{
 "metadata": {
  "name": "",
  "signature": "sha256:96746f6bc03e6fec9c102fd49c4dffc99ad3edefff7266618321c39d3d020abb"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "<img src=\"images/blaze_med.png\" alt=\"Blaze Logo\"\n",
      "                                align=\"right\"\n",
      "                                width=\"30%\">\n",
      "\n",
      "# Getting Started with Blaze\n",
      "\n",
      "* Full tutorial available at http://github.com/ContinuumIO/blaze-tutorial\n",
      "* Install software with `conda install -c blaze blaze`"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "## 3.  Storing results with `into`\n",
      "\n",
      "We just played with some interesting queries on baseball statistics"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from blaze import Data, into, by, join\n",
      "db = Data('sqlite:///data/lahman2013.sqlite')\n",
      "joined = join(db.Salaries, db.Teams)\n",
      "result = by(joined[['name', 'yearID']], avg=joined.salary.mean())\n",
      "result"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "How do we now store this result or use it with other libraries?\n",
      "\n",
      "The result itself is a Blaze expression, not terribly useful if we're not using Blaze."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "type(result)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Use normal Python collections like `list` or `np.array`\n",
      "\n",
      "Blaze follows normal conventions and so can be converted by standard constructors"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "list(result)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import numpy as np\n",
      "np.array(result)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "### Use `into`\n",
      "\n",
      "Alternatively, Blaze has registered itself into the `into` project and so can migrate its results to any of those formats."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "into('salaries.csv', result)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "!head salaries.csv"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "#### Exercise\n",
      "\n",
      "Dump `results` into the following formats"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Dump results into a Python set\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Dump results into a Pandas DataFrame\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Dump results into a JSON file, inspect the file to make sure that it came out ok\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}