter(MRJob):\n",
    "    \"\"\" For testing, to ensure inline / hadoop runner use multiple reducers\n",
    "\n",
    "    def jobconf(self):\n",
    "        orig_jobconf = super(NodeCounter, self).jobconf()        \n",
    "        custom_jobconf = {\n",
    "            'mapred.reduce.tasks': '2',\n",
    "        }\n",
    "        combined_jobconf = orig_jobconf\n",
    "        combined_jobconf.update(custom_jobconf)\n",
    "\n",
    "        self.jobconf = combined_jobconf\n",
    "        return combined_jobconf\n",
    "    \"\"\"\n",
    "    def mapper_init(self):\n",
    "        self.seen_nodes = set()\n",
    "\n",
    "    def mapper(self, _, line):\n",
    "        components = line.strip().split(\"\\t\")\n",
    "\n",
    "        if len(components) == 2:\n",
    "\n",
    "            node = components[0]\n",
    "            try:\n",
    "                neighbors = ast.literal_eval(components[1])\n",
    "            except ValueError:\n",
    "                print \"Bad neighbors: \" + components[1]\n",
    "                return\n",
    "\n",
    "            for n in itertools.chain([node], neighbors.keys()):\n",
    "                if not n in self.seen_nodes:\n",
    "                    self.seen_nodes.add(n)\n",
    "                    yield n, None\n",
    "\n",
    "    def reducer_init(self):\n",
    "        self.node_count = 0\n",
    "\n",
    "    def reducer(self, key, values):\n",
    "        self.node_count += 1\n",
    "\n",
    "    def reducer_final(self):\n",
    "        yield \"NodeCount\", self.node_count\n",
    "\n",
    "    def summary_mapper(self, k, v):\n",
    "        yield k, v\n",
    "\n",
    "    def summary_reducer(self, k, values):\n",
    "        yield k, sum(values)\n",
    "\n",
    "    def steps(self):\n",
    "        summary_conf = {\n",
    "            'mapred.reduce.tasks': '1'\n",
    "        }\n",
    "\n",
    "        return [\n",
    "            MRStep(mapper_init=self.mapper_init, mapper=self.mapper,\\\n",
    "                   reducer_init=self.reducer_init, reducer=self.reducer,\\\n",
    "                   reducer_final=self.reducer_final),\n",
    "            MRStep(mapper=self.summary_mapper, reducer=self.summary_reducer,\\\n",
    "                jobconf=summary_conf)\n",
    "        ]\n",
    "\n",
    "if __name__ == '__main__':\n",
    "    NodeCounter.run()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "using configs in /home/steven/.mrjob.conf\n",
      "creating tmp directory /tmp/NodeCounter.steven.20151031.164538.792789\n",
      "writing to /tmp/NodeCounter.steven.20151031.164538.792789/step-0-mapper_part-00000\n",
      "Counters from step 1:\n",
      "  (no counters found)\n",
      "writing to /tmp/NodeCounter.steven.20151031.164538.792789/step-0-mapper-sorted\n",
      "> sort /tmp/NodeCounter.steven.20151031.164538.792789/step-0-mapper_part-00000\n",
      "writing to /tmp/NodeCounter.steven.20151031.164538.792789/step-0-reducer_part-00000\n",
      "Counters from step 1:\n",
      "  (no counters found)\n",
      "writing to /tmp/NodeCounter.steven.20151031.164538.792789/step-1-mapper_part-00000\n",
      "Counters from step 2:\n",
      "  (no counters found)\n",
      "writing to /tmp/NodeCounter.steven.20151031.164538.792789/step-1-mapper-sorted\n",
      "> sort /tmp/NodeCounter.steven.20151031.164538.792789/step-1-mapper_part-00000\n",
      "writing to /tmp/NodeCounter.steven.20151031.164538.792789/step-1-reducer_part-00000\n",
      "Counters from step 2:\n",
      "  (no counters found)\n",
      "Moving /tmp/NodeCounter.steven.20151031.164538.792789/step-1-reducer_part-00000 -> /tmp/NodeCounter.steven.20151031.164538.792789/output/part-00000\n",
      "Streaming final output from /tmp/NodeCounter.steven.20151031.164538.792789/output\n",
      "\"NodeCount\"\t11\n",
      "removing tmp directory /tmp/NodeCounter.steven.20151031.164538.792789\n"
     ]
    }
   ],
   "source": [
    "!python NodeCounter.py data/PageRank-test_indexed.txt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Implementation Notes for PageRank\n",
    "\n",
    "*This version of PageRank is implemented to use two steps per iteration, for a specified number of iterations. The second step accumulates lost pagerank mass, and does the final PageRank calculation in a single reducer. This is of course a disadvantage for the ultimate scalability of the implementation, but sufficient for the datasets in this assignment. The key advantage however is that it simplifies implementation in MRJob, eliminating the need for explicit synchronization in the driver, e.g. reading counters. This design simplification proved valuable: during the run of topic page rank on the wikipedia dataset, the VM that launched the EMR job crashed. Had it been actively driving the job, this would have at a minimum caused headaches restarting from the last good iteration, or at worst wasted over 20 hours of instance time. Had the worst happened, we would not have completed the topic run on EMR, we would not have likely had time to redo it.*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Overwriting PageRank.py\n"
     ]
    }
   ],
   "source": [
    "%%writefile PageRank.py\n",
    "#!/usr/bin/python\n",
    "\n",
    "from mrjob.job import MRJob\n",
    "from mrjob.step import MRStep\n",
    "\n",
    "import ast\n",
    "import math\n",
    "\n",
    "class PageRank(MRJob):\n",
    "    def configure_options(self):\n",
    "        super(PageRank, self).configure_options()\n",
    "\n",
    "        self.add_passthrough_option('--node-count', type='float')\n",
    "        self.add_passthrough_option('--damping', type='float')\n",
    "        self.add_passthrough_option('--iterations', type='int')\n",
    "\n",
    "    def pr_mapper_init(self):\n",
    "        self.mass = 0.0\n",
    "\n",
    "    def pr_mapper(self, key, line):\n",
    "\n",
    "        # First iteration only\n",
    "        if isinstance(line, basestring):\n",
    "\n",
    "            components = line.strip().split('\\t')\n",
    "\n",
    "            node = components[0]\n",
    "\n",
    "            # Initialize page-rank to 1 / |G|\n",
    "            node_pr = 1 / self.options.node_count\n",
    "\n",
    "            neighbors = ast.literal_eval(components[1])\n",
    "        \n",
    "        # Subsequent iterations\n",
    "        else:\n",
    "            node = key\n",
    "            node_pr = line[0] # pagerank from previous iteration\n",
    "            neighbors = line[1]\n",
    "\n",
    "        if neighbors:\n",
    "            # Factor in multiple links to the same page\n",
    "            num_out_links = sum(neighbors.values())\n",
    "            \n",
    "            for neighbor, num_links in neighbors.iteritems():\n",
    "                yield neighbor, (\"pr\", num_links * (node_pr / num_out_links))\n",
    "\n",
    "        else:\n",
    "            # Accumulate mass for dangling nodes\n",
    "            # This will only run for subsequent iterations\n",
    "            self.mass += node_pr\n",
    "\n",
    "        yield node, (\"node\", neighbors)\n",
    "\n",
    "    def pr_mapper_final(self):\n",
    "        yield \"*loss\", self.mass\n",
    "\n",
    "    def pr_reducer_init(self):\n",
    "        self.loss = 0.0\n",
    "\n",
    "    def pr_reducer(self, key, values):\n",
    "\n",
    "        if key == \"*loss\":\n",
    "            # Won't be emitted by first iteration, dangling node\n",
    "            # block below will\n",
    "            self.loss = sum(values)\n",
    "        else:\n",
    "            next_pr = 0.0\n",
    "            node_data = None\n",
    "\n",
    "            for value in values:\n",
    "                if value[0] == \"pr\":\n",
    "                    next_pr += value[1]\n",
    "                elif value[0] == \"node\":\n",
    "                    node_data = value\n",
    "\n",
    "\n",
    "            # Regular node\n",
    "            if node_data:\n",
    "                neighbors = node_data[1]\n",
    "                # Output the accumulated pagerank mass. Next step will do final calc w/ damping\n",
    "                yield key, (next_pr, neighbors)\n",
    "\n",
    "            # Dangling node, first iteration only will have \"pr\" records with mass contributions\n",
    "            # but will itself not have a \"node\" entry, i.e it does not have its own record in the\n",
    "            # original input file, it's only listed as a neighbor. So we ensure the dangling node\n",
    "            # has its own record, but with an empty list of neighbors.\n",
    "            else:\n",
    "                self.loss += (1 / self.options.node_count)\n",
    "                \n",
    "                yield key, (next_pr, {})\n",
    "\n",
    "    def pr_reducer_final(self):\n",
    "        yield \"*loss\", self.loss\n",
    "\n",
    "    def update_mapper(self, key, value):\n",
    "        yield key, value\n",
    "\n",
    "    def update_reducer_init(self):\n",
    "        self.total_pr = 0.0\n",
    "\n",
    "    def update_reducer(self, key, values):\n",
    "        if key == \"*loss\":\n",
    "            self.loss = sum(values)\n",
    "        else:\n",
    "            (next_pr, neighbors) = values.next()\n",
    "\n",
    "            final_pr = (1 - self.options.damping) * (1 / self.options.node_count) + \\\n",
    "                self.options.damping * ( (self.loss / self.options.node_count) + next_pr)\n",
    "\n",
    "            self.total_pr += final_pr\n",
    "            yield key, (final_pr, neighbors)\n",
    "\n",
    "    def update_reducer_final(self):\n",
    "        #print \"Loss: %f; Total Pr: %f\" % (self.loss, self.total_pr)\n",
    "        pass\n",
    "\n",
    "    def steps(self):\n",
    "\n",
    "        # Unfortunately using the \"special key\" approach means we can only use a single reducer\n",
    "        summary_conf = {\n",
    "            'mapred.reduce.tasks': '1'\n",
    "        }\n",
    "\n",
    "        return [\n",
    "            MRStep(mapper_init=self.pr_mapper_init, mapper=self.pr_mapper, mapper_final=self.pr_mapper_final, \\\n",
    "                  reducer_init=self.pr_reducer_init, reducer = self.pr_reducer, reducer_final=self.pr_reducer_final),\n",
    "            MRStep(mapper=self.update_mapper, reducer=self.update_reducer, jobconf=summary_conf,\\\n",
    "                reducer_init=self.update_reducer_init, reducer_final=self.update_reducer_final)\n",
    "        ] * self.options.iterations\n",
    "\n",
    "if __name__ == '__main__':\n",
    "    PageRank.run()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\"1\"\t[0.032781493159347676, {}]\r\n",
      "\"10\"\t[0.016169479016858935, {\"5\": 1}]\r\n",
      "\"11\"\t[0.016169479016858935, {\"5\": 1}]\r\n",
      "\"2\"\t[0.38436978095287694, {\"3\": 1}]\r\n",
      "\"3\"\t[0.34294145336903575, {\"2\": 1}]\r\n",
      "\"4\"\t[0.03908709209997012, {\"1\": 1, \"2\": 1}]\r\n",
      "\"5\"\t[0.08088569323450434, {\"2\": 1, \"4\": 1, \"6\": 1}]\r\n",
      "\"6\"\t[0.03908709209997012, {\"2\": 1, \"5\": 1}]\r\n",
      "\"7\"\t[0.016169479016858935, {\"2\": 1, \"5\": 1}]\r\n",
      "\"8\"\t[0.016169479016858935, {\"2\": 1, \"5\": 1}]\r\n",
      "\"9\"\t[0.016169479016858935, {\"2\": 1, \"5\": 1}]\r\n"
     ]
    }
   ],
   "source": [
    "!python PageRank.py --iterations 50 --node-count 11 --damping \".85\" data/PageRank-test_indexed.txt 2>/dev/null"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## HW 9.2: Exploring PageRank teleportation and network plots\n",
    "\n",
    "**In order to overcome  problems such as disconnected components, the damping factor (a typical value for d is 0.85) can be varied. Using the graph in HW1, plot the test graph (using networkx, https://networkx.github.io/) for several values of the damping parameter alpha, so that each nodes radius is proportional to its PageRank score. In particular you should\n",
    "do this for the following damping factors: [0,0.25,0.5,0.75, 0.85, 1]**  \n",
    "  \n",
    "*First, we pre-compute the pagerank scores for the various damping factors and then save them off to a file. Then, we read them back in and plot each one*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "%reload_ext autoreload\n",
    "%autoreload 2\n",
    "\n",
    "from PageRank import PageRank\n",
    "\n",
    "def run_with_dampings(values):\n",
    "    for damping in values:\n",
    "        mr_job = PageRank(args=[\"--iterations\", '50', \"--node-count\", '11', \"--damping\", str(damping), 'data/PageRank-test_indexed.txt'])\n",
    "        with mr_job.make_runner() as runner: \n",
    "            runner.run()\n",
    "\n",
    "            with open(\"damping/test_\" + str(int(damping * 100)), 'w') as outputFile:\n",
    "                for line in runner.stream_output():\n",
    "                    key,value =  mr_job.parse_output_line(line)\n",
    "\n",
    "                    outputFile.write(key + \"\\t\" + str(value) + \"\\n\")\n",
    "            \n",
    "run_with_dampings([0,0.25,0.5,0.75, 0.85, 1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6wAAAVmCAYAAACdmDOsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nsAIAALAUAXJOP/HqsGEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f01e81e1310>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from os import listdir\n",
    "from os.path import isfile, join\n",
    "\n",
    "import ast\n",
    "\n",
    "def get_damping_from_filename(file):\n",
    "    i = file.index(\"test_\")\n",
    "    damping = file[i + len(\"test_\"):]\n",
    "    return int(damping)\n",
    "\n",
    "# Ensure the results files are processed in increasing order of dampness so the graphs are ordered reasonably\n",
    "def file_cmp(a, b):\n",
    "    a_damping = get_damping_from_filename(a)\n",
    "    b_damping = get_damping_from_filename(b)\n",
    "    \n",
    "    return cmp(a_damping, b_damping)\n",
    "\n",
    "def graph_pagerank(files):\n",
    "    \n",
    "    plt.figure(figsize=(16, 24))\n",
    "    \n",
    "    for i, file in enumerate(files):\n",
    "        nodes = {}\n",
    "\n",
    "        with open(file, 'r') as inputFile:\n",
    "\n",
    "            for line in inputFile:\n",
    "                line = line.strip()\n",
    "                node, data = line.split('\\t')\n",
    "                nodes[node] = ast.literal_eval(data)\n",
    "\n",
    "        node_list = sorted(nodes.keys())\n",
    "        node_sizes = map(lambda x: x * 10000, [nodes[n][0] for n in node_list])\n",
    "        title = \"Damping = \" + str(get_damping_from_filename(file) / 100.0)\n",
    "\n",
    "        G = nx.DiGraph()\n",
    "\n",
    "        G.add_nodes_from(node_list)\n",
    "\n",
    "        for node, data in nodes.iteritems():\n",
    "\n",
    "\n",
    "            edges = [(node, e) for e in data[1].keys()]\n",
    "            G.add_edges_from(edges)\n",
    "\n",
    "        pos=nx.spring_layout(G)\n",
    "\n",
    "        ax = plt.subplot(len(files) / 2, 2, i + 1)\n",
    "\n",
    "        nx.draw(G, pos, ax, node_size=node_sizes, node_color='w')\n",
    "\n",
    "        nx.draw_networkx_labels(G, pos, {node: node for node in node_list})\n",
    "        plt.title(title)\n",
    "    plt.show()\n",
    "    \n",
    "\n",
    "# Grab the filenames from the run above and graph them!\n",
    "mypath = \"damping/\"\n",
    "pagerank_files = sorted([ join(mypath, f) for f in listdir(mypath) if isfile(join(mypath,f)) ], file_cmp)\n",
    "graph_pagerank(pagerank_files)        "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## HW 9.3\n",
    "\n",
    "*Step 0: Count the number of nodes in the wikipedia set*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "!python NodeCounter.py -r emr --num-ec2-instances 3 --ec2-instance-type 'm1.small' \\\n",
    "s3://ucb-mids-mls-networks/wikipedia/all-pages-indexed-out.txt &> out/wiki_counter_out 1>out/wiki_counter_result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "using configs in /home/steven/.mrjob.conf\n",
      "Got unexpected opts from /home/steven/.mrjob.conf: ec2_instace_type\n",
      "inferring aws_region from scratch bucket's region (us-west-1)\n",
      "creating tmp directory /tmp/NodeCounter.steven.20151031.165544.873170\n",
      "writing master bootstrap script to /tmp/NodeCounter.steven.20151031.165544.873170/b.py\n",
      "Copying non-input files into s3://stevenschaefer-mids-mlscale/tmp/NodeCounter.steven.20151031.165544.873170/files/\n",
      "Waiting 5.0s for S3 eventual consistency\n",
      "Creating Elastic MapReduce job flow\n",
      "Job flow created with ID: j-ULLD4PWX7PVD\n",
      "Created new job flow j-ULLD4PWX7PVD\n",
      "Job launched 31.2s ago, status STARTING: Provisioning Amazon EC2 capacity\n",
      "Job launched 62.5s ago, status STARTING: Provisioning Amazon EC2 capacity\n",
      "Job launched 93.5s ago, status STARTING: Provisioning Amazon EC2 capacity\n",
      "Job launched 124.6s ago, status STARTING: Provisioning Amazon EC2 capacity\n",
      "Job launched 155.6s ago, status STARTING: Provisioning Amazon EC2 capacity\n",
      "Job launched 186.6s ago, status STARTING: Configuring cluster software\n",
      "Job launched 217.6s ago, status BOOTSTRAPPING: Running bootstrap actions\n",
      "Job launched 248.8s ago, status BOOTSTRAPPING: Running bootstrap actions\n",
      "Job launched 279.8s ago, status BOOTSTRAPPING: Running bootstrap actions\n",
      "Job launched 310.9s ago, status RUNNING: Running step (NodeCounter.steven.20151031.165544.873170: Step 1 of 2)\n",
      "Job launched 341.9s ago, status RUNNING: Running step (NodeCounter.steven.20151031.165544.873170: Step 1 of 2)\n",
      "Job launched 372.9s ago, status RUNNING: Running step (NodeCounter.steven.20151031.165544.873170: Step 1 of 2)\n",
      "Job launched 403.9s ago, status RUNNING: Running step (NodeCounter.steven.20151031.165544.873170: Step 1 of 2)\n",
      "Job launched 434.9s ago, status RUNNING: Running step (NodeCounter.steven.20151031.165544.873170: Step 1 of 2)\n",
      "Job launched 465.8s ago, status RUNNING: Running step (NodeCounter.steven.20151031.165544.873170: Step 1 of 2)\n",
      "Job launched 496.8s ago, status RUNNING: Running step (NodeCounter.steven.20151031.165544.873170: Step 1 of 2)\n",
      "Job launched 527.7s ago, status RUNNING: Running step (NodeCounter.steven.20151031.165544.873170: Step 1 of 2)\n",
      "Job launched 558.7s ago, status RUNNING: Running step (NodeCounter.steven.20151031.165544.873170: Step 1 of 2)\n",
      "Job launched 589.6s ago, status RUNNING: Running step (NodeCounter.steven.20151031.165544.873170: Step 1 of 2)\n",
      "Job launched 620.6s ago, status RUNNING: Running step (NodeCounter.steven.20151031.165544.873170: Step 1 of 2)\n",
      "...\n",
      "    Launched map tasks: 9\n",
      "    Launched reduce tasks: 1\n",
      "    Rack-local map tasks: 4\n",
      "    SLOTS_MILLIS_MAPS: 155784\n",
      "    SLOTS_MILLIS_REDUCES: 38559\n",
      "    Total time spent by all maps waiting after reserving slots (ms): 0\n",
      "    Total time spent by all reduces waiting after reserving slots (ms): 0\n",
      "  Map-Reduce Framework:\n",
      "    CPU time spent (ms): 9440\n",
      "    Combine input records: 0\n",
      "    Combine output records: 0\n",
      "    Map input bytes: 60\n",
      "    Map input records: 3\n",
      "    Map output bytes: 60\n",
      "    Map output materialized bytes: 210\n",
      "    Map output records: 3\n",
      "    Physical memory (bytes) snapshot: 1867718656\n",
      "    Reduce input groups: 1\n",
      "    Reduce input records: 3\n",
      "    Reduce output records: 1\n",
      "    Reduce shuffle bytes: 210\n",
      "    SPLIT_RAW_BYTES: 1395\n",
      "    Spilled Records: 6\n",
      "    Total committed heap usage (bytes): 1367908352\n",
      "    Virtual memory (bytes) snapshot: 6455361536\n",
      "Streaming final output from s3://stevenschaefer-mids-mlscale/tmp/NodeCounter.steven.20151031.165544.873170/output/\n",
      "removing tmp directory /tmp/NodeCounter.steven.20151031.165544.873170\n",
      "Removing all files in s3://stevenschaefer-mids-mlscale/tmp/NodeCounter.steven.20151031.165544.873170/\n",
      "Removing all files in s3://stevenschaefer-mids-mlscale/tmp/logs/j-ULLD4PWX7PVD/\n",
      "Terminating job flow: j-ULLD4PWX7PVD\n"
     ]
    }
   ],
   "source": [
    "!head -n 30 out/wiki_counter_out\n",
    "!echo \"...\"\n",
    "!tail -n 30 out/wiki_counter_out"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\"NodeCount\"\t15192277\r\n"
     ]
    }
   ],
   "source": [
    "!cat out/wiki_counter_result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "!python PageRank.py -r emr --num-ec2-instances 5 --ec2-instance-type 'm1.medium' --no-output \\\n",
    "--output-dir s3://stevenschaefer-mids-mlscale/wiki-pr-10 --iterations 10 --node-count 15192277 --damping \".85\" \\\n",
    "    s3://ucb-mids-mls-networks/wikipedia/all-pages-indexed-out.txt &> out/wki_pr_10_out"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "using configs in /home/steven/.mrjob.conf\n",
      "Got unexpected opts from /home/steven/.mrjob.conf: ec2_instace_type\n",
      "inferring aws_region from scratch bucket's region (us-west-1)\n",
      "creating tmp directory /tmp/PageRank.steven.20151031.221127.738607\n",
      "writing master bootstrap script to /tmp/PageRank.steven.20151031.221127.738607/b.py\n",
      "Copying non-input files into s3://stevenschaefer-mids-mlscale/tmp/PageRank.steven.20151031.221127.738607/files/\n",
      "Waiting 5.0s for S3 eventual consistency\n",
      "Creating Elastic MapReduce job flow\n",
      "Job flow created with ID: j-1AOCJO40K86BC\n",
      "Created new job flow j-1AOCJO40K86BC\n",
      "Job launched 31.4s ago, status STARTING: Provisioning Amazon EC2 capacity\n",
      "Job launched 62.5s ago, status STARTING: Provisioning Amazon EC2 capacity\n",
      "Job launched 93.6s ago, status STARTING: Provisioning Amazon EC2 capacity\n",
      "Job launched 124.9s ago, status STARTING: Provisioning Amazon EC2 capacity\n",
      "Job launched 156.1s ago, status STARTING: Configuring cluster software\n",
      "Job launched 187.4s ago, status BOOTSTRAPPING: Running bootstrap actions\n",
      "Job launched 218.5s ago, status BOOTSTRAPPING: Running bootstrap actions\n",
      "Job launched 249.7s ago, status RUNNING: Running step (PageRank.steven.20151031.221127.738607: Step 1 of 20)\n",
      "Job launched 281.0s ago, status RUNNING: Running step (PageRank.steven.20151031.221127.738607: Step 1 of 20)\n",
      "Job launched 312.3s ago, status RUNNING: Running step (PageRank.steven.20151031.221127.738607: Step 1 of 20)\n",
      "Job launched 343.5s ago, status RUNNING: Running step (PageRank.steven.20151031.221127.738607: Step 1 of 20)\n",
      "Job launched 374.8s ago, status RUNNING: Running step (PageRank.steven.20151031.221127.738607: Step 1 of 20)\n",
      "Job launched 406.0s ago, status RUNNING: Running step (PageRank.steven.20151031.221127.738607: Step 1 of 20)\n",
      "Job launched 437.2s ago, status RUNNING: Running step (PageRank.steven.20151031.221127.738607: Step 1 of 20)\n",
      "Job launched 468.4s ago, status RUNNING: Running step (PageRank.steven.20151031.221127.738607: Step 1 of 20)\n",
      "Job launched 499.6s ago, status RUNNING: Running step (PageRank.steven.20151031.221127.738607: Step 1 of 20)\n",
      "Job launched 530.7s ago, status RUNNING: Running step (PageRank.steven.20151031.221127.738607: Step 1 of 20)\n",
      "Job launched 561.9s ago, status RUNNING: Running step (PageRank.steven.20151031.221127.738607: Step 1 of 20)\n",
      "Job launched 593.0s ago, status RUNNING: Running step (PageRank.steven.20151031.221127.738607: Step 1 of 20)\n",
      "Job launched 624.2s ago, status RUNNING: Running step (PageRank.steven.20151031.221127.738607: Step 1 of 20)\n",
      "...\n",
      "    Data-local map tasks: 23\n",
      "    Launched map tasks: 24\n",
      "    Launched reduce tasks: 1\n",
      "    Rack-local map tasks: 1\n",
      "    SLOTS_MILLIS_MAPS: 2425783\n",
      "    SLOTS_MILLIS_REDUCES: 1445899\n",
      "    Total time spent by all maps waiting after reserving slots (ms): 0\n",
      "    Total time spent by all reduces waiting after reserving slots (ms): 0\n",
      "  Map-Reduce Framework:\n",
      "    CPU time spent (ms): 2309030\n",
      "    Combine input records: 0\n",
      "    Combine output records: 0\n",
      "    Map input bytes: 2601914792\n",
      "    Map input records: 15192284\n",
      "    Map output bytes: 2608102545\n",
      "    Map output materialized bytes: 1491948465\n",
      "    Map output records: 15192284\n",
      "    Physical memory (bytes) snapshot: 7793111040\n",
      "    Reduce input groups: 15192278\n",
      "    Reduce input records: 15192284\n",
      "    Reduce output records: 15192277\n",
      "    Reduce shuffle bytes: 1491948465\n",
      "    SPLIT_RAW_BYTES: 3192\n",
      "    Spilled Records: 44008240\n",
      "    Total committed heap usage (bytes): 6515412992\n",
      "    Virtual memory (bytes) snapshot: 20576526336\n",
      "removing tmp directory /tmp/PageRank.steven.20151031.221127.738607\n",
      "Removing all files in s3://stevenschaefer-mids-mlscale/tmp/PageRank.steven.20151031.221127.738607/\n",
      "Removing all files in s3://stevenschaefer-mids-mlscale/tmp/logs/j-1AOCJO40K86BC/\n",
      "Terminating job flow: j-1AOCJO40K86BC\n"
     ]
    }
   ],
   "source": [
    "!head -n 30 out/wki_pr_10_out\n",
    "!echo \"...\"\n",
    "!tail -n 30 out/wki_pr_10_out"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* * *\n",
    "\n",
    "Now to process the results and plot the graphs, and revive an old friend, TopN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Now, run for 50 iterations\n",
    "!python PageRank.py -r emr --num-ec2-instances 12 --ec2-instance-type 'm1.large' --no-output \\\n",
    "--output-dir s3://stevenschaefer-mids-mlscale/wiki-pr-50 --iterations 50 --node-count 15192277 --damping \".85\" \\\n",
    "    s3://ucb-mids-mls-networks/wikipedia/all-pages-indexed-out.txt &> out/wki_pr_50_out"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "using configs in /home/steven/.mrjob.conf\r\n",
      "Got unexpected opts from /home/steven/.mrjob.conf: ec2_instace_type\r\n",
      "inferring aws_region from scratch bucket's region (us-west-1)\r\n",
      "creating tmp directory /tmp/PageRank.steven.20151101.190848.600827\r\n",
      "writing master bootstrap script to /tmp/PageRank.steven.20151101.190848.600827/b.py\r\n",
      "Copying non-input files into s3://stevenschaefer-mids-mlscale/tmp/PageRank.steven.20151101.190848.600827/files/\r\n",
      "Waiting 5.0s for S3 eventual consistency\r\n",
      "Creating Elastic MapReduce job flow\r\n",
      "Job flow created with ID: j-MPNWDRKJFKSR\r\n",
      "Created new job flow j-MPNWDRKJFKSR\r\n",
      "Job launched 32.0s ago, status STARTING: Provisioning Amazon EC2 capacity\r\n",
      "Job launched 64.1s ago, status STARTING: Provisioning Amazon EC2 capacity\r\n",
      "Job launched 95.9s ago, status STARTING: Provisioning Amazon EC2 capacity\r\n",
      "Job launched 127.7s ago, status STARTING: Provisioning Amazon EC2 capacity\r\n",
      "Job launched 159.4s ago, status STARTING: Configuring cluster software\r\n",
      "Job launched 191.6s ago, status BOOTSTRAPPING: Running bootstrap actions\r\n",
      "Job launched 228.4s ago, status RUNNING: Running step (PageRank.steven.20151101.190848.600827: Step 1 of 100)\r\n",
      "Job launched 260.2s ago, status RUNNING: Running step (PageRank.steven.20151101.190848.600827: Step 1 of 100)\r\n",
      "Job launched 292.4s ago, status RUNNING: Running step (PageRank.steven.20151101.190848.600827: Step 1 of 100)\r\n",
      "Job launched 326.5s ago, status RUNNING: Running step (PageRank.steven.20151101.190848.600827: Step 1 of 100)\r\n",
      "Job launched 362.6s ago, status RUNNING: Running step (PageRank.steven.20151101.190848.600827: Step 1 of 100)\r\n",
      "Job launched 394.8s ago, status RUNNING: Running step (PageRank.steven.20151101.190848.600827: Step 1 of 100)\r\n",
      "Job launched 426.3s ago, status RUNNING: Running step (PageRank.steven.20151101.190848.600827: Step 1 of 100)\r\n",
      "Job launched 458.3s ago, status RUNNING: Running step (PageRank.steven.20151101.190848.600827: Step 1 of 100)\r\n",
      "Job launched 489.9s ago, status RUNNING: Running step (PageRank.steven.20151101.190848.600827: Step 1 of 100)\r\n",
      "Job launched 521.8s ago, status RUNNING: Running step (PageRank.steven.20151101.190848.600827: Step 1 of 100)\r\n",
      "Job launched 556.8s ago, status RUNNING: Running step (PageRank.steven.20151101.190848.600827: Step 1 of 100)\r\n",
      "Job launched 588.5s ago, status RUNNING: Running step (PageRank.steven.20151101.190848.600827: Step 1 of 100)\r\n",
      "Job launched 620.4s ago, status RUNNING: Running step (PageRank.steven.20151101.190848.600827: Step 1 of 100)\r\n",
      "Job launched 652.6s ago, status RUNNING: Running step (PageRank.steven.20151101.190848.600827: Step 1 of 100)\r\n"
     ]
    }
   ],
   "source": [
    "!head -n 30 out/wiki_pr_50_out"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%%writefile ResultsProcessor.py\n",
    "#!/usr/bin/python\n",
    "\n",
    "import argparse\n",
    "import ast\n",
    "\n",
    "from TopN import TopN\n",
    "\n",
    "if __name__ == \"__main__\":\n",
    "\n",
    "    parser = argparse.ArgumentParser()\n",
    "    parser.add_argument('file', type=argparse.FileType('r'))\n",
    "\n",
    "    args = parser.parse_args()\n",
    "\n",
    "    top = TopN(100)\n",
    "\n",
    "    for line in args.file:\n",
    "        node, values = line.strip().split('\\t')\n",
    "\n",
    "        node = node.translate(None, '\"')\n",
    "        values = ast.literal_eval(values)\n",
    "\n",
    "        pr = values[0]\n",
    "\n",
    "        top.insert((pr, node))\n",
    "\n",
    "    for result in top.get():\n",
    "        print result[1] + \"\\t\" + str(result[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%%writefile TopN.py\n",
    "import heapq\n",
    "\n",
    "class TopN:\n",
    "    def __init__(self, n):\n",
    "        self.limit = n\n",
    "        self.heap = []\n",
    "        \n",
    "    def insert(self, item):\n",
    "        self.heap.append(item)\n",
    "        # Using a max-heap to work around comparison issues between\n",
    "        # the tuples. Combined with negating confidence, this will\n",
    "        # ensure what's left are the highest confidence rules with\n",
    "        # the lowest-ordered item ids\n",
    "        heapq.heapify(self.heap)\n",
    "        \n",
    "        # Optimize out the need to check the limit every insert.\n",
    "        if len(self.heap) == self.limit:\n",
    "            self.insert = self.insert_replace\n",
    "\n",
    "        return None\n",
    "        \n",
    "    def insert_replace(self, item):\n",
    "        return heapq.heappushpop(self.heap, item)\n",
    "        \n",
    "    def get(self):\n",
    "        return heapq.nlargest(self.limit, self.heap)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Not shown - downloading from S3\n",
    "!python ResultsProcessor.py results/wiki_10_pr > wiki_10_top_100"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The top 100 nodes in the wikipedia set for 10 iterations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "13455888\t0.00154472471298\r\n",
      "4695850\t0.000671024071888\r\n",
      "5051368\t0.000598385680972\r\n",
      "1184351\t0.000598207353645\r\n",
      "2437837\t0.000462492892892\r\n",
      "6076759\t0.000455094006413\r\n",
      "4196067\t0.000442377888835\r\n",
      "13425865\t0.000441553517141\r\n",
      "6172466\t0.000422400200183\r\n",
      "1384888\t0.000401289560406\r\n",
      "6113490\t0.000395789247717\r\n",
      "14112583\t0.000394384728374\r\n",
      "7902219\t0.000370098784734\r\n",
      "10390714\t0.000365026496431\r\n",
      "12836211\t0.00036199488631\r\n",
      "6237129\t0.000351955584761\r\n",
      "6416278\t0.000348662356451\r\n",
      "13432150\t0.000339365106373\r\n",
      "1516699\t0.000332975002861\r\n",
      "7990491\t0.000307609062657\r\n",
      "9386580\t0.000295697041955\r\n",
      "15164193\t0.000284301851455\r\n",
      "7576704\t0.000280702869462\r\n",
      "12074312\t0.000278939094991\r\n",
      "3191491\t0.00027331798906\r\n",
      "2797855\t0.000272895494663\r\n",
      "11253108\t0.000272424060445\r\n",
      "13725487\t0.000270209455375\r\n",
      "2155467\t0.000266492981678\r\n",
      "7835160\t0.000261256778085\r\n",
      "4198751\t0.000258929983074\r\n",
      "9276255\t0.000254633864535\r\n",
      "10469541\t0.000253701471535\r\n",
      "11147327\t0.000251546450909\r\n",
      "5154210\t0.000249270296206\r\n",
      "2396749\t0.000248115891295\r\n",
      "3603527\t0.000246513672565\r\n",
      "3069099\t0.00023478481004\r\n",
      "12430985\t0.000233492889263\r\n",
      "9355455\t0.000226580046968\r\n",
      "14881689\t0.000221607447992\r\n",
      "9391762\t0.000212352481341\r\n",
      "981395\t0.000205927337584\r\n",
      "8697871\t0.000205546663299\r\n",
      "11245362\t0.000203904731349\r\n",
      "14112408\t0.000195789187005\r\n",
      "9562547\t0.000194412655091\r\n",
      "12067030\t0.000189164848726\r\n",
      "2614581\t0.000189084319667\r\n",
      "10527224\t0.000187113393458\r\n",
      "10728264\t0.00018541236935\r\n",
      "11582765\t0.000184152279344\r\n",
      "3191268\t0.000180898206284\r\n",
      "9997298\t0.000180411333911\r\n",
      "1441065\t0.000179322488292\r\n",
      "5490435\t0.000176548479333\r\n",
      "10566120\t0.000176325397113\r\n",
      "994890\t0.000173887413276\r\n",
      "14503460\t0.000173028765159\r\n",
      "13280859\t0.00017301606208\r\n",
      "9394907\t0.000170678409269\r\n",
      "1637982\t0.000168133426614\r\n",
      "2614578\t0.000166615773627\r\n",
      "8019937\t0.000165539243898\r\n",
      "14725161\t0.000164759920424\r\n",
      "12447593\t0.000164230495193\r\n",
      "6172167\t0.000152942939577\r\n",
      "3577363\t0.000151704426879\r\n",
      "12038331\t0.000146761798995\r\n",
      "10345830\t0.000144223304873\r\n",
      "10917716\t0.000143879855442\r\n",
      "13853369\t0.00014373406055\r\n",
      "1332806\t0.00014256592668\r\n",
      "4344962\t0.000142137700337\r\n",
      "6171937\t0.000141458395652\r\n",
      "1813634\t0.000141017343736\r\n",
      "1523975\t0.000140819964444\r\n",
      "1575979\t0.000139884810817\r\n",
      "2778099\t0.000138472414712\r\n",
      "9742161\t0.000137007707223\r\n",
      "4568647\t0.000135211831145\r\n",
      "9924814\t0.000135107535041\r\n",
      "12785678\t0.000134495476038\r\n",
      "14565507\t0.000134358646024\r\n",
      "14981725\t0.000132538141803\r\n",
      "4978429\t0.000130654545776\r\n",
      "1175360\t0.000130450657957\r\n",
      "8641167\t0.000130253525201\r\n",
      "1947095\t0.000129889420659\r\n",
      "14963657\t0.000129490102303\r\n",
      "10246542\t0.000128365218308\r\n",
      "4624519\t0.000126868447599\r\n",
      "3591832\t0.000125973918079\r\n",
      "13328060\t0.000125862437589\r\n",
      "2578813\t0.000125479277954\r\n",
      "5908108\t0.000124594108286\r\n",
      "12685893\t0.000123757420739\r\n",
      "9390959\t0.000123671537329\r\n",
      "12048800\t0.000122542503965\r\n",
      "3328327\t0.000118928013647\r\n"
     ]
    }
   ],
   "source": [
    "!cat wiki_10_top_100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Not shown, downloading from S3\n",
    "!python ResultsProcessor.py results/wiki_50_pr > wiki_50_top_100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "13455888\t0.00154482798353\r\n",
      "4695850\t0.000670847961875\r\n",
      "5051368\t0.000598232607382\r\n",
      "1184351\t0.000597707726193\r\n",
      "2437837\t0.000462634927125\r\n",
      "6076759\t0.000454793361456\r\n",
      "4196067\t0.000442596928982\r\n",
      "13425865\t0.000441785647576\r\n",
      "6172466\t0.000421653861826\r\n",
      "1384888\t0.000400985069817\r\n",
      "6113490\t0.000395477730263\r\n",
      "14112583\t0.000394460913044\r\n",
      "7902219\t0.000369622052276\r\n",
      "10390714\t0.000364245830988\r\n",
      "12836211\t0.000362500304597\r\n",
      "6237129\t0.000352006901858\r\n",
      "6416278\t0.000348430312772\r\n",
      "13432150\t0.000339325215501\r\n",
      "1516699\t0.00033300814843\r\n",
      "7990491\t0.000307852396813\r\n",
      "9386580\t0.000295946268457\r\n",
      "15164193\t0.000283967750637\r\n",
      "7576704\t0.000280495058846\r\n",
      "12074312\t0.000278926989105\r\n",
      "2797855\t0.000272922937003\r\n",
      "3191491\t0.000272922439925\r\n",
      "11253108\t0.000272292754986\r\n",
      "13725487\t0.000269524520408\r\n",
      "2155467\t0.000266298878671\r\n",
      "7835160\t0.000261072939724\r\n",
      "4198751\t0.000258912013576\r\n",
      "9276255\t0.000254604973381\r\n",
      "10469541\t0.000253098130271\r\n",
      "11147327\t0.00025129230839\r\n",
      "5154210\t0.00024868666872\r\n",
      "2396749\t0.000248155875074\r\n",
      "3603527\t0.000246342846012\r\n",
      "3069099\t0.000234627628013\r\n",
      "12430985\t0.000233348220623\r\n",
      "9355455\t0.000226589579724\r\n",
      "14881689\t0.000221443553638\r\n",
      "9391762\t0.000212495519503\r\n",
      "981395\t0.000205589431576\r\n",
      "8697871\t0.000205438375357\r\n",
      "11245362\t0.000203673077319\r\n",
      "14112408\t0.000195859502439\r\n",
      "9562547\t0.000194213280537\r\n",
      "12067030\t0.000189149769805\r\n",
      "2614581\t0.000188830322532\r\n",
      "10527224\t0.00018705925316\r\n",
      "10728264\t0.000185362368294\r\n",
      "11582765\t0.000184308991007\r\n",
      "3191268\t0.000180652287498\r\n",
      "9997298\t0.000180453899922\r\n",
      "1441065\t0.000179255238937\r\n",
      "5490435\t0.000176489132119\r\n",
      "10566120\t0.000176092532228\r\n",
      "994890\t0.000173729379253\r\n",
      "14503460\t0.000172987038732\r\n",
      "13280859\t0.000172963972351\r\n",
      "9394907\t0.000170666580444\r\n",
      "1637982\t0.000167926147051\r\n",
      "2614578\t0.000166385497588\r\n",
      "8019937\t0.000165617362893\r\n",
      "14725161\t0.00016465763045\r\n",
      "12447593\t0.000164205619746\r\n",
      "6172167\t0.000152737953956\r\n",
      "3577363\t0.000151841890731\r\n",
      "12038331\t0.00014673304813\r\n",
      "10345830\t0.000144001256343\r\n",
      "10917716\t0.000143990957894\r\n",
      "13853369\t0.000143917774019\r\n",
      "1332806\t0.000142490470368\r\n",
      "4344962\t0.000142117292657\r\n",
      "6171937\t0.00014129974658\r\n",
      "1813634\t0.000141048810446\r\n",
      "1523975\t0.00014078664044\r\n",
      "1575979\t0.000140044332978\r\n",
      "2778099\t0.000138573968878\r\n",
      "9742161\t0.000137113736561\r\n",
      "4568647\t0.000135113845335\r\n",
      "9924814\t0.000135030272948\r\n",
      "12785678\t0.000134656578243\r\n",
      "14565507\t0.00013431412645\r\n",
      "14981725\t0.000132453076434\r\n",
      "4978429\t0.000130609729426\r\n",
      "8641167\t0.000130466763883\r\n",
      "1175360\t0.000130319392406\r\n",
      "1947095\t0.000129731102208\r\n",
      "14963657\t0.000129439696894\r\n",
      "10246542\t0.000128334243903\r\n",
      "4624519\t0.000126621443125\r\n",
      "3591832\t0.00012592167663\r\n",
      "13328060\t0.000125862453716\r\n",
      "2578813\t0.000125588774308\r\n",
      "5908108\t0.000124509884841\r\n",
      "12685893\t0.000123783957834\r\n",
      "9390959\t0.000123781733269\r\n",
      "12048800\t0.000122455741951\r\n",
      "3328327\t0.000118780041689\r\n"
     ]
    }
   ],
   "source": [
    "!cat wiki_50_top_100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA8gAAAH4CAYAAACWt8SGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f7a0497de90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%reload_ext autoreload\n",
    "%autoreload 2\n",
    "%matplotlib inline\n",
    "\n",
    "from matplotlib import pyplot as py\n",
    "\n",
    "def fetch_pageranks(file):\n",
    "    pageranks = []\n",
    "    with open(file, 'r') as input_file:\n",
    "        for line in input_file:\n",
    "            _, pr = line.strip().split('\\t')\n",
    "            pageranks.append(pr)\n",
    "            \n",
    "    return pageranks\n",
    "\n",
    "def plot_pagerank(pageranks):\n",
    "        \n",
    "    x = range(1, len(pageranks) + 1)\n",
    "    \n",
    "    py.plot(x, pageranks, 'o')\n",
    "    \n",
    "    \"\"\"\n",
    "    # min() is used to hack limits for the wikipedia graph\n",
    "    lims = py.ylim()\n",
    "    py.ylim(0, min(lims[1] + 1, 50000))\n",
    "    \n",
    "    lims = py.xlim()\n",
    "    py.xlim(0, min(lims[1] + 1, 10000))\n",
    "    \"\"\"\n",
    "\n",
    "    py.xlabel('Rank')\n",
    "    py.ylabel('PageRank')\n",
    "    \n",
    "    \"\"\"\n",
    "    py.gca().get_yaxis().set_major_locator(MaxNLocator(integer=True))\n",
    "    py.gca().get_xaxis().set_major_locator(MaxNLocator(integer=True))\n",
    "    \"\"\"\n",
    "    \n",
    "    py.title('Distribution of Incoming Edges')\n",
    "\n",
    "py.figure(figsize=(16, 8))\n",
    "\n",
    "py.subplot(1, 2, 1)\n",
    "    \n",
    "wiki10 = fetch_pageranks('wiki_10_top_100')\n",
    "plot_pagerank(wiki10)\n",
    "\n",
    "py.title(\"Top-100 PageRanks for Wikipedia 10 Iterations\")\n",
    "\n",
    "py.subplot(1, 2, 2)\n",
    "wiki50 = fetch_pageranks('wiki_50_top_100')\n",
    "plot_pagerank(wiki50)\n",
    "\n",
    "py.title(\"Top-100 PageRanks for Wikipedia 50 Iterations\")\n",
    "    \n",
    "py.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Comment on your findings**  \n",
    "  \n",
    "We see in the plots very similar graphs, indicating that at 10 iterations the algorithm had already started to converge. The extra 40 iterations yielded minor changes in the pageranks, but overall the results look fairly stable: the top 20 and bottom 20 look to be in the same order for both 10 & 50 iterations. For this dataset, 10 iterations appears to be more than sufficient."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## HW 9.4: Topic-specific PageRank implementation using MRJob\n",
    "\n",
    "**Modify your PageRank implementation to produce a topic specific PageRank implementation,\n",
    "as described in:**\n",
    "  \n",
    "**http://www-cs-students.stanford.edu/~taherh/papers/topic-sensitive-pagerank.pdf**  \n",
    "  \n",
    "**Note in this article that there is a special caveat to ensure that the transition matrix is irreducible. This caveat lies in footnote 3 on page 3:**\n",
    "  \n",
    "\tA minor caveat: to ensure that M is irreducible when p\n",
    "\tcontains any 0 entries, nodes not reachable from nonzero\n",
    "\tnodes in p should be removed. In practice this is not problematic.\n",
    "  \n",
    "**and must be adhered to for convergence to be guaranteed.**\n",
    "\n",
    "**This is not a small task, especially as it it must be performed\n",
    "separately for each of the (10) topics.**\n",
    "  \n",
    "**So, instead of using this method for irreducibility, please comment on why the literature's method is difficult to implement, and what what extra computation it will require.**\n",
    "  \n",
    "*The difficulty with this \"minor caveat\", is that it implies we need to check connectivity from the topic nodes to the non-topic nodes, and exclude the non-topic nodes that are unreachable. This means for each topic, we would have to process the graph checking for connectivity, resulting in different copies of the graph for each topic. At a minimum, this adds extra bookkeeping state tracked in the graph file, and at worst yields a copy of the entire graph for each topic. Either way, data storage needs increase, and managing these extra graphs adds administrative costs.*  \n",
    "\n",
    "### Implementation Notes\n",
    "\n",
    "*This implementation adds functionality to the prior generic PageRank version used in the earlier parts of this assignment. Topics are kept in an array, in addition to the pagerank score and neighbors dictionary. This enables a single map-reduce job to calculate the general page rank, along with the topic-specific pageranks.*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%%writefile TopicPageRank.py\n",
    "#!/usr/bin/python\n",
    "\n",
    "from mrjob.job import MRJob\n",
    "from mrjob.step import MRStep\n",
    "\n",
    "import ast\n",
    "import math\n",
    "from collections import defaultdict\n",
    "\n",
    "class PageRank(MRJob):\n",
    "\n",
    "    LOSS_KEY = \"*loss\"\n",
    "\n",
    "    def configure_options(self):\n",
    "        super(PageRank, self).configure_options()\n",
    "\n",
    "        self.add_passthrough_option('--node-count', type='float')\n",
    "        self.add_passthrough_option('--damping', type='float')\n",
    "        self.add_passthrough_option('--iterations', type='int')\n",
    "        self.add_passthrough_option('--beta', type='float')\n",
    "\n",
    "        # Could get this from topic-file... but we know ahead of time so optimize that away\n",
    "        self.add_passthrough_option('--num-topics', type='int')\n",
    "        self.add_passthrough_option('--debug', action='store_true')\n",
    "\n",
    "        self.add_file_option('--topic-file')\n",
    "\n",
    "    def pr_mapper_init(self):\n",
    "        self.init_loss_accumulators()\n",
    "\n",
    "    def pr_mapper(self, key, line):\n",
    "\n",
    "        # First iteration only\n",
    "        if isinstance(line, basestring):\n",
    "\n",
    "            components = line.strip().split('\\t')\n",
    "\n",
    "            node = components[0]\n",
    "\n",
    "            # Initialize page-rank to 1 / |G|\n",
    "            node_pr = 1 / self.options.node_count\n",
    "            topic_pr = [node_pr] * self.options.num_topics\n",
    "\n",
    "            neighbors = ast.literal_eval(components[1])\n",
    "        \n",
    "        # Subsequent iterations\n",
    "        else:\n",
    "            node = key\n",
    "            node_pr = line[0] # pagerank from previous iteration\n",
    "            topic_pr = line[1]\n",
    "            neighbors = line[2]\n",
    "\n",
    "        if neighbors:\n",
    "            # Factor in multiple links to the same page\n",
    "            num_out_links = sum(neighbors.values())\n",
    "            \n",
    "            # Output general pagerank mass contributions\n",
    "            for neighbor, num_links in neighbors.iteritems():\n",
    "                yield neighbor, (\"pr\", num_links * (node_pr / num_out_links))\n",
    "\n",
    "                # Output topic-specific pagerank mass\n",
    "                for i, pr in enumerate(topic_pr):\n",
    "                    yield neighbor, (\"pr\" + str(i), num_links * (pr / num_out_links))\n",
    "\n",
    "        else:\n",
    "            # Accumulate mass for dangling nodes\n",
    "            # This will only run for subsequent iterations\n",
    "            self.loss += node_pr\n",
    "\n",
    "            for i, pr in enumerate(topic_pr):\n",
    "                self.topic_loss[i] += pr\n",
    "\n",
    "        yield node, (\"node\", neighbors)\n",
    "\n",
    "    def pr_mapper_final(self):\n",
    "        yield self.LOSS_KEY, (self.loss, self.topic_loss)\n",
    "\n",
    "    \"\"\"\n",
    "    These next two functions are used by both steps to accumulate the lost PageRank\n",
    "    mass.\n",
    "    \"\"\"\n",
    "    def init_loss_accumulators(self):\n",
    "        self.loss = 0.0\n",
    "        self.topic_loss = [0.0] * self.options.num_topics\n",
    "\n",
    "    def process_loss_accumulators(self, values):\n",
    "        for loss, topic_loss in values:\n",
    "            self.loss += loss\n",
    "\n",
    "            for i, lost_pr in enumerate(topic_loss):\n",
    "                self.topic_loss[i] += lost_pr\n",
    "\n",
    "    def pr_reducer_init(self):\n",
    "        self.init_loss_accumulators()        \n",
    "\n",
    "    def pr_reducer(self, key, values):\n",
    "\n",
    "        if key == self.LOSS_KEY:\n",
    "            # Won't be emitted by first iteration, dangling node\n",
    "            # block below will\n",
    "            self.process_loss_accumulators(values)\n",
    "\n",
    "        else:\n",
    "            next_pr = 0.0\n",
    "            next_topic_pr = [0.0] * self.options.num_topics\n",
    "            node_data = None\n",
    "\n",
    "            for value in values:\n",
    "                record_type = value[0]\n",
    "\n",
    "                if record_type == \"pr\":\n",
    "                    next_pr += value[1]\n",
    "\n",
    "                elif record_type.startswith(\"pr\"):\n",
    "                    topic_index = int(record_type[2:])\n",
    "                    next_topic_pr[topic_index] += value[1]\n",
    "\n",
    "                elif record_type == \"node\":\n",
    "                    node_data = value\n",
    "\n",
    "\n",
    "            # Regular node\n",
    "            if node_data:\n",
    "                neighbors = node_data[1]\n",
    "                # Output the accumulated pagerank mass. Next step will do final calc w/ damping\n",
    "                yield key, (next_pr, next_topic_pr, neighbors)\n",
    "\n",
    "            # Dangling node, first iteration only will have \"pr\" records with mass contributions\n",
    "            # but will itself not have a \"node\" entry, i.e it does not have its own record in the\n",
    "            # original input file, it's only listed as a neighbor. So we ensure the dangling node\n",
    "            # has its own record, but with an empty list of neighbors.\n",
    "            else:\n",
    "                lost_mass = (1 / self.options.node_count) # initial mass for all nodes\n",
    "                self.loss += lost_mass\n",
    "                for i, pr in enumerate(self.topic_loss):\n",
    "                    self.topic_loss[i] = pr + lost_mass\n",
    "                \n",
    "                yield key, (next_pr, next_topic_pr, {})\n",
    "\n",
    "    def pr_reducer_final(self):\n",
    "        yield self.LOSS_KEY, (self.loss, self.topic_loss)\n",
    "\n",
    "    def update_mapper(self, key, value):\n",
    "        yield key, value\n",
    "\n",
    "    def update_reducer_init(self):\n",
    "        self.init_loss_accumulators()\n",
    "\n",
    "        self.total_pr = 0.0\n",
    "        self.total_topic_pr = [0.0] * self.options.num_topics\n",
    "\n",
    "        self.node_to_topic_map = {}\n",
    "        self.topic_size = defaultdict(float)\n",
    "\n",
    "        with open(self.options.topic_file, 'r') as topic_file:\n",
    "            for line in topic_file:\n",
    "                (node, topic) = line.strip().split('\\t')\n",
    "\n",
    "                topic = int(topic)\n",
    "                self.node_to_topic_map[node] = topic\n",
    "                self.topic_size[topic] += 1\n",
    "\n",
    "        self.topic_order = map(int, sorted(self.node_to_topic_map.keys(), key=int))\n",
    "\n",
    "    def update_reducer(self, key, values):\n",
    "        \n",
    "        if key == self.LOSS_KEY:\n",
    "            self.process_loss_accumulators(values)        \n",
    "        else:\n",
    "            (general_pr, topic_pageranks, neighbors) = values.next()\n",
    "\n",
    "            node_count = self.options.node_count\n",
    "            damping = self.options.damping\n",
    "\n",
    "            # Calculate pagerank for the topics\n",
    "            for i, next_pr in enumerate(topic_pageranks):\n",
    "                topic = self.topic_order[i]\n",
    "                in_topic = self.node_to_topic_map[key] == topic\n",
    "                beta = self.options.beta\n",
    "                loss = self.topic_loss[i]\n",
    "                \n",
    "                if in_topic:\n",
    "                    teleportation = beta * (1 / self.topic_size[topic])\n",
    "                else:\n",
    "                    teleportation = (1 - beta) * (1 / (node_count - self.topic_size[topic]))\n",
    "\n",
    "                final_pr = (1 - damping) * teleportation + \\\n",
    "                            damping * ( (loss / node_count) + next_pr)\n",
    "                topic_pageranks[i] = final_pr\n",
    "\n",
    "            # Do the general non-topic pagerank calc. Loss/pagerank is maintained\n",
    "            # separately from the topic array/map system to more easily cope with\n",
    "            # base-1 topics in randNet, but base-0 topics for wikipedia set.\n",
    "            loss = self.loss\n",
    "\n",
    "            final_pr = (1 - damping) * (1 / node_count) + \\\n",
    "                        damping * ( (loss / node_count) + general_pr)\n",
    "\n",
    "            if self.options.debug:\n",
    "                self.total_pr += final_pr\n",
    "                for i, pr in enumerate(topic_pageranks):\n",
    "                    self.total_topic_pr[i] += pr\n",
    "\n",
    "            yield key, (final_pr, topic_pageranks, neighbors)\n",
    "\n",
    "    def update_reducer_final(self):\n",
    "        if self.options.debug:\n",
    "            print \"Loss: %f; Total Pr: %f, Topic PR: %s\" % (self.loss, self.total_pr, str(self.total_topic_pr))\n",
    "        \n",
    "    def steps(self):\n",
    "\n",
    "        summary_conf = {\n",
    "            'mapred.reduce.tasks': '1'\n",
    "        }\n",
    "\n",
    "        return [\n",
    "            MRStep(mapper_init=self.pr_mapper_init, mapper=self.pr_mapper, mapper_final=self.pr_mapper_final, \\\n",
    "                  reducer_init=self.pr_reducer_init, reducer = self.pr_reducer, reducer_final=self.pr_reducer_final),\n",
    "            MRStep(mapper=self.update_mapper, reducer=self.update_reducer, jobconf=summary_conf,\\\n",
    "                reducer_init=self.update_reducer_init, reducer_final=self.update_reducer_final)\n",
    "        ] * self.options.iterations\n",
    "\n",
    "if __name__ == '__main__':\n",
    "    PageRank.run()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "!python TopicPageRank.py --iterations 50 --damping \".85\" --num-topics 10 --node-count 100 \\\n",
    "--topic-file data/randNet_topics.txt --beta \".99\" data/randNet.txt > randNet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\"1\"\t[0.007873143501279354, [0.006552240111082188, 0.00634945180596396, 0.005212904846035831, 0.00676498741106525, 0.006493061250751263, 0.005847867902675189, 0.008030658548099444, 0.005466175975529963, 0.005605253155922045, 0.018998088144569804], {\"11\": 1, \"27\": 1, \"46\": 1, \"47\": 1, \"35\": 1, \"63\": 1, \"89\": 1, \"5\": 1}]\r\n",
      "\"10\"\t[0.011115047597052393, [0.018565527659378488, 0.008473360067136343, 0.01059253762478958, 0.00926149012415711, 0.009453018233792363, 0.01037203905165704, 0.00997003291561642, 0.010127214686835154, 0.00946113378557876, 0.008989150910780716], {\"38\": 1, \"49\": 1, \"46\": 1, \"47\": 1, \"28\": 1, \"53\": 1, \"61\": 1, \"89\": 1, \"64\": 1, \"80\": 1}]\r\n",
      "\"100\"\t[0.015376515070515785, [0.013137293483664292, 0.01472557845487658, 0.014892904535859198, 0.013253554450432569, 0.01674152091638443, 0.011724384381648887, 0.011952593611148895, 0.03286601507341609, 0.01329311881302321, 0.013403739875762518], {\"10\": 1, \"39\": 1, \"48\": 1, \"33\": 1, \"51\": 1, \"53\": 1, \"52\": 1, \"67\": 1}]\r\n",
      "\"11\"\t[0.009341776412699586, [0.008122262742021864, 0.008273925491453832, 0.006635748404099196, 0.00867394191547394, 0.006860159658086346, 0.03133966220222701, 0.007815538114046231, 0.007368579328700516, 0.009284037555838618, 0.008188664424556492], {\"39\": 1, \"12\": 1, \"21\": 1, \"55\": 1, \"43\": 1, \"46\": 1, \"89\": 1, \"7\": 1, \"84\": 1}]\r\n",
      "\"12\"\t[0.009678112703601907, [0.008243712705804365, 0.02688896411823312, 0.006400134643728505, 0.008929034873403277, 0.006350833167788286, 0.01142586031532957, 0.008202868403468396, 0.006945680352069378, 0.006765747553539189, 0.010109294230579786], {\"26\": 1, \"59\": 1, \"16\": 1, \"33\": 1, \"29\": 1, \"99\": 1, \"88\": 1, \"2\": 1, \"100\": 1, \"95\": 1}]\r\n",
      "\"13\"\t[0.01317823212124274, [0.011589656706398341, 0.012952345927144155, 0.013315772635059182, 0.010390019189834052, 0.010731150424696722, 0.03457155767187721, 0.01309626368285583, 0.011691998209092285, 0.009428810022771341, 0.0130127860772345], {\"54\": 1, \"52\": 1, \"77\": 1, \"65\": 1, \"92\": 1, \"85\": 1}]\r\n",
      "\"14\"\t[0.0098582880297808, [0.009638587282091033, 0.008284570684299912, 0.007458093543797289, 0.007683038893532848, 0.008314573190035151, 0.006968382525822248, 0.006795758423087716, 0.010118067628805466, 0.029495186887110746, 0.008884815631746144], {\"11\": 1, \"34\": 1, \"3\": 1, \"64\": 1, \"65\": 1, \"91\": 1, \"86\": 1}]\r\n",
      "\"15\"\t[0.01635632305602751, [0.01366301699592087, 0.013695868332907299, 0.031529080815593365, 0.01675216795369466, 0.016341356373975845, 0.017242596786219998, 0.015936482639623935, 0.011908490701162847, 0.013608803232421253, 0.014997614998378626], {\"13\": 1, \"21\": 1, \"49\": 1, \"45\": 1, \"37\": 1, \"99\": 1, \"74\": 1, \"92\": 1, \"87\": 1}]\r\n",
      "\"16\"\t[0.007682872962050437, [0.014879821301546914, 0.0066590476079980195, 0.006065061903552272, 0.006904849192320102, 0.006076327027914411, 0.005968160601281155, 0.005213184575169975, 0.00697760409544042, 0.004783212027182009, 0.00655611426345631], {\"24\": 1, \"10\": 1, \"58\": 1, \"57\": 1, \"65\": 1, \"9\": 1, \"85\": 1}]\r\n",
      "\"17\"\t[0.013065806979675947, [0.009652810935926954, 0.015158688037819222, 0.009382112248811754, 0.011943259791472885, 0.010807562340733757, 0.010022061890798087, 0.015580598429983125, 0.011300468515212546, 0.01193252267651535, 0.023590506777885652], {\"27\": 1, \"48\": 1, \"35\": 1, \"61\": 1, \"75\": 1, \"71\": 1, \"70\": 1, \"92\": 1, \"78\": 1}]\r\n"
     ]
    }
   ],
   "source": [
    "!head randNet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%%writefile TopicResultsProcessor.py\n",
    "#!/usr/bin/python\n",
    "\n",
    "import argparse\n",
    "import ast\n",
    "\n",
    "from TopN import TopN\n",
    "\n",
    "if __name__ == \"__main__\":\n",
    "\n",
    "    parser = argparse.ArgumentParser()\n",
    "    parser.add_argument('file', type=argparse.FileType('r'))\n",
    "    parser.add_argument('n', type=int)\n",
    "    parser.add_argument('offset', type=int, default=0)\n",
    "\n",
    "    args = parser.parse_args()\n",
    "\n",
    "    top = TopN(args.n)\n",
    "\n",
    "    topic_top = [None] * 10\n",
    "    for i in range(0, 10):\n",
    "        topic_top[i] = TopN(args.n)\n",
    "\n",
    "    for line in args.file:\n",
    "        node, values = line.strip().split('\\t')\n",
    "\n",
    "        node = node.translate(None, '\"')\n",
    "        values = ast.literal_eval(values)\n",
    "\n",
    "        pr = values[0]\n",
    "\n",
    "        topic_pr = values[1]\n",
    "\n",
    "        top.insert((pr, node))\n",
    "\n",
    "        for i, pr in enumerate(topic_pr):\n",
    "            topic_top[i].insert((pr, node))\n",
    "\n",
    "    print \"Vanilla PageRank\"\n",
    "    for result in top.get():\n",
    "        print result[1] + \"\\t\" + str(result[0])\n",
    "\n",
    "    print \"\"\n",
    "    for i in range(0, 10):\n",
    "        print \"PageRank for Topic \" + str(i + args.offset)\n",
    "        for result in topic_top[i].get():\n",
    "            print result[1] + \"\\t\" + str(result[0])\n",
    "        print \"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "!python TopicResultsProcessor.py randNet 10 1 > randNet_topic_results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Vanilla PageRank\r\n",
      "15\t0.016356323056\r\n",
      "74\t0.0159691890062\r\n",
      "63\t0.0157709405122\r\n",
      "100\t0.0153765150705\r\n",
      "85\t0.0151785240576\r\n",
      "9\t0.0150325119447\r\n",
      "58\t0.014828168909\r\n",
      "71\t0.0144908657999\r\n",
      "61\t0.0144070472986\r\n",
      "52\t0.014311039396\r\n",
      "\r\n",
      "PageRank for Topic 1\r\n",
      "32\t0.0206458512866\r\n",
      "77\t0.0205475762655\r\n",
      "52\t0.0197543048268\r\n",
      "92\t0.0195292443227\r\n",
      "10\t0.0185655276594\r\n",
      "27\t0.0185225299503\r\n",
      "85\t0.0178405128139\r\n",
      "98\t0.0176924092197\r\n",
      "46\t0.0175141303283\r\n",
      "74\t0.0160281417572\r\n",
      "\r\n",
      "PageRank for Topic 2\r\n",
      "58\t0.0308474697246\r\n",
      "71\t0.0296652160451\r\n",
      "9\t0.0292968277252\r\n",
      "73\t0.0289148327104\r\n",
      "12\t0.0268889641182\r\n",
      "59\t0.0257996585334\r\n",
      "75\t0.0248495901988\r\n",
      "82\t0.0228582190028\r\n",
      "52\t0.0163221073853\r\n",
      "17\t0.0151586880378\r\n",
      "\r\n",
      "PageRank for Topic 3\r\n",
      "15\t0.0315290808156\r\n",
      "70\t0.0270765896211\r\n",
      "86\t0.0265279680485\r\n",
      "91\t0.024463316203\r\n",
      "66\t0.0241485260743\r\n",
      "2\t0.0237050889109\r\n",
      "31\t0.0227671069051\r\n",
      "40\t0.0221785039384\r\n",
      "20\t0.0197450498654\r\n",
      "74\t0.0158999904173\r\n",
      "\r\n",
      "PageRank for Topic 4\r\n",
      "63\t0.0262020476066\r\n",
      "83\t0.0217600597112\r\n",
      "65\t0.0206237950517\r\n",
      "78\t0.0202100877919\r\n",
      "41\t0.0199084505097\r\n",
      "84\t0.0195198810945\r\n",
      "79\t0.018428641499\r\n",
      "38\t0.0175154143581\r\n",
      "15\t0.0167521679537\r\n",
      "72\t0.0166947212551\r\n",
      "\r\n",
      "PageRank for Topic 5\r\n",
      "99\t0.0289632796232\r\n",
      "90\t0.0283449752279\r\n",
      "88\t0.0271687028902\r\n",
      "51\t0.0268308059561\r\n",
      "45\t0.0255532965188\r\n",
      "5\t0.0239199680924\r\n",
      "34\t0.0239097236729\r\n",
      "4\t0.0233632611864\r\n",
      "80\t0.0228395918386\r\n",
      "100\t0.0167415209164\r\n",
      "\r\n",
      "PageRank for Topic 6\r\n",
      "13\t0.0345715576719\r\n",
      "56\t0.0328539108785\r\n",
      "37\t0.0317771730682\r\n",
      "11\t0.0313396622022\r\n",
      "69\t0.0301207775979\r\n",
      "23\t0.0283489020801\r\n",
      "15\t0.0172425967862\r\n",
      "85\t0.0169894360702\r\n",
      "52\t0.0166021385904\r\n",
      "74\t0.0154610802586\r\n",
      "\r\n",
      "PageRank for Topic 7\r\n",
      "85\t0.0267942407048\r\n",
      "25\t0.0266087818235\r\n",
      "28\t0.02482723761\r\n",
      "53\t0.0247676602638\r\n",
      "35\t0.0242008493561\r\n",
      "97\t0.0233950349816\r\n",
      "47\t0.0228650194907\r\n",
      "55\t0.0225608364195\r\n",
      "30\t0.0221348262738\r\n",
      "50\t0.0200890700894\r\n",
      "\r\n",
      "PageRank for Topic 8\r\n",
      "100\t0.0328660150734\r\n",
      "61\t0.0278585573034\r\n",
      "39\t0.027195517001\r\n",
      "8\t0.0271530914456\r\n",
      "62\t0.0253468181687\r\n",
      "87\t0.025298398627\r\n",
      "6\t0.0235068261757\r\n",
      "54\t0.0228917422812\r\n",
      "18\t0.0206226544265\r\n",
      "9\t0.0153797664428\r\n",
      "\r\n",
      "PageRank for Topic 9\r\n",
      "94\t0.0301988683317\r\n",
      "14\t0.0294951868871\r\n",
      "42\t0.0291978782081\r\n",
      "21\t0.0283995917011\r\n",
      "57\t0.0274618061283\r\n",
      "96\t0.0262609478642\r\n",
      "24\t0.0257730811357\r\n",
      "63\t0.0171614985751\r\n",
      "61\t0.0163773404251\r\n",
      "74\t0.0142777214983\r\n",
      "\r\n",
      "PageRank for Topic 10\r\n",
      "74\t0.0263318213012\r\n",
      "17\t0.0235905067779\r\n",
      "49\t0.0235742624199\r\n",
      "95\t0.0206293139069\r\n",
      "7\t0.0199138056816\r\n",
      "43\t0.0193641716344\r\n",
      "68\t0.0190382686762\r\n",
      "48\t0.0190050464152\r\n",
      "1\t0.0189980881446\r\n",
      "3\t0.018639552329\r\n",
      "\r\n"
     ]
    }
   ],
   "source": [
    "!cat randNet_topic_results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*The results look fairly good, for the most part the top results for a topic are unique. A few documents like '15' show up multiple times, but as that was the top-ranked page for regular page-rank, that's not too surprising.*"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## HW 9.5: Applying topic-specific PageRank to Wikipedia\n",
    "\n",
    "**Here you will apply your topic-specific PageRank implementation to Wikipedia, defining topics (very arbitrarily) for each page by the length (number of characters) of the name of the article mod 10, so that there are 10 topics. Once again, print out the top ten ranking nodes and their topics for each of the 11 versions, and comment on your result. Assume a teleportation factor of 0.15 in all your analyses.**  \n",
    "\n",
    "*Step 0: First, we need to generate the topics list using the indices.txt file from the wikipedia dataset*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%%writefile CreateWikiTopics.py\n",
    "#!/usr/bin/python\n",
    "\n",
    "import sys\n",
    "\n",
    "with open('data/wiki_topics.txt', 'w') as out:\n",
    "    with open('data/indices.txt', 'r') as wiki_sites:\n",
    "\n",
    "        for line in wiki_sites:\n",
    "            name, index, _ = line.strip().split('\\t', 2)\n",
    "\n",
    "            out.write(\"%s\\t%d\\n\" % (index, len(name) % 10))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "!python CreateWikiTopics.py"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1\t1\n",
      "2\t4\n",
      "3\t9\n",
      "4\t8\n",
      "5\t3\n",
      "6\t1\n",
      "7\t0\n",
      "8\t9\n",
      "9\t7\n",
      "10\t1\n",
      "11\t5\n",
      "12\t3\n",
      "13\t8\n",
      "14\t8\n",
      "15\t4\n",
      "16\t2\n",
      "17\t8\n",
      "18\t2\n",
      "19\t3\n",
      "20\t1\n",
      "21\t8\n",
      "22\t2\n",
      "23\t8\n",
      "24\t3\n",
      "25\t1\n",
      "\n",
      "...\n",
      "\n",
      "15192253\t3\n",
      "15192254\t5\n",
      "15192255\t9\n",
      "15192256\t2\n",
      "15192257\t9\n",
      "15192258\t5\n",
      "15192259\t7\n",
      "15192260\t3\n",
      "15192261\t1\n",
      "15192262\t1\n",
      "15192263\t3\n",
      "15192264\t1\n",
      "15192265\t5\n",
      "15192266\t4\n",
      "15192267\t6\n",
      "15192268\t1\n",
      "15192269\t2\n",
      "15192270\t5\n",
      "15192271\t9\n",
      "15192272\t9\n",
      "15192273\t1\n",
      "15192274\t7\n",
      "15192275\t6\n",
      "15192276\t9\n",
      "15192277\t2\n"
     ]
    }
   ],
   "source": [
    "!head -n 25 data/wiki_topics.txt\n",
    "!echo \"\\n...\\n\"\n",
    "!tail -n 25 data/wiki_topics.txt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "!# Now, run for 50 iterations\n",
    "!python TopicPageRank.py -r emr --num-ec2-instances 12 --ec2-instance-type 'm1.large' --no-output \\\n",
    "--output-dir s3://stevenschaefer-mids-mlscale/topic-pr-10 --iterations 10 --node-count 15192277 --damping \".85\" --beta \".99\" --num-topics 10 \\\n",
    "    --topic-file data\\wiki_topics.txt s3://ucb-mids-mls-networks/wikipedia/all-pages-indexed-out.txt &> out/topic_pr_10_out"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "using configs in /home/steven/.mrjob.conf\r\n",
      "Got unexpected opts from /home/steven/.mrjob.conf: ec2_instace_type\r\n",
      "inferring aws_region from scratch bucket's region (us-west-1)\r\n",
      "creating tmp directory /tmp/TopicPageRank.steven.20151101.234922.802695\r\n",
      "writing master bootstrap script to /tmp/TopicPageRank.steven.20151101.234922.802695/b.py\r\n",
      "Copying non-input files into s3://stevenschaefer-mids-mlscale/tmp/TopicPageRank.steven.20151101.234922.802695/files/\r\n",
      "Waiting 5.0s for S3 eventual consistency\r\n",
      "Creating Elastic MapReduce job flow\r\n",
      "Job flow created with ID: j-1WWC6STQ0A33O\r\n",
      "Created new job flow j-1WWC6STQ0A33O\r\n",
      "Job launched 31.2s ago, status STARTING: Provisioning Amazon EC2 capacity\r\n",
      "Job launched 62.2s ago, status STARTING: Provisioning Amazon EC2 capacity\r\n",
      "Job launched 93.3s ago, status STARTING: Provisioning Amazon EC2 capacity\r\n",
      "Job launched 124.4s ago, status STARTING: Provisioning Amazon EC2 capacity\r\n",
      "Job launched 155.5s ago, status STARTING: Configuring cluster software\r\n",
      "Job launched 186.5s ago, status STARTING: Configuring cluster software\r\n",
      "Job launched 217.6s ago, status BOOTSTRAPPING: Running bootstrap actions\r\n",
      "Job launched 248.7s ago, status BOOTSTRAPPING: Running bootstrap actions\r\n",
      "Job launched 279.7s ago, status RUNNING: Running step (TopicPageRank.steven.20151101.234922.802695: Step 1 of 20)\r\n",
      "Job launched 310.8s ago, status RUNNING: Running step (TopicPageRank.steven.20151101.234922.802695: Step 1 of 20)\r\n",
      "Job launched 341.8s ago, status RUNNING: Running step (TopicPageRank.steven.20151101.234922.802695: Step 1 of 20)\r\n",
      "Job launched 372.9s ago, status RUNNING: Running step (TopicPageRank.steven.20151101.234922.802695: Step 1 of 20)\r\n",
      "Job launched 404.0s ago, status RUNNING: Running step (TopicPageRank.steven.20151101.234922.802695: Step 1 of 20)\r\n",
      "Job launched 435.0s ago, status RUNNING: Running step (TopicPageRank.steven.20151101.234922.802695: Step 1 of 20)\r\n",
      "Job launched 466.1s ago, status RUNNING: Running step (TopicPageRank.steven.20151101.234922.802695: Step 1 of 20)\r\n",
      "Job launched 497.1s ago, status RUNNING: Running step (TopicPageRank.steven.20151101.234922.802695: Step 1 of 20)\r\n",
      "Job launched 528.3s ago, status RUNNING: Running step (TopicPageRank.steven.20151101.234922.802695: Step 1 of 20)\r\n",
      "Job launched 559.4s ago, status RUNNING: Running step (TopicPageRank.steven.20151101.234922.802695: Step 1 of 20)\r\n",
      "Job launched 590.5s ago, status RUNNING: Running step (TopicPageRank.steven.20151101.234922.802695: Step 1 of 20)\r\n",
      "Job launched 621.5s ago, status RUNNING: Running step (TopicPageRank.steven.20151101.234922.802695: Step 1 of 20)\r\n"
     ]
    }
   ],
   "source": [
    "!head -n 30 out/topic_pr_10_out"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "!python TopicResultsProcessor.py results/topic_pr_10 10 1 > wiki_topic_results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Vanilla PageRank\r\n",
      "13455888\t0.00154472471298\r\n",
      "4695850\t0.000671024071891\r\n",
      "5051368\t0.000598385680975\r\n",
      "1184351\t0.000598207353647\r\n",
      "2437837\t0.000462492892894\r\n",
      "6076759\t0.000455094006414\r\n",
      "4196067\t0.000442377888837\r\n",
      "13425865\t0.000441553517143\r\n",
      "6172466\t0.000422400200185\r\n",
      "1384888\t0.000401289560407\r\n",
      "\r\n",
      "PageRank for Topic 0\r\n",
      "13455888\t0.00155597693995\r\n",
      "4695850\t0.000653511684341\r\n",
      "1184351\t0.00062144726401\r\n",
      "5051368\t0.000598729921087\r\n",
      "2437837\t0.000467148417263\r\n",
      "6076759\t0.000455271103185\r\n",
      "4196067\t0.000447795248159\r\n",
      "13425865\t0.000443956022325\r\n",
      "1384888\t0.000419122567486\r\n",
      "6172466\t0.000413101357393\r\n",
      "\r\n",
      "PageRank for Topic 1\r\n",
      "13455888\t0.00156270329548\r\n",
      "4695850\t0.000651521978627\r\n",
      "1184351\t0.000592438642617\r\n",
      "5051368\t0.000589728378876\r\n",
      "2437837\t0.000469852987697\r\n",
      "6076759\t0.000452358745013\r\n",
      "4196067\t0.000451878141503\r\n",
      "13425865\t0.000448287019656\r\n",
      "6172466\t0.000404927876503\r\n",
      "14112583\t0.000396008626215\r\n",
      "\r\n",
      "PageRank for Topic 2\r\n",
      "13455888\t0.00154996571911\r\n",
      "4695850\t0.00065138950481\r\n",
      "10527224\t0.000630428473371\r\n",
      "5051368\t0.00058469645903\r\n",
      "5490435\t0.000571540951524\r\n",
      "1184351\t0.000566554068438\r\n",
      "2437837\t0.000470617359045\r\n",
      "4196067\t0.000454028353634\r\n",
      "6076759\t0.000450378016278\r\n",
      "13425865\t0.000445499171073\r\n",
      "\r\n",
      "PageRank for Topic 3\r\n",
      "13455888\t0.0015633112645\r\n",
      "4695850\t0.000659839942594\r\n",
      "5051368\t0.000584555372333\r\n",
      "1184351\t0.000555518430254\r\n",
      "2437837\t0.000476514849715\r\n",
      "4196067\t0.000450076836207\r\n",
      "6076759\t0.000443362838036\r\n",
      "13425865\t0.000443272300637\r\n",
      "6172466\t0.000399667910038\r\n",
      "14112583\t0.000395593125506\r\n",
      "\r\n",
      "PageRank for Topic 4\r\n",
      "13455888\t0.00154227005483\r\n",
      "4695850\t0.000676835868157\r\n",
      "5051368\t0.000594548687625\r\n",
      "1184351\t0.000561418965801\r\n",
      "2437837\t0.000461638735085\r\n",
      "6076759\t0.0004510245051\r\n",
      "4196067\t0.000444396036995\r\n",
      "13425865\t0.000440800651898\r\n",
      "6172466\t0.000421513019758\r\n",
      "14112583\t0.000396427118629\r\n",
      "\r\n",
      "PageRank for Topic 5\r\n",
      "13455888\t0.00153465233762\r\n",
      "4695850\t0.000693814294346\r\n",
      "5051368\t0.000599383489649\r\n",
      "1184351\t0.000578283714422\r\n",
      "6076759\t0.000457979283975\r\n",
      "2437837\t0.000456039210725\r\n",
      "13425865\t0.000438981240886\r\n",
      "6172466\t0.000438492515728\r\n",
      "4196067\t0.000436291100312\r\n",
      "14112583\t0.000395056511415\r\n",
      "\r\n",
      "PageRank for Topic 6\r\n",
      "13455888\t0.00153633369543\r\n",
      "4695850\t0.000694559130644\r\n",
      "5051368\t0.000604781384756\r\n",
      "1184351\t0.000601043884179\r\n",
      "6076759\t0.000457762281977\r\n",
      "2437837\t0.00045604846945\r\n",
      "6172466\t0.000448000123214\r\n",
      "13425865\t0.000435565658504\r\n",
      "4196067\t0.000431198367894\r\n",
      "1384888\t0.000403768272473\r\n",
      "\r\n",
      "PageRank for Topic 7\r\n",
      "13455888\t0.00152287850631\r\n",
      "4695850\t0.000693465558222\r\n",
      "1184351\t0.000629191103665\r\n",
      "5051368\t0.000611046922676\r\n",
      "6076759\t0.000464581189065\r\n",
      "2437837\t0.000450362176992\r\n",
      "6172466\t0.000446370632704\r\n",
      "13425865\t0.000436840453355\r\n",
      "4196067\t0.000430664040483\r\n",
      "1384888\t0.000426190611508\r\n",
      "\r\n",
      "PageRank for Topic 8\r\n",
      "13455888\t0.00152786697208\r\n",
      "4695850\t0.000677151115862\r\n",
      "1184351\t0.000644272964353\r\n",
      "5051368\t0.000612212082995\r\n",
      "6076759\t0.000461287822482\r\n",
      "2437837\t0.000452131388443\r\n",
      "13425865\t0.000438260294369\r\n",
      "1384888\t0.000438106501603\r\n",
      "6172466\t0.000434870869559\r\n",
      "4196067\t0.000433959015036\r\n",
      "\r\n",
      "PageRank for Topic 9\r\n",
      "13455888\t0.000384923968726\r\n",
      "4695850\t0.000161603029958\r\n",
      "5051368\t0.000144438050829\r\n",
      "13425865\t0.000119041011506\r\n",
      "1184351\t0.00011834015492\r\n",
      "2437837\t0.000115256592768\r\n",
      "12836211\t0.000109643039586\r\n",
      "4196067\t0.000107120550013\r\n",
      "6076759\t0.000106928709142\r\n",
      "14112583\t0.000105993258998\r\n",
      "\r\n"
     ]
    }
   ],
   "source": [
    "!cat wiki_topic_results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**comment on your result**  \n",
    "  \n",
    "*The wikipedia topics were created mostly randomly, the length of the title mod 10. Assuming there's no sort of correlation between the length of the title and other pages, we'd expect the results of the topic page rank to include a lot of the same pages, and we see that's the case here. The order changes, but overall there are a lot of repeat pages. The page rank values themselves are also very similar across topics and the general pagerank results. It's especially noteworthy that the top page of the general pagerank algorithm is at the top of all the topics: 13455888.*  \n",
    "  \n",
    "*What is this top page?*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "United States\t13455888\t555979\t1448\r\n"
     ]
    }
   ],
   "source": [
    "!grep 13455888 data/indices.txt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Final Note about Instance Sizing\n",
    "\n",
    "For counting the nodes of the wikipedia data set 3 m1.smalls were used. 4 m1.mediums were used for the 10 iteration PageRank job. The runtime of the steps for this job, combined with the runtimes other groups were experiencing informed the decision to use around 11 m1.larges for the 50 iteration job with the goal of total runtime being around 24 hours. This job ran in 25 hours. For the 10 iteration topic page rank job, a similar goal of 24 hours was also desired. A couple of 2 hour runs using m1.mediums where the first step had not completed indicated m1.large was also necessary despite fewer iterations, along with around 11 instances. This job completed in around 22 hours."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}