tutorials/BigQuery/SQL Query Composition.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# SQL Query Composition\n",
    "\n",
    "This notebook builds on previous notebooks that illustrate relatively simple SQL queries (queries with a single `SELECT` statement to retrieve and analyze BigQuery data). This notebooks illustrates how Google BigQuery allows you to use nested queries to create complex queries that successively filter or slice/dice results or perform joins in a single query issued to the service. Constructing these complex queries can often be made easier when done step-by-step.\n",
    "\n",
    "This notebook introduces a pattern for developing queries independently, and then stitching them together to construct a complete solution."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Scenario\n",
    "\n",
    "As in other sample notebooks, this notebook will build on the HTTP logs dataset. The goal is to build a latency distribution curve of requests. This helps visualize the latency pattern of a web server.\n",
    "\n",
    "**Preface on Notebook Content**\n",
    "\n",
    "Although query composition through step-by-step building is intrinsically iterative, this notebook solution represents a snapshot of the final state. For illustrarive purposes, each step of the iteration is broken out into separate sections, but it's unlikely you'll do the same in a real-task scenario."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import datalab.bigquery as bq"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Step 1 - Inspecting Latencies\n",
    "\n",
    "The first step is to get an initial slice of the data we're working with with a very basic query."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "    <div class=\"bqtv\" id=\"1_146281575355\"><table><tr><th>latency</th></tr><tr><td>427</td></tr><tr><td>331</td></tr><tr><td>317</td></tr><tr><td>329</td></tr><tr><td>75</td></tr><tr><td>231</td></tr><tr><td>197</td></tr><tr><td>293</td></tr><tr><td>187</td></tr><tr><td>226</td></tr></table></div>\n",
       "    <br />(rows: 10, time: 1.0s,    10MB processed, job: job_vh_2XP65gHhyVVB3pF6Ueiy_CXo)<br />\n",
       "    <script>\n",
       "      require(['datalab/charting', 'datalab/element!1_146281575355', 'base/js/events',\n",
       "          'datalab/style!/nbextensions/gcpdatalab/charting.css'],\n",
       "        function(charts, dom, events) {\n",
       "          charts.render('gcharts', dom, events, 'table', [], {\"rows\": [{\"c\": [{\"v\": 427}]}, {\"c\": [{\"v\": 331}]}, {\"c\": [{\"v\": 317}]}, {\"c\": [{\"v\": 329}]}, {\"c\": [{\"v\": 75}]}, {\"c\": [{\"v\": 231}]}, {\"c\": [{\"v\": 197}]}, {\"c\": [{\"v\": 293}]}, {\"c\": [{\"v\": 187}]}, {\"c\": [{\"v\": 226}]}], \"cols\": [{\"type\": \"number\", \"id\": \"latency\", \"label\": \"latency\"}]},\n",
       "            {\n",
       "              pageSize: 25,\n",
       "              cssClassNames:  {\n",
       "                tableRow: 'gchart-table-row',\n",
       "                headerRow: 'gchart-table-headerrow',\n",
       "                oddTableRow: 'gchart-table-oddrow',\n",
       "                selectedTableRow: 'gchart-table-selectedrow',\n",
       "                hoverTableRow: 'gchart-table-hoverrow',\n",
       "                tableCell: 'gchart-table-cell',\n",
       "                headerCell: 'gchart-table-headercell',\n",
       "                rowNumberCell: 'gchart-table-rownumcell'\n",
       "              }\n",
       "            },\n",
       "            {source_index: 0, fields: 'latency'},\n",
       "            0,\n",
       "            10);\n",
       "        }\n",
       "      );\n",
       "    </script>\n",
       "  "
      ],
      "text/plain": [
       "QueryResultsTable job_vh_2XP65gHhyVVB3pF6Ueiy_CXo"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%sql\n",
    "SELECT latency FROM [cloud-datalab-samples:httplogs.logs_20140615] WHERE endpoint = 'Recent' LIMIT 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "    <div class=\"bqtv\" id=\"2_146281575587\"><table><tr><th>min</th><th>max</th><th>count</th></tr><tr><td>2</td><td>18715</td><td>734</td></tr></table></div>\n",
       "    <br />(rows: 1, time: 0.7s,    10MB processed, job: job_hWOaBTfTzK_mzXtF8o8XNdQynpY)<br />\n",
       "    <script>\n",
       "      require(['datalab/charting', 'datalab/element!2_146281575587', 'base/js/events',\n",
       "          'datalab/style!/nbextensions/gcpdatalab/charting.css'],\n",
       "        function(charts, dom, events) {\n",
       "          charts.render('gcharts', dom, events, 'table', [], {\"rows\": [{\"c\": [{\"v\": 2}, {\"v\": 18715}, {\"v\": 734}]}], \"cols\": [{\"type\": \"number\", \"id\": \"min\", \"label\": \"min\"}, {\"type\": \"number\", \"id\": \"max\", \"label\": \"max\"}, {\"type\": \"number\", \"id\": \"count\", \"label\": \"count\"}]},\n",
       "            {\n",
       "              pageSize: 25,\n",
       "              cssClassNames:  {\n",
       "                tableRow: 'gchart-table-row',\n",
       "                headerRow: 'gchart-table-headerrow',\n",
       "                oddTableRow: 'gchart-table-oddrow',\n",
       "                selectedTableRow: 'gchart-table-selectedrow',\n",
       "                hoverTableRow: 'gchart-table-hoverrow',\n",
       "                tableCell: 'gchart-table-cell',\n",
       "                headerCell: 'gchart-table-headercell',\n",
       "                rowNumberCell: 'gchart-table-rownumcell'\n",
       "              }\n",
       "            },\n",
       "            {source_index: 1, fields: 'min,max,count'},\n",
       "            0,\n",
       "            1);\n",
       "        }\n",
       "      );\n",
       "    </script>\n",
       "  "
      ],
      "text/plain": [
       "QueryResultsTable job_hWOaBTfTzK_mzXtF8o8XNdQynpY"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%sql\n",
    "SELECT MIN(latency) AS min, MAX(latency) AS max, COUNT(latency) AS count\n",
    "FROM [cloud-datalab-samples:httplogs.logs_20140615]\n",
    "WHERE endpoint = 'Recent'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Step 2 - Bucketing Latencies\n",
    "\n",
    "Clearly, latencies are discrete numbers (milliseconds) over a wide range. One way to deal with this is to bucket them into groups."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "%%sql --module latency\n",
    "SELECT INTEGER(latency / 25) * 25 AS latency, COUNT(latency) AS requests\n",
    "FROM [cloud-datalab-samples:httplogs.logs_20140615]\n",
    "WHERE endpoint = 'Recent'\n",
    "GROUP BY latency\n",
    "ORDER BY latency"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "    <div class=\"bqtv\" id=\"3_146281575852\"><table><tr><th>latency</th><th>requests</th></tr><tr><td>0</td><td>66</td></tr><tr><td>25</td><td>4</td></tr><tr><td>50</td><td>12</td></tr><tr><td>75</td><td>45</td></tr><tr><td>100</td><td>28</td></tr><tr><td>125</td><td>9</td></tr><tr><td>150</td><td>4</td></tr><tr><td>175</td><td>76</td></tr><tr><td>200</td><td>100</td></tr><tr><td>225</td><td>104</td></tr><tr><td>250</td><td>80</td></tr><tr><td>275</td><td>46</td></tr><tr><td>300</td><td>40</td></tr><tr><td>325</td><td>46</td></tr><tr><td>350</td><td>26</td></tr><tr><td>375</td><td>26</td></tr><tr><td>400</td><td>9</td></tr><tr><td>425</td><td>4</td></tr><tr><td>450</td><td>4</td></tr><tr><td>500</td><td>2</td></tr><tr><td>550</td><td>1</td></tr><tr><td>600</td><td>1</td></tr><tr><td>18700</td><td>1</td></tr></table></div>\n",
       "    <br />(rows: 23, time: 0.8s,    10MB processed, job: job_01iap04gpHTfDEnm3tO29csZJJw)<br />\n",
       "    <script>\n",
       "      require(['datalab/charting', 'datalab/element!3_146281575852', 'base/js/events',\n",
       "          'datalab/style!/nbextensions/gcpdatalab/charting.css'],\n",
       "        function(charts, dom, events) {\n",
       "          charts.render('gcharts', dom, events, 'table', [], {\"rows\": [{\"c\": [{\"v\": 0}, {\"v\": 66}]}, {\"c\": [{\"v\": 25}, {\"v\": 4}]}, {\"c\": [{\"v\": 50}, {\"v\": 12}]}, {\"c\": [{\"v\": 75}, {\"v\": 45}]}, {\"c\": [{\"v\": 100}, {\"v\": 28}]}, {\"c\": [{\"v\": 125}, {\"v\": 9}]}, {\"c\": [{\"v\": 150}, {\"v\": 4}]}, {\"c\": [{\"v\": 175}, {\"v\": 76}]}, {\"c\": [{\"v\": 200}, {\"v\": 100}]}, {\"c\": [{\"v\": 225}, {\"v\": 104}]}, {\"c\": [{\"v\": 250}, {\"v\": 80}]}, {\"c\": [{\"v\": 275}, {\"v\": 46}]}, {\"c\": [{\"v\": 300}, {\"v\": 40}]}, {\"c\": [{\"v\": 325}, {\"v\": 46}]}, {\"c\": [{\"v\": 350}, {\"v\": 26}]}, {\"c\": [{\"v\": 375}, {\"v\": 26}]}, {\"c\": [{\"v\": 400}, {\"v\": 9}]}, {\"c\": [{\"v\": 425}, {\"v\": 4}]}, {\"c\": [{\"v\": 450}, {\"v\": 4}]}, {\"c\": [{\"v\": 500}, {\"v\": 2}]}, {\"c\": [{\"v\": 550}, {\"v\": 1}]}, {\"c\": [{\"v\": 600}, {\"v\": 1}]}, {\"c\": [{\"v\": 18700}, {\"v\": 1}]}], \"cols\": [{\"type\": \"number\", \"id\": \"latency\", \"label\": \"latency\"}, {\"type\": \"number\", \"id\": \"requests\", \"label\": \"requests\"}]},\n",
       "            {\n",
       "              pageSize: 25,\n",
       "              cssClassNames:  {\n",
       "                tableRow: 'gchart-table-row',\n",
       "                headerRow: 'gchart-table-headerrow',\n",
       "                oddTableRow: 'gchart-table-oddrow',\n",
       "                selectedTableRow: 'gchart-table-selectedrow',\n",
       "                hoverTableRow: 'gchart-table-hoverrow',\n",
       "                tableCell: 'gchart-table-cell',\n",
       "                headerCell: 'gchart-table-headercell',\n",
       "                rowNumberCell: 'gchart-table-rownumcell'\n",
       "              }\n",
       "            },\n",
       "            {source_index: 2, fields: 'latency,requests'},\n",
       "            0,\n",
       "            23);\n",
       "        }\n",
       "      );\n",
       "    </script>\n",
       "  "
      ],
      "text/plain": [
       "QueryResultsTable job_01iap04gpHTfDEnm3tO29csZJJw"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%bigquery execute --query latency"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f703f22dd90>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeYAAAF8CAYAAAAabhjqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQShmAAAM8v8BIDHDFHpYnTIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f703f2365d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "bq.Query(latency).to_dataframe().set_index('latency').plot(kind='bar')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Step 3 - Latency Distribution\n",
    "\n",
    "In this step, we want to determine the percentage of requests that fall within each bucket. In order to compute percentages, we need to issue a count query to determine counts, then use the buckets along with counts to produce percentage values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "%%sql --module latency\n",
    "\n",
    "DEFINE QUERY buckets\n",
    "  SELECT INTEGER(latency / 25) * 25 AS latency, COUNT(latency) AS requests\n",
    "  FROM [cloud-datalab-samples:httplogs.logs_20140615]\n",
    "  WHERE endpoint = 'Recent'\n",
    "  GROUP BY latency\n",
    "  ORDER BY latency\n",
    "\n",
    "DEFINE QUERY stats\n",
    "  SELECT COUNT(latency) AS total\n",
    "  FROM [cloud-datalab-samples:httplogs.logs_20140615]\n",
    "  WHERE endpoint = 'Recent'\n",
    "\n",
    "SELECT bucket.latency AS latency,\n",
    "       bucket.requests AS requests,\n",
    "       bucket.requests * 100 / stats.total AS percent\n",
    "FROM $buckets AS bucket CROSS JOIN $stats AS stats\n",
    "ORDER BY latency"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Notice that the first query was named`buckets` and the second was named `stats`, and these were referenced in the third query as `$buckets` and `$stats`, which uses the parameter syntax seen earlier to parameterize queries to pass in Python variables.\n",
    "\n",
    "The third query was left as anonymous, and is treated as the _main_ query within the SQL module named `latency`. We can inspect the fully composed SQL represented by the module by creating a `Query` object and printing its `sql` property."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SELECT bucket.latency AS latency,\n",
      "       bucket.requests AS requests,\n",
      "       bucket.requests * 100 / stats.total AS percent\n",
      "FROM (SELECT INTEGER(latency / 25) * 25 AS latency, COUNT(latency) AS requests\n",
      "  FROM [cloud-datalab-samples:httplogs.logs_20140615]\n",
      "  WHERE endpoint = 'Recent'\n",
      "  GROUP BY latency\n",
      "  ORDER BY latency) AS bucket CROSS JOIN (SELECT COUNT(latency) AS total\n",
      "  FROM [cloud-datalab-samples:httplogs.logs_20140615]\n",
      "  WHERE endpoint = 'Recent') AS stats\n",
      "ORDER BY latency\n"
     ]
    }
   ],
   "source": [
    "latency_distribution = bq.Query(latency)\n",
    "print latency_distribution.sql"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Notice that the `buckets` and `stats` queries have been inlined as nested queries and used in a `JOIN`. The resulting SQL represents a single query that is issued and executed in BigQuery."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "    <div class=\"bqtv\" id=\"4_146281576606\"><table><tr><th>latency</th><th>requests</th><th>percent</th></tr><tr><td>0</td><td>66</td><td>8.99182561308</td></tr><tr><td>25</td><td>4</td><td>0.544959128065</td></tr><tr><td>50</td><td>12</td><td>1.6348773842</td></tr><tr><td>75</td><td>45</td><td>6.13079019074</td></tr><tr><td>100</td><td>28</td><td>3.81471389646</td></tr><tr><td>125</td><td>9</td><td>1.22615803815</td></tr><tr><td>150</td><td>4</td><td>0.544959128065</td></tr><tr><td>175</td><td>76</td><td>10.3542234332</td></tr><tr><td>200</td><td>100</td><td>13.6239782016</td></tr><tr><td>225</td><td>104</td><td>14.1689373297</td></tr><tr><td>250</td><td>80</td><td>10.8991825613</td></tr><tr><td>275</td><td>46</td><td>6.26702997275</td></tr><tr><td>300</td><td>40</td><td>5.44959128065</td></tr><tr><td>325</td><td>46</td><td>6.26702997275</td></tr><tr><td>350</td><td>26</td><td>3.54223433243</td></tr><tr><td>375</td><td>26</td><td>3.54223433243</td></tr><tr><td>400</td><td>9</td><td>1.22615803815</td></tr><tr><td>425</td><td>4</td><td>0.544959128065</td></tr><tr><td>450</td><td>4</td><td>0.544959128065</td></tr><tr><td>500</td><td>2</td><td>0.272479564033</td></tr><tr><td>550</td><td>1</td><td>0.136239782016</td></tr><tr><td>600</td><td>1</td><td>0.136239782016</td></tr><tr><td>18700</td><td>1</td><td>0.136239782016</td></tr></table></div>\n",
       "    <br />(rows: 23, time: 3.0s,    10MB processed, job: job_zpKjhmCSKzd7QmSaA1xwZBuBZaE)<br />\n",
       "    <script>\n",
       "      require(['datalab/charting', 'datalab/element!4_146281576606', 'base/js/events',\n",
       "          'datalab/style!/nbextensions/gcpdatalab/charting.css'],\n",
       "        function(charts, dom, events) {\n",
       "          charts.render('gcharts', dom, events, 'table', [], {\"rows\": [{\"c\": [{\"v\": 0}, {\"v\": 66}, {\"v\": 8.991825613079019}]}, {\"c\": [{\"v\": 25}, {\"v\": 4}, {\"v\": 0.5449591280653951}]}, {\"c\": [{\"v\": 50}, {\"v\": 12}, {\"v\": 1.6348773841961852}]}, {\"c\": [{\"v\": 75}, {\"v\": 45}, {\"v\": 6.130790190735695}]}, {\"c\": [{\"v\": 100}, {\"v\": 28}, {\"v\": 3.8147138964577656}]}, {\"c\": [{\"v\": 125}, {\"v\": 9}, {\"v\": 1.2261580381471389}]}, {\"c\": [{\"v\": 150}, {\"v\": 4}, {\"v\": 0.5449591280653951}]}, {\"c\": [{\"v\": 175}, {\"v\": 76}, {\"v\": 10.354223433242506}]}, {\"c\": [{\"v\": 200}, {\"v\": 100}, {\"v\": 13.623978201634877}]}, {\"c\": [{\"v\": 225}, {\"v\": 104}, {\"v\": 14.168937329700272}]}, {\"c\": [{\"v\": 250}, {\"v\": 80}, {\"v\": 10.899182561307901}]}, {\"c\": [{\"v\": 275}, {\"v\": 46}, {\"v\": 6.267029972752043}]}, {\"c\": [{\"v\": 300}, {\"v\": 40}, {\"v\": 5.449591280653951}]}, {\"c\": [{\"v\": 325}, {\"v\": 46}, {\"v\": 6.267029972752043}]}, {\"c\": [{\"v\": 350}, {\"v\": 26}, {\"v\": 3.542234332425068}]}, {\"c\": [{\"v\": 375}, {\"v\": 26}, {\"v\": 3.542234332425068}]}, {\"c\": [{\"v\": 400}, {\"v\": 9}, {\"v\": 1.2261580381471389}]}, {\"c\": [{\"v\": 425}, {\"v\": 4}, {\"v\": 0.5449591280653951}]}, {\"c\": [{\"v\": 450}, {\"v\": 4}, {\"v\": 0.5449591280653951}]}, {\"c\": [{\"v\": 500}, {\"v\": 2}, {\"v\": 0.2724795640326976}]}, {\"c\": [{\"v\": 550}, {\"v\": 1}, {\"v\": 0.1362397820163488}]}, {\"c\": [{\"v\": 600}, {\"v\": 1}, {\"v\": 0.1362397820163488}]}, {\"c\": [{\"v\": 18700}, {\"v\": 1}, {\"v\": 0.1362397820163488}]}], \"cols\": [{\"type\": \"number\", \"id\": \"latency\", \"label\": \"latency\"}, {\"type\": \"number\", \"id\": \"requests\", \"label\": \"requests\"}, {\"type\": \"number\", \"id\": \"percent\", \"label\": \"percent\"}]},\n",
       "            {\n",
       "              pageSize: 25,\n",
       "              cssClassNames:  {\n",
       "                tableRow: 'gchart-table-row',\n",
       "                headerRow: 'gchart-table-headerrow',\n",
       "                oddTableRow: 'gchart-table-oddrow',\n",
       "                selectedTableRow: 'gchart-table-selectedrow',\n",
       "                hoverTableRow: 'gchart-table-hoverrow',\n",
       "                tableCell: 'gchart-table-cell',\n",
       "                headerCell: 'gchart-table-headercell',\n",
       "                rowNumberCell: 'gchart-table-rownumcell'\n",
       "              }\n",
       "            },\n",
       "            {source_index: 3, fields: 'latency,requests,percent'},\n",
       "            0,\n",
       "            23);\n",
       "        }\n",
       "      );\n",
       "    </script>\n",
       "  "
      ],
      "text/plain": [
       "QueryResultsTable job_zpKjhmCSKzd7QmSaA1xwZBuBZaE"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "latency_distribution.results()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f703d03b990>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeAAAAFqCAYAAADC/4krAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f703d0cb990>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "latency_distribution.to_dataframe().drop(['requests'], axis = 1).set_index('latency').plot(logx = True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Looking Ahead\n",
    "\n",
    "Composing queries one step at a time and validating the results along the way (by either executing or sampling indermediate queries) allows you to build sophisticated queries while containing the complexity.\n",
    "\n",
    "Look at some of the BigQuery samples for more examples."
   ]
  }
 ],
 "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": "2.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}