4-analyzing-experiments.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "source": [
    "First we load `rpy2.ipython` extension so we can execute `R` code."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [],
   "source": [
    "%load_ext rpy2.ipython"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Analyzing Experimental Data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Getting started\n",
    "\n",
    "## Load up the R packages we're going to use"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "%%R\n",
    "suppressMessages({\n",
    "    # install.packages(c('dplyr', 'tidyr', 'ggplot2', 'broom'))\n",
    "    library(dplyr)\n",
    "    library(tidyr)\n",
    "    library(ggplot2)\n",
    "    library(broom)\n",
    "})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Exposure logs\n",
    "First we logged everytime a user loaded up the user interface. Often we want to only take the first exposure by a user."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "We have 492 exposures. \n",
       "We have 164 potential subjects. \n",
       "\n",
       "Source: local data frame [6 x 6]\n",
       "\n",
       "  userid story_key  source position       time source.ideology\n",
       "1   1388         2   msnbc        1 1431954455              -1\n",
       "2   1388         7 foxnews        0 1431954455               1\n",
       "3   1388         8     cnn        2 1431954455               0\n",
       "4   2025         3 foxnews        0 1431955391               1\n",
       "5   2025         5     cnn        2 1431955391               0\n",
       "6   2025         8   msnbc        1 1431955391              -1\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "src2ideology <- c(\n",
    "    # <=liberal=====moderate====conservative=>\n",
    "    'msnbc' = -1, 'cnn' = 0, 'foxnews' = 1)\n",
    "\n",
    "Exp <- read.csv('data/exposure_mturk2_log.csv', stringsAsFactors = FALSE) %>%\n",
    "    group_by(userid, story_key) %>%\n",
    "    arrange(time) %>%\n",
    "    summarise(source = first(source), position = first(position), time = first(time)) %>%\n",
    "    mutate(source.ideology = src2ideology[source]) %>%\n",
    "    ungroup\n",
    "\n",
    "cat('We have', nrow(Exp), 'exposures.', '\\n')\n",
    "cat('We have', with(Exp, length(unique(userid))), 'potential subjects.', '\\n\\n')\n",
    "head(Exp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Source: local data frame [3 x 2]\n",
       "\n",
       "  source.ideology   n\n",
       "1              -1 164\n",
       "2               0 164\n",
       "3               1 164\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "Exp %>% group_by(source.ideology) %>% summarise(n = n())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Events logs\n",
    "\n",
    "The events log is produced by user actions that take place sometime after the exposure. So in this experiment we logged two kinds of these:  summaries and survey responses."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "We have 390 summaries. \n",
       "We have 130 subjects who wrote summaries. \n",
       "0.37 stories were summarized. \n",
       "\n",
       "Source: local data frame [6 x 3]\n",
       "\n",
       "  userid story_key summary\n",
       "1   1388         2       0\n",
       "2   1388         7       0\n",
       "3   1388         8       1\n",
       "4   2025         3       0\n",
       "5   2025         5       1\n",
       "6   2025         8       0\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "Sum <- read.csv('data/summary_mturk2_log.csv', stringsAsFactors = FALSE) %>%\n",
    "    group_by(userid, story_key) %>%\n",
    "    arrange(time) %>%\n",
    "    summarise(summary = first(summary)) %>%\n",
    "    ungroup\n",
    "\n",
    "cat('We have', nrow(Sum), 'summaries.', '\\n')\n",
    "cat('We have', with(Sum, length(unique(userid))), 'subjects who wrote summaries.', '\\n')\n",
    "cat(with(Sum, round(mean(summary), 2)), 'stories were summarized.', '\\n\\n')\n",
    "head(Sum)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#Survey responses\n",
    "\n",
    "After the experiment, we survey the subjects about their political ideology."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "We have  129 survey responses. \n",
       "Source: local data frame [6 x 4]\n",
       "\n",
       "  userid response reversed_scale subject.ideology\n",
       "1   1388        1              0                1\n",
       "2   2025        2              0                0\n",
       "3   2263        1              1                1\n",
       "4   3856        2              0                0\n",
       "5   4875        3              0               -1\n",
       "6   5477        4              1               -2\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "Survey <- read.csv('data/survey_mturk2_log.csv', stringsAsFactors = FALSE) %>%\n",
    "  filter(!is.na(response)) %>%\n",
    "  group_by(userid) %>%\n",
    "  arrange(time) %>%\n",
    "  summarise(response = first(response), reversed_scale = first(reversed_scale)) %>%\n",
    "  # Original scale is [0, 4] (from conservative to liberal)\n",
    "  # let's convert it to [-2, 2] (liberal to conservative)\n",
    "  mutate(subject.ideology = 2 - response) %>%\n",
    "  ungroup\n",
    "\n",
    "cat('We have ', nrow(Survey), 'survey responses.', '\\n')\n",
    "head(Survey)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Joining exposures to events\n",
    "\n",
    "Typically we need to perform joins from the exposures to the outcomes in order to analyze the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "  0   1 \n",
       "180  90 \n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "\n",
    "Results <- Exp %>% \n",
    "  inner_join(Sum, c('userid', 'story_key')) %>%\n",
    "  inner_join(Survey, 'userid') %>%\n",
    "  mutate(opposed = as.numeric((subject.ideology * source.ideology) < 0)) %>%\n",
    "  group_by(userid) %>%\n",
    "  mutate(num_stories = n(), num_summaries = sum(summary)) %>%\n",
    "  ungroup %>%\n",
    "  # We only study the people who followed instructions and have a political affiliation.\n",
    "  filter(num_stories == 3, subject.ideology != 0)\n",
    "\n",
    "with(Results, summary(factor(opposed)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#EDA\n",
    "\n",
    "It's good to do some EDA on your data to make sure all the data we collected looks good."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "stat_bin: binwidth defaulted to range/30. Use 'binwidth = x' to adjust this.\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeAAAAHgCAYAAAB91L6VAAAD8GlDQ1BJQ0MgUHJvZmlsZQAAOI2N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\noFgBAVysr9YJECBAgECuwP8CUTO8n/5RwGoAAAAASUVORK5CYII=\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "# When did people show up to do the experiment?\n",
    "ggplot(Exp, aes(x = time)) + \n",
    "  geom_histogram() +\n",
    "  theme_bw()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeAAAAHgCAYAAAB91L6VAAAD8GlDQ1BJQ0MgUHJvZmlsZQAAOI2N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\nqUldCCCAAAIIeAr8P2/nD3fKR5SOAAAAAElFTkSuQmCC\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "# Check that we have balanced exposures.\n",
    "Results %>%\n",
    "  group_by(source.ideology) %>%\n",
    "  summarise(n = n(), num.users = length(unique(userid))) %>%\n",
    "  ggplot(aes(x = source.ideology, y = n)) +\n",
    "  geom_bar(stat = 'identity') +\n",
    "  xlab('(Manipulated) Source Ideology') +\n",
    "  ylab('Number of observations') +\n",
    "  theme_bw()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeAAAAHgCAYAAAB91L6VAAAD8GlDQ1BJQ0MgUHJvZmlsZQAAOI2N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\nKiGAAAII2C/wv+l+u9Kn1iabAAAAAElFTkSuQmCC\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "# Summary rate by story to see if any stories are weird.\n",
    "Sum %>%\n",
    "  group_by(story_key) %>%\n",
    "  summarise(k = sum(summary), n = n(),\n",
    "            p = mean(summary),\n",
    "            lcl = binom.test(k, n)$conf.int[1], \n",
    "            ucl = binom.test(k, n)$conf.int[2]) %>%\n",
    "  ggplot(aes(x = factor(story_key), y = p, ymin = lcl, ymax = ucl)) +\n",
    "  geom_pointrange() +\n",
    "  xlab('Story ID') +\n",
    "  ylab('Probability of Summary') +\n",
    "  theme_bw()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeAAAAHgCAYAAAB91L6VAAAD8GlDQ1BJQ0MgUHJvZmlsZQAAOI2N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\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "\n",
    "# Make sure our survey results look sane.\n",
    "ideology.labels <- c('very liberal', 'liberal', 'moderate', 'conservative', 'very conservative')\n",
    "\n",
    "Survey %>%\n",
    "  ggplot(aes(x = factor(subject.ideology, labels = ideology.labels), fill = subject.ideology)) +\n",
    "  geom_histogram() +\n",
    "  scale_fill_gradient(low = \"blue\", high = \"red\") + # haha\n",
    "  theme_bw() +\n",
    "  theme(legend.position = 'none') +\n",
    "  xlab('Political Affiliation')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Measuring the ATE\n",
    "Now we can tabulate the selections by source ideology to measure the ATE.  We use the standard formulas:\n",
    "\n",
    "$\\hat{ATE} = \\frac{1}{n_1} \\sum_{i \\in T} Y_i - \\frac{1}{n_0} \\sum_{i \\in C} Y_i$\n",
    "\n",
    "$\\hat{SE}(\\hat{ATE}) = \\sqrt{\\frac{1}{n_0 + n_1}} \\sqrt{\\frac{n_1}{n_0} \\mbox{Var}(Y_i(0)) + \\frac{n_0}{n_1} \\mbox{Var}(Y_i(1))}$\n",
    "\n",
    "Note these are not appropriate for our experiment because our data are not i.i.d."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Source: local data frame [1 x 3]\n",
       "\n",
       "         ate        lcl         ucl\n",
       "1 -0.1277778 -0.2152437 -0.04031181\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "Results %>%\n",
    "  group_by(opposed) %>%\n",
    "  summarise(y = mean(summary),\n",
    "            n = n(),\n",
    "            v = var(summary)) %>%\n",
    "  gather(measure, value, y:v) %>%\n",
    "  unite(quantity, measure, opposed, sep = '') %>%\n",
    "  mutate(blah = 1) %>% # must add this because I DONT KNOW WHY\n",
    "  spread(quantity, value) %>%\n",
    "  mutate(ate = y1 - y0, \n",
    "         se = sqrt(1 / (n0 + n1)) * sqrt((n1 / n0) * v0 + (n0/n1) * v1),\n",
    "         lcl = ate - 1.96 * se,\n",
    "         ucl = ate + 1.96 * se) %>%\n",
    "  select(ate, lcl, ucl)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#Modeling the Results\n",
    "We usually prefer modeling the data using a regression.  We'll show how to use linear regression to perform correct inference on data with dependence, control variables, and larger than you can fit into memory.\n",
    "\n",
    "##Outline\n",
    "\n",
    "* i.i.d. case (with and without covariates)\n",
    "* bootstrapping to deal with dependence\n",
    "* using data reduction with weighted regressions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# I.i.d. case\n",
    "If your data are i.i.d (which is pretty rare), then you can use an OLS regression and trust the standard errors on the treatment variable that `R` produces.\n",
    "\n",
    "This analysis is inappropriate for our experiment because of user- and story-level dependence across observations.  Later we'll show how we account for this using a bootstrap."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeAAAAHgCAYAAAB91L6VAAAD8GlDQ1BJQ0MgUHJvZmlsZQAAOI2N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\nY0MAAQQQMFbg/wFfVbOVUnqRgAAAAABJRU5ErkJggg==\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "lm(summary ~ opposed, data = Results) %>%\n",
    "  tidy %>%\n",
    "  ggplot(aes(x = term, y = estimate, \n",
    "             ymin = estimate - 1.96*std.error, \n",
    "             ymax = estimate + 1.96*std.error)) +\n",
    "  geom_pointrange() +\n",
    "  geom_hline(yintercept = 0.0, linetype = 'dashed') +\n",
    "  coord_flip() +\n",
    "  theme_bw()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# I.i.d. case with control variables\n",
    "Including control variables in your regression shouldn't change the ATE estimate, but it might help with precision (giving you smaller confidence intervals)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeAAAAHgCAYAAAB91L6VAAAD8GlDQ1BJQ0MgUHJvZmlsZQAAOI2N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\nAAAAAElFTkSuQmCC\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "# In this case we think the position should help explain variation in the outcome (which is plausible).\n",
    "lm(summary ~ opposed + factor(position), data = Results) %>%\n",
    "  tidy %>%\n",
    "  ggplot(aes(x = term, y = estimate, \n",
    "             ymin = estimate - 1.96*std.error, \n",
    "             ymax = estimate + 1.96*std.error)) +\n",
    "  geom_pointrange() +\n",
    "  geom_hline(yintercept = 0.0, linetype = 'dashed') +\n",
    "  coord_flip() +\n",
    "  theme_bw()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Bootstrapping to get confidence intervals\n",
    "\n",
    "If your data exhibits dependence, then you want to use a bootstrap to get correct inference.  In our experiment we resample both users and stories, so the true model has two way dependence.  We'll start with just user-level dependence:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [],
   "source": [
    "%%R\n",
    "suppressMessages({\n",
    "    # install.packages(c('lmtest', 'Hmisc', 'foreach'))\n",
    "    source('css_stats.R') # Provides `clustered.bootstrap`\n",
    "})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Source: local data frame [6 x 5]\n",
       "\n",
       "         term   estimate  std.error statistic      p.value\n",
       "1 (Intercept)  0.4700000 0.04795154  9.801561 6.296919e-18\n",
       "2     opposed -0.2200000 0.08003333 -2.748855 6.695621e-03\n",
       "3 (Intercept)  0.4250000 0.05307832  8.007036 5.768641e-13\n",
       "4     opposed -0.1750000 0.08456729 -2.069358 4.049329e-02\n",
       "5 (Intercept)  0.4137931 0.05141050  8.048805 4.118301e-13\n",
       "6     opposed -0.1429598 0.08621804 -1.658119 9.965016e-02\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R \n",
    "\n",
    "fitter <- function(.data) {\n",
    "    lm(summary ~ opposed, data = .data, weights = .weights) %>%\n",
    "    tidy\n",
    "}\n",
    "\n",
    "iid.replicates    <- iid.bootstrap(Results, fitter, .combine = bind_rows)\n",
    "oneway.replicates <- clustered.bootstrap(Results, c('userid'), fitter, .combine = bind_rows)\n",
    "twoway.replicates <- clustered.bootstrap(Results, c('userid', 'story_key'), fitter, .combine = bind_rows)\n",
    "\n",
    "head(iid.replicates)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeAAAAHgCAYAAAB91L6VAAAD8GlDQ1BJQ0MgUHJvZmlsZQAAOI2N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\n/wHyxu9QJz++WwAAAABJRU5ErkJggg==\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "replicates <- bind_rows(\n",
    "    iid.replicates %>% mutate(bootstrap = 'iid'),\n",
    "    oneway.replicates %>% mutate(bootstrap = 'oneway'),\n",
    "    twoway.replicates %>% mutate(bootstrap = 'twoway')\n",
    ")\n",
    "\n",
    "replicates %>%\n",
    "  group_by(term, bootstrap) %>%\n",
    "  summarise(y = mean(estimate), lcl = quantile(estimate, 0.025), ucl = quantile(estimate, 0.975)) %>%\n",
    "  ggplot(aes(x = bootstrap, y = y, ymin = lcl, ymax = ucl)) +\n",
    "  facet_grid(term ~ .) +\n",
    "  geom_pointrange() +\n",
    "  geom_hline(yintercept = 0.0, linetype = 'dashed') +\n",
    "  coord_flip() +\n",
    "  theme_bw()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#Using Data Reduction\n",
    "\n",
    "If you have a lot of data, sometimes it's a lot easier to group it and fit weighted models.  We show how to do that here."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Source: local data frame [4 x 3]\n",
       "Groups: summary\n",
       "\n",
       "  summary opposed cases\n",
       "1       0       0   105\n",
       "2       0       1    64\n",
       "3       1       0    75\n",
       "4       1       1    26\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "\n",
    "Results.small <- Results %>%\n",
    "  group_by(summary, opposed) %>%\n",
    "  summarise(cases = n())\n",
    "\n",
    "Results.small"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Fitting Reduced Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeAAAAHgCAYAAAB91L6VAAAD8GlDQ1BJQ0MgUHJvZmlsZQAAOI2N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\nAAAASUVORK5CYII=\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "lm(summary ~ opposed, data = Results.small, weights = cases) %>%\n",
    "  coeftest(vcov = sandwich.lm) %>%\n",
    "  tidy %>%\n",
    "  ggplot(aes(x = term, y = estimate, \n",
    "             ymin = estimate - 1.96*std.error, \n",
    "             ymax = estimate + 1.96*std.error)) +\n",
    "  geom_pointrange() +\n",
    "  geom_hline(yintercept = 0.0, linetype = 'dashed') +\n",
    "  coord_flip() +\n",
    "  theme_bw()  \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Reducing data when you have covariates\n",
    "\n",
    "If your covariates are discrete and low-cardinality, you can still use data reduction."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeAAAAHgCAYAAAB91L6VAAAD8GlDQ1BJQ0MgUHJvZmlsZQAAOI2N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\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R \n",
    "Results.medium <- Results %>%\n",
    "  group_by(summary, opposed, position) %>%\n",
    "  summarise(cases = n())\n",
    "\n",
    "lm(summary ~ opposed + factor(position), data = Results.medium, weights = cases) %>%\n",
    "  coeftest(vcov = sandwich.lm) %>%\n",
    "  tidy %>%\n",
    "  ggplot(aes(x = term, y = estimate, \n",
    "             ymin = estimate - 1.96*std.error, \n",
    "             ymax = estimate + 1.96*std.error)) +\n",
    "  geom_pointrange() +\n",
    "  geom_hline(yintercept = 0.0, linetype = 'dashed') +\n",
    "  coord_flip() +\n",
    "  theme_bw()  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Bootstrapping Reduced Data\n",
    "\n",
    "If your data are dependent and large, you can combine data reduction with the bootstrap:\n",
    "\n",
    "1. Generate a bootstrap replicate\n",
    "2. Reduce that replicate\n",
    "3. Fit a model on the reduced replicate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Source: local data frame [6 x 5]\n",
       "\n",
       "               term    estimate std.error  statistic    p.value\n",
       "1       (Intercept)  0.46132972 0.3211212  1.4366218 0.19397433\n",
       "2           opposed -0.15196744 0.4581708 -0.3316830 0.74983810\n",
       "3 factor(position)1  0.05234593 0.4479023  0.1168691 0.91024625\n",
       "4 factor(position)2 -0.16617868 0.4447901 -0.3736114 0.71975175\n",
       "5       (Intercept)  0.58501678 0.2951675  1.9819821 0.08279434\n",
       "6           opposed -0.14535970 0.3517462 -0.4132516 0.69027714\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "\n",
    "reduced.fitter <- function(.data) {\n",
    "    reduced.data <- .data %>%\n",
    "       group_by(summary, opposed, position) %>%\n",
    "       summarise(cases = sum(.weights))\n",
    "    lm(summary ~ opposed + factor(position), data = reduced.data, weights = cases) %>%\n",
    "    tidy\n",
    "}\n",
    "\n",
    "twoway.replicates <- clustered.bootstrap(Results, c('userid', 'story_key'), reduced.fitter, .combine = bind_rows)\n",
    "\n",
    "head(twoway.replicates)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {
    "collapsed": false,
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeAAAAHgCAYAAAB91L6VAAAD8GlDQ1BJQ0MgUHJvZmlsZQAAOI2N\nVd1v21QUP4lvXKQWP6Cxjg4Vi69VU1u5GxqtxgZJk6XpQhq5zdgqpMl1bhpT1za2021Vn/YCbwz4\nA4CyBx6QeEIaDMT2su0BtElTQRXVJKQ9dNpAaJP2gqpwrq9Tu13GuJGvfznndz7v0TVAx1ea45hJ\nGWDe8l01n5GPn5iWO1YhCc9BJ/RAp6Z7TrpcLgIuxoVH1sNfIcHeNwfa6/9zdVappwMknkJsVz19\nHvFpgJSpO64PIN5G+fAp30Hc8TziHS4miFhheJbjLMMzHB8POFPqKGKWi6TXtSriJcT9MzH5bAzz\nHIK1I08t6hq6zHpRdu2aYdJYuk9Q/881bzZa8Xrx6fLmJo/iu4/VXnfH1BB/rmu5ScQvI77m+Bkm\nfxXxvcZcJY14L0DymZp7pML5yTcW61PvIN6JuGr4halQvmjNlCa4bXJ5zj6qhpxrujeKPYMXEd+q\n00KR5yNAlWZzrF+Ie+uNsdC/MO4tTOZafhbroyXuR3Df08bLiHsQf+ja6gTPWVimZl7l/oUrjl8O\ncxDWLbNU5D6JRL2gxkDu16fGuC054OMhclsyXTOOFEL+kmMGs4i5kfNuQ62EnBuam8tzP+Q+tSqh\nz9SuqpZlvR1EfBiOJTSgYMMM7jpYsAEyqJCHDL4dcFFTAwNMlFDUUpQYiadhDmXteeWAw3HEmA2s\n15k1RmnP4RHuhBybdBOF7MfnICmSQ2SYjIBM3iRvkcMki9IRcnDTthyLz2Ld2fTzPjTQK+Mdg8y5\nnkZfFO+se9LQr3/09xZr+5GcaSufeAfAww60mAPx+q8u/bAr8rFCLrx7s+vqEkw8qb+p26n11Aru\nq6m1iJH6PbWGv1VIY25mkNE8PkaQhxfLIF7DZXx80HD/A3l2jLclYs061xNpWCfoB6WHJTjbH0mV\n35Q/lRXlC+W8cndbl9t2SfhU+Fb4UfhO+F74GWThknBZ+Em4InwjXIyd1ePnY/Psg3pb1TJNu15T\nMKWMtFt6ScpKL0ivSMXIn9QtDUlj0h7U7N48t3i8eC0GnMC91dX2sTivgloDTgUVeEGHLTizbf5D\na9JLhkhh29QOs1luMcScmBXTIIt7xRFxSBxnuJWfuAd1I7jntkyd/pgKaIwVr3MgmDo2q8x6IdB5\nQH162mcX7ajtnHGN2bov71OU1+U0fqqoXLD0wX5ZM005UHmySz3qLtDqILDvIL+iH6jB9y2x83ok\n898GOPQX3lk3Itl0A+BrD6D7tUjWh3fis58BXDigN9yF8M5PJH4B8Gr79/F/XRm8m241mw/wvur4\nBGDj42bzn+Vmc+NL9L8GcMn8F1kAcXgSteGGAAA/SElEQVR4Ae3dB5gUVbbA8TPknIYofuzCooCC\n6GOFFRAJ4ltQUbIZQVfeElz1KQqYHoICggISVgVUMiIKukoUn6uygJmwZMkCLpJzfJ77ttqemZ6Z\nnp6u7rpV//o+mO4KN/xOd5++t6q7Uy78sggLAggggAACCCRUIE9Ca6MyBBBAAAEEEDACJGAeCAgg\ngAACCCRBgAScBHSqRAABBBBAgATMYwABBBBAAIEkCJCAk4BOlQgggAACCJCAeQwggAACCCCQBAES\ncBLQqRIBBBBAAAESMI8BBBBAAAEEkiBAAk4COlUigAACCCBAAuYxgAACCCCAQBIESMBJQKdKBBBA\nAAEESMA8BhBAAAEEEEiCAAk4CehUiQACCCCAQD4IvCFw5swZmTNnjuTNmzdDg5wfrEpJScmwzbYV\n2pfz589H7KdtfdH2al80Ln6Iza5du6RQoUKSmppqYyjStNlPzxnt2Llz53zznPFbbC6++GKpX79+\nmsdftHdIwNFKubzf2bNnpWzZstKsWbMMNWly1sScJ4/9ExaasI4dOybFixfP0E8bV5w+fVry58/v\niwTcpUsXqV27tvzpT3+yMRRp2qyPM01aGhs/LIcPH5YSJUr4oSsmLpqE8+XzR/qZNWtWzAnY/ld0\nXzwk6QQCCCCAQNAESMBBizj9RSATgXLlykmpUqUy2cpqBBCIt4A/5gDirUJ5CARQYMCAAb6Zsg1g\n+OiyhQKMgC0MGk1GAAEEELBfgARsfwzpAQIIIICAhQIkYAuDRpMRQAABBOwXIAHbH0N6gEBcBJ54\n4gkZP358XMqiEAQQyF6ABJy9EXsgEAiBQ4cOydGjRwPRVzqJgBcESMBeiAJtQAABBBAInAAJOHAh\np8MIIIAAAl4Q4HPAXogCbUDAAwJdu3b1xfdAe4CSJiAQlQAJOComdkLA/wL6hfJ++e5k/0eLHvpB\ngCloP0SRPiCAAAIIWCdAArYuZDQYAQQQQMAPAiRgP0SRPiAQB4EvvvhCVq9eHYeSKAIBBKIRIAFH\no8Q+CARAYMqUKbJ48eIA9JQuIuANARKwN+JAKxBAAAEEAiZAAg5YwOkuAggggIA3BEjA3ogDrUAg\n6QJly5aVkiVLJr0dNACBoAjwOeCgRJp+IpCNwHPPPcfngLMxYjMC8RRgBBxPTcpCAAEEEEAgSgES\ncJRQ7IYAAggggEA8BUjA8dSkLAQQQAABBKIUIAFHCcVuCPhdoF+/fjJ+/Hi/d5P+IeAZARKwZ0JB\nQxBIrsCBAwfk6NGjyW0EtSMQIAEScICCTVcRQAABBLwjQAL2TixoCQIIIIBAgAT4HHCAgk1XEchK\n4N5775XU1NSsdmEbAgjEUYAEHEdMikLAZoEGDRrwRRw2B5C2WyfAFLR1IaPBCCCAAAJ+ECAB+yGK\n9AEBBBBAwDoBErB1IaPBCLgjsHTpUlmzZo07hVMqAghkECABZyBhBQLBFJg8ebIsWrQomJ2n1wgk\nQYAEnAR0qkQAAQQQQIAEzGMAAQQQQACBJAiQgJOATpUIeFFAPwNcokQJLzaNNiHgSwE+B+zLsNIp\nBHIuMHDgQD4HnHM2jkAgZgFGwDHTcSACCCCAAAKxC5CAY7fjSAQQQAABBGIWIAHHTMeBCCCAAAII\nxC5AAo7djiMR8JVA//79ZcKECb7qE51BwMsCJGAvR4e2IZBAgf3798uRI0cSWCNVIRBsARJwsONP\n7xFAAAEEkiRAAk4SPNUigAACCARbgM8BBzv+9B6BkMA999wjZcuWDd3nBgIIuCtAAnbXl9IRsEbg\nmmuu4Ys4rIkWDfWDAFPQfogifUAAAQQQsE6ABGxdyGgwAggggIAfBEjAfogifUAgDgL/+Mc/5J//\n/GccSqIIBBCIRoAEHI0S+yAQAIFJkybJwoULA9BTuoiANwRIwN6IA61AAAEEEAiYAAk4YAGnuwgg\ngAAC3hAgAXsjDrQCgaQLlClTRooXL570dtAABIIiwOeAgxJp+olANgKDBg3ic8DZGLEZgXgKMAKO\npyZlIYAAAgggEKUACThKKHZDAAEEEEAgngIk4HhqUhYCCCCAAAJRCpCAo4TK7W4tW7bMbREc71OB\nbdu2yahRo2TRokVJ7eGTTz4pEydOTGobqByBIAkE4iKsM2fOZLi45Pz583Lu3LkM60+ePCmFChVK\n8xiIdLz+cHn6K0Yj7Xfq1CkpWLCg7Nq1K02Z3EFABfRxceWVV8rBgwcNyCuvvCK9evVKCs7PP/8s\nhw8fTkrdVIpAEAVSLvyy+LXjGzdulN69e5sEWLJkSTPK0MTZpk0b0fuahO+8807p3r27NGjQQEqV\nKiX58uWTihUryvjx402C1u3Hjx+XvHnzyrPPPiu/+c1v5NZbb5UKFSqIjlwmT54sv/3tb+Xhhx+W\nvXv3mjIHDBggtWvXlh49esi6deskT548snXrVtm0aVOIeu7cuWa7syI1NVWGDBkijRo1claF/uob\nBS0jJSUltM7WG/pw+8///E9jZ2sf4tlufWyFJ738+fOLPhaSseibAK2/aNGiyag+2zqbN28u48aN\ny3Y/3UEfZ/r81uetH5ZIAwNb+6Vx0UVf0/ywzJs3Tzp37hxTV3w9Av7rX/8qOq3WuHFjGTlypMyc\nOVNat24t27dvN8lSHwBVqlSRLl26yL59+0T3v+qqq6RDhw7yySefyLFjx6ROnTqmjO+//14GDx5s\nkmaxYsVk7Nixsnv3boOuX9+nCXnMmDGyefNms1/Xrl1N0l2yZIn5ft30U9DNmjVLM+V49uxZ0RFI\niRIlMgRSR9b6QuKHB6w++V599dUMMw8ZOm3JitOnT5u+xPrmaM2aNdKpU6dQb/V2v379QvcTeaNv\n375yySWXSLdu3RJZbdR16YxTpOdHpAL0cRZphivSvrasi7bvXu+PxkXfIOlgxw9Lbl6X/SGQSRQ1\nMWrS1KVhw4YydOhQk4Dr168fCv7ll19uRmOFCxc2U4G6b7169URHz99995107NhRV0ndunVl2bJl\nMnXqVLnuuuukRYsWZhStU4aLFy+W5cuXy9dff2321SeKHq/l6HLZZZeZfc2df/+n++h6Zzlx4oRJ\nwM59P/+tVq1ahul7W/ub2wSsjwF9Y6jfw1yrVi0zy5KsEajOFpUtWzbN49LWuNBuBGwQ8MccQCbS\nmnS//PJLs1WTZ/Xq1c1tTaz6Dkzfia1evVqqVq0qmgB1ulgXTab6wtikSROTdHXdqlWrzNS0/lqM\njhJ0RHz//ffL6NGj5dprrxUd4eq0so6idTq7Zs2asmLFCj1UtmzZIocOHTK3+Q+B9AI66v3b3/4m\nL774YlKnf/X5om9IWRBAIDECvh4B6zv6/v37G0lNgO+8845JtDr1oVPAeoHUCy+8IAUKFDBTvD17\n9jTTiZqoNakePXpU/vznP0v79u3N1PJrr71mknifPn3k7bffNlPM+qKpI119AdUX0v3795spRE3C\nNWrUMIlZE7Z+zR8LAggggAACjoCvL8JyOqkXuhQpUsTc3bFjhxm5asLUUbAmX110+k9Ht7pv+inA\n8OPNzr/8p8lZ9ws/96ejaJ3KDl+ivXhCj9VRur4xSL/47RywnltPfwV5+j7bcj+3U9Be6qc+zvUi\nLP1n++K3c8B6oR7ngL35qJw1a1boVGVOW+jrKWgHw0m+el9PmJcuXdq8yDjJV9fruS9NpumTr24L\nP17v66IXYoUnX12XPvnquvQfadJ1LAh4UUBPvaxdu9aLTaNNCPhSIBAJODxylStXlhkzZoSvMrc/\n++yzDOtYgUCQBN58801ZsGBBkLpMXxFIqkDgEnBStakcAQQQQACBfwuQgHkoIIAAAgggkAQBEnAS\n0KkSAS8K6DfB6bUNLAggkBgBX38MKTGE1IKAPwT0I3l+uALaH9GgF0EQYAQchCjTRwQQQAABzwmQ\ngD0XEhqEAAIIIBAEARJwEKJMHxFAAAEEPCdAAvZcSGgQAskReOqpp+SNN95ITuXUikAABUjAAQw6\nXUYgkoD+JCc/GhJJhnUIuCNAAnbHlVIRQAABBBDIUoAEnCUPGxFAAAEEEHBHgM8Bu+NKqQhYJ3DX\nXXeZHyWxruE0GAFLBUjAlgaOZiMQb4FGjRrxRRzxRqU8BLIQYAo6Cxw2IYAAAggg4JYACdgtWcpF\nAAEEEEAgCwEScBY4bEIgSAIrVqyQdevWBanL9BWBpAqQgJPKT+UIeEdAv4Rj/vz53mkQLUHA5wIk\nYJ8HmO4hgAACCHhTgATszbjQKgQQQAABnwuQgH0eYLqHQLQCJUuWlGLFikW7O/shgEAuBfgccC4B\nORwBvwgMHjyYzwH7JZj0wwoBRsBWhIlGIoAAAgj4TYAE7LeI0h8EEEAAASsESMBWhIlGIoAAAgj4\nTYAE7LeI0h8EYhR45pln5M0334zxaA5DAIGcCpCAcyrG/gj4VOCnn36SgwcP+rR3dAsB7wmQgL0X\nE1qEAAIIIBAAARJwAIJMFxFAAAEEvCfA54C9FxNahEBSBO68804pW7ZsUuqmUgSCKEACDmLU6TMC\nEQQaN27MF3FEcGEVAm4JMAXtlizlIoAAAgggkIUACTgLHDYhgAACCCDglgAJ2C1ZykXAMoEvv/xS\n1q9fb1mraS4C9gqQgO2NHS1HIK4CEydOlHnz5sW1TApDAIHMBUjAmduwBQEEEEAAAdcESMCu0VIw\nAggggAACmQuQgDO3YQsCgRIoUaKEFC1aNFB9prMIJFOAzwEnU5+6EfCQwJAhQ/gcsIfiQVP8L8AI\n2P8xpocIIIAAAh4UIAF7MCg0CQEEEEDA/wIkYP/HmB4igAACCHhQgATswaDQJASSIfDss8/KW2+9\nlYyqqROBQAqQgAMZdjqNQEaBvXv3yoEDBzJuYA0CCLgiQAJ2hZVCEUAAAQQQyFqABJy1D1sRQAAB\nBBBwRYDPAbvCSqEI2Cdw++23S7ly5exrOC1GwFIBErClgaPZCMRboEmTJnwRR7xRKQ+BLASYgs4C\nh00IIIAAAgi4JUACdkuWchFAAAEEEMhCgAScBQ6bEAiSwFdffSXr168PUpfpKwJJFSABJ5WfyhHw\njsCECRNk3rx53mkQLUHA5wIkYJ8HmO4hgAACCHhTgATszbjQKgQQQAABnwuQgH0eYLqHQLQCxYsX\nl6JFi0a7O/shgEAuBfgccC4BORwBvwgMHTqUzwH7JZj0wwoBRsBWhIlGIoAAAgj4TYAE7LeI0h8E\nEEAAASsESMBWhIlGIoAAAgj4TYAE7LeI0h8EYhT4n//5H5k0aVKMR3MYAgjkVIAEnFMx9kfApwJ7\n9uyR/fv3+7R3dAsB7wmQgL0XE1qEAAIIIBAAARJwAIJMFxFAAAEEvCfA54C9FxNahEBSBG677TYp\nV65cUuqmUgSCKEACDmLU6TMCEQSuu+46vogjggurEHBLgClot2QpFwEEEEAAgSwESMBZ4LAJAQQQ\nQAABtwRIwG7JUi4Clgl88803smHDBstaTXMRsFeABGxv7Gg5AnEVeP311+Wjjz6Ka5kUhgACmQuQ\ngDO3YQsCCCCAAAKuCZCAXaOlYAQQQAABBDIXIAFnbsMWBAIlULRoUSlSpEig+kxnEUimAJ8DTqY+\ndSPgIYFhw4bxOWAPxYOm+F+AEbD/Y0wPEUAAAQQ8KEAC9mBQaBICCCCAgP8FSMD+jzE9RAABBBDw\noAAJ2INBoUkIJENgwIABMnny5GRUTZ0IBFKABBzIsNNpBDIK7N69W37++eeMG1iDAAKuCJCAXWGl\nUAQQQAABBLIWIAFn7cNWBBBAAAEEXBHgc8CusFIoAvYJdO7cWcqVK2dfw2kxApYKkIAtDRzNRiDe\nAk2bNuWLOOKNSnkIZCHAFHQWOGxCAAEEEEDALQESsFuylIsAAggggEAWAiTgLHDYhECQBL799lvZ\nuHFjkLpMXxFIqgAJOKn8VI6AdwRee+01+fDDD73TIFqCgM8FSMA+DzDdQwABBBDwpgAJ2JtxoVUI\nIIAAAj4XIAH7PMB0D4FoBYoUKSKFCxeOdnf2QwCBXArwOeBcAnI4An4RGD58OJ8D9ksw6YcVAoyA\nrQgTjUQAAQQQ8JsACdhvEaU/CCCAAAJWCJCArQgTjUQAAQQQ8JuAqwl4/fr1UqdOHenfv3/UbmvX\nrpWlS5dGvf++fftyVH5mBV9//fVmk1O//jbq3XffndnuEdefPn1aevXqFdo2Y8YMadu2reiX3P/9\n738PreeGtwV++ukneeutt2T58uXebmicWzdw4ECZMmVKnEulOAQQyEzA1QT86aefmiQ2aNAgU//5\n8+fl2LFjGdpy6tSp0LoPPvhANmzYELp/9uzZ0O3wG5rsdNGyw5NeeFnh+585cyb8ruh+Thm6Yfr0\n6Wa7U3+FChVk5MiRaY5JX4ZuPHHiRGifAgUKmDccn3zyiflhcz1+5syZMnbsWHnkkUfkwoULoX25\n4U2BvXv3yhVXXCH33nuv/OEPf5CJEyd6s6EutGrXrl2ib2hZEEAgMQIpvyQFV7KCPplvueUWk3Re\nfPFFWbVqlcydO9dcZanJ7+OPP5atW7fKbbfdJhdffLFo8tJv4rnuuuukYMGC5p24Jq/PP/9cUlJS\nzAuijiTvv/9+2blzpxw6dEg0Wd51110yf/58mTp1qjleP0rxr3/9S8aNGydXX321vP/++zJixAgp\nWrSo1K1bV/RdvibEWbNmmTqvvPJKGTJkiEmcOvK+9tprTf3Dhg2T559/XubNmxexjAceeMC04ejR\no7Jt2zb56KOPpEqVKibx3nnnnaav2sby5cubNx2XXnqpbN++XfLmzWsiq+U++uijoSiXLl3ajOQb\nNWoUWufcOHfunOTJk8c4OOsi/X377bdF2+31RR9yGlMvLhqzPXv2hJqmH8vRuAZh0Vkffe6VKVMm\nYd0tVqyYLFmyJO716WNM3/A7z7e4V5DgAk+ePCmFChVKcK3uVKdx0UVf0/yw6Gu55qZYFtc+hlS5\ncmXp1q2beRI0b95cVq5cKQsXLjT3GzduLD/88IP07dvXJLmWLVuaBLp//3654447TNLSFwFNiJpc\ndRT8+9//Xjp27Ghu69/77rtPdIRds2ZN028NaqlSpUziW716tfTo0cNM++pHKxYsWGAevDodrNPi\ns2fPlieeeEK03vfee88cr3UUL148VH/VqlVNXboxUhmaFOvVqyd9+vSR5557zrRTk3Jqaqp5AdcX\nMk2+Osq+5557TH3hLwb65uD11183det/2n4dYZcoUSK0zrmh6/XY7B6wN910k1x22WXOYZ78q/3U\nWQN9Q+TFRR873bt3DzVNHyOPP/546H76GxqbfPnyefYNRfr2ZnVfH8fVqlUTfQOZqEXtIj3mc1u/\nPs70OZo/f/7cFuWZ491wSkbnNC76Bklj74clu9flrPqYMAHFbtWqlehI78CBAybZbN68WerXr2/a\nl/5Jv3jxYpN0daMeqwldz8/qUqtWLfNXR9DhPyDesGFDs/7yyy+XHTt2iE4naqCdd466XcvVEfGA\nAQOkZ8+e5p1L+/btzXGR/susDN3XqU+TbvjUt7b3yJEjpt36ZqFFixbSu3fvNMWXLVtW9J+zaFJa\ntmyZczemvxUrVhT95+VFXxj1NIS+2fHiojHVJ5SeA9bH2dChQ80bu8zaqm+w9EXeqyP6zNodab2e\nytHnk/OcjLQP6xBAIH4CCZkD0JHtqFGjzChRz7XqdLO+EDdt2lS++eYb05uXXnpJVqxYYUZ6uk1H\nyXpfF72v++moVBdnJOlM65qVv/znlPX999+bd/J6Hvf48eOh5KgJrnr16mZEPGnSJDMtPmfOHNHp\ncmfRsrU+Z8msDN3utMPZ1/mrL8o6wrv11lvl9ttvl4cfftjZxF8LBPQ0x2effWZOaeisSlCWZs2a\nyVVXXRWU7tJPBJIukJARsE4n66i0a9euZkpWR4w//vijmerr16+fOV+rSfmhhx4yo2MdLdauXdtc\nBKPnkfXc1DPPPCN6fjd80XO6OqXsLHqeWUfZOmrVEYwuOs2tiVCnPfRcnk4p6lWuuk5HYVqGjq6d\nRevV+sNH1pHK0POtkRY9/6z1vPPOO+b8tdblnJfVc106A8CCAAIIIICAaxdhRaLVadlIU4/pLzAI\nn9bT2zrSzGy0qeeC9RydfmREL8568MEHTaJOPyWoU7zh33OrU9M6FaoXgaRfwusP35a+jPBtzm29\n4KxGjRrSpk0bZ1VUf50paB2FpF+iPQec/jgv3tfZBS9PQefULLPHSk7L8cL+Oluk0+l+OG+qjzM/\nnQM+fPiwK+fKk/G489s5YL2gV081xrIkZAraaVik5KvbnHO0zn46GnYSqN7OLPnq/npVs17xrGXo\ntK/+c451ytO/4clX7+s+kZKvbguvX+87S/oynPXOX30x1ot4cpp8neP5i0AyBb799lvZtGlTMptA\n3QgESiAhU9BuilaqVMlM8WaVpN2sP7xsTdxvvPFG+CpuI2CNgH4MUE/BOBc5WtNwGoqApQIJHQG7\nZeSF5Ov0LTeXpDtl8BcBBBBAwP8CvkjA/g8TPUQAAQQQ8JsACdhvEaU/CMQooNc4pL8eI8aiOAwB\nBKIQsP4ccBR9ZBcEEIhCQD+L74croKPoKrsg4AkBRsCeCAONQAABBBAImgAJOGgRp78IIIAAAp4Q\nIAF7Igw0AgEEEEAgaAIk4KBFnP4ikImA/rb2lClTMtnKagQQiLcAF2HFW5TyELBUQL/KNUg/PmFp\nmGi2jwQYAfsomHQFAQQQQMAeARKwPbGipQgggAACPhJgCtpHwaQrCORGoEOHDqK/f82CAAKJESAB\nJ8aZWhDwvECLFi34Ig7PR4kG+kmAKWg/RZO+IIAAAghYI0ACtiZUNBQBBBBAwE8CJGA/RZO+IJAL\nge+++042b96cixI4FAEEciJAAs6JFvsi4GOBV199VT744AMf95CuIeAtARKwt+JBaxBAAAEEAiJA\nAg5IoOkmAggggIC3BEjA3ooHrUEgaQKFChWSggULJq1+KkYgaAJ8DjhoEae/CGQi8PLLL/M54Exs\nWI2AGwKMgN1QpUwEEEAAAQSyESABZwPEZgQQQAABBNwQIAG7oUqZCCCAAAIIZCNAAs4GiM0IBEVg\n0KBBMnXq1KB0l34ikHQBLsJKeghoAALeENi5c6eUKlXKG42hFQgEQIARcACCTBcRQAABBLwnQAL2\nXkxoEQIIIIBAAASYgg5AkOkiAtEItG/fXipUqBDNruyDAAJxECABxwGRIhDwg8D111/PF3H4IZD0\nwRoBpqCtCRUNRQABBBDwkwAJ2E/RpC8IIIAAAtYIxDwFffDgQVm3bp1cuHAh1NkaNWpImTJlQve5\ngQAC9gisXLlSSpcuLfo8ZkEAAfcFYkrATz/9tIwZM8ZcsJE3b95QK4cPHy433HBD6D43EEDAHoFx\n48ZJ7dq15bHHHrOn0bQUAYsFYkrAEyZMkA0bNkhqaqrFXafpCCCAAAIIJE8gpnPAlSpVkqNHjyav\n1dSMAAIIIICA5QIxjYBHjx4tDRo0kFatWknZsmVDBHfffbdcccUVofvcQAABewQKFiwoBQoUsKfB\ntBQBywViSsCPPPKIXH755XLRRRdJnjy/DqLDzwdb7kLzEQicwIgRI/gccOCiToeTKRBTAt64caPs\n2rWLd8vJjBx1I4AAAghYLfDr8DUH3WjRooUsXrw4B0ewKwIIIIAAAgiEC8SUgE+ePCk33nijOf9b\ns2ZNcf7Nnz8/vGxuI4AAAggggEAmAjFNQfft21f69++focjq1atnWMcKBBCwQ+D555+XSy65RLp0\n6WJHg2klApYLxJSAO3XqJIsWLZJLL73U8u7TfAQQcAR27NghJUuWdO7yFwEEXBaIaQr64osvllWr\nVrncNIpHAAEEEEDAvwIxJWD9zdAOHTpI+fLlzceR9CNJ+m/BggX+laJnCCCAAAIIxFEgpinoJ598\nUvr06ZOhGUxJZyBhBQLWCLRr104qVqxoTXtpKAK2C8SUgP/jP/7D9Pvs2bOiv4pUqlQpyZcvpqJs\n96P9CPhGoGXLlnwRh2+iSUdsEIhpCnrr1q1y//33m2/C0l9Qeeihh2Tw4ME29Jc2IoAAAggg4AmB\nmBJwt27dpFq1ajJgwADTCZ2OnjZtmvmFJE/0ikYggAACCCDgcYEcJ+ALFy7ImjVrzG+GFipUyHSv\nSpUq0rlzZ1myZInHu0vzEEAgMwH9ZMOWLVsy28x6BBCIs0COE3BKSooUK1YszceQzp8/L3/7299E\nf6aQBQEE7BQYO3aszJ07187G02oELBSI6cqpoUOHSpMmTaRq1armoo1XX31VLrvsMrnpppssJKDJ\nCCCAAAIIJF4gpgTctGlT+frrr+XDDz+UM2fOiH58Qb9BR6+ITk1NTXwvqBEBBBBAAAHLBHKUgPVH\nGHTRBPzll19Kjx49zH2dgtYLs/74xz/Kvffea9bxHwII2CWQP39+PoZkV8horeUCOUrA+gtIzoVW\nhQsXDnVdzwtXrlw5dFV0aAM3EEDAGoFRo0aRgK2JFg31g0COLsLS3wDWKefbb7/d/NXb+k+/kEO/\nyJ1vwvLDQ4I+IIAAAggkQiBHI2Ad6eo3XulnflkQQAABBBBAIHaBHI2AY6+GIxFAAAEEEEAgXIAE\nHK7BbQQCLKBfJzt9+vQAC9B1BBIrkKMp6MQ2jdoQQCCRAtu2bZPixYsnskrqQiDQAoyAAx1+Oo8A\nAgggkCwBEnCy5KkXAQQQQCDQAkxBBzr8dB6BXwXatm0rFStW/HUFtxBAwFUBErCrvBSOgD0CN9xw\nA1/EYU+4aKkPBJiC9kEQ6QICCCCAgH0CJGD7YkaLEUAAAQR8IEAC9kEQ6QIC8RBYvXq1bN26NR5F\nUQYCCEQhQAKOAoldEAiCwJgxY2TOnDlB6Cp9RMATAiRgT4SBRiCAAAIIBE2ABBy0iNNfBBBAAAFP\nCJCAPREGGoFA8gX0l870HwsCCCRGgGdbYpypBQHPC7zyyit8DtjzUaKBfhJgBOynaNIXBBBAAAFr\nBEjA1oSKhiKAAAII+EmABOynaNIXBBBAAAFrBEjA1oSKhiLgrsCQIUNkxowZ7lZC6QggEBLgIqwQ\nBTcQCLaAfgtWsWLFgo1A7xFIoAAj4ARiUxUCCCCAAAKOAAnYkeAvAggggAACCRRgCjqB2FSFgJcF\nbrnlFqlUqZKXm0jbEPCVAAnYV+GkMwjELvDHP/6RL+KInY8jEcixAFPQOSbjAAQQQAABBHIvQALO\nvSElIIAAAgggkGMBEnCOyTgAAX8KrFmzRrZt2+bPztErBDwoQAL2YFBoEgLJEBg9erS89957yaia\nOhEIpAAJOJBhp9MIIIAAAskWIAEnOwLUjwACCCAQSAEScCDDTqcRyCiQL18+yZs3b8YNrEEAAVcE\n+BywK6wUioB9Aq+88gqfA7YvbLTYYgFGwBYHj6YjgAACCNgrQAK2N3a0HAEEEEDAYgESsMXBo+kI\nIIAAAvYKuJqA169fL3Xq1JH+/ftHLbR27VpZunRp1Pvv27cvR+VnVvD1119vNjn17969W+6+++7M\ndo+4/vTp09KrV6802/Q3VvVL7lkQiIfA5s2bzWd19XEf72Xo0KEyc+bMeBdLeQggkImAqwn4008/\nNUls0KBBpvrz58/LsWPHMjTl1KlToXUffPCBbNiwIXT/7NmzodvhNzTZ6aJlhye98LLC9z9z5kz4\nXdH9nDJ0w/Tp0812p/4KFSrIyJEj0xyTvgzdeOLEidA+BQoUMG84PvnkE7Nu/vz50qFDB/nhhx9C\n+3ADgVgFFi5cKLVq1ZJ27dqZv9u3b4+1qIjHbdmyRfSNJwsCCCRGwLWroHft2iWvvfaaXLhwQX7/\n+9/LqlWrZO7cueYqS01+H3/8sejo8LbbbpOLL75YNHnp/tOmTZOCBQtKo0aNzLvxzz//XFJSUuTe\ne++Vzp07y/333y87d+6UQ4cOiSZLHbHqT6hNnTrVHF+kSBH517/+JePGjZOrr75a3n//fRkxYoQU\nLVpU6tatKwMHDpSxY8fKrFmzTJ1XXnmlDBkyRJo3b25G3k79v/vd7+T555+XefPmRSzjgQceMG04\nevSo+fq+jz76SKpUqWIS7p133inNmjUT3fbhhx+KM7oOD+nixYvl6aefDq0qWbKkPPjgg3L48OHQ\nOufGuXPnJE+ePMbBWWfrX3086Buf8ePHh9702NoXbbf2Rx+fiVg2bdokzptAHQHrcySePx+obxT1\nzaM+N/yy6PO0U6dO1nfn5MmT1vfB6YAOxHTR1zQ/LE5/YumLawm4cuXK0q1bN9HGaXJbuXKl6Dt4\nvd+4cWMzKuzbt69Jci1btjQJdP/+/XLHHXdI+fLlpUyZMiYh6ihSR8GaxDt27Ghu69/77rtPdIRd\ns2ZN028tt1SpUibJr169Wnr06CF///vfZfjw4bJgwQIpVKiQtG3bVnRafPbs2fLEE0+I1ut89Z7W\nUbx48VD9VatWNXVp4ZHK0KRYr1496dOnjzz33HOi7dQne2pqquzZs8e0SUe/mY3gL7/8cnnqqafM\nfs5/2sYSJUo4d0N/9UVXP5/phwesxklnQW688cZQ7EIdtfCGxkY/P5uIJDxhwgTz2HWY7rnnHvNc\ncu7n9u+wYcPkt7/9rXkTmduykn28Ps70X+3atSM+p5Ldvljqj/TaEEs5yT5GXzv1jas+b/yw5OZ1\nOWECit2qVSspXbq0HDhwwLyT1/NZ9evXNzHQUWP4oiNETbq66LGa0HW0q4tOw+miI+hy5cqZ2/pf\nw4YNzW1Nbjt27JC9e/eaQGtic7ZruToiHjBggPTs2dOMqtu3b2+2R/ovszKc8vSvJt3wqW9t75Ej\nR0xCj1SmrtORS/joRaeyly1bltnuvlt/6aWXiv6zfdHRfP78+ROSgHXEW6xYMfnqq6/MNLQ+huO5\nHD9+3DwmnedRPMtOdFmafPWFXmPDgoBXBRIyB6Aj21GjRplRop5r1elmfYI0bdpUvvnmG2Pz0ksv\nyYoVK8xIT7fpKFnv66L3dT8dlerifFuPvoCHnwdzyvr++++lWrVqoudx9UXFSY6a4KpXr25GxJMm\nTTLT4nPmzBGdLncWLVvrc5bMytDtTjucfZ2/+qKso2kWBOIpoCOgN998U3SGJ97JV9upb5D1tA0L\nAggkRiAhI2CdTtZRadeuXc35Px0x/vjjj9K9e3fp16+fOV+rSfmhhx4yo+PevXubqaM//OEP5gpi\nvTDkmWeeET2/G77oOV2dUnYWPc+sLyI6an3rrbfMap3mvvXWW827YT1Hq9POP/30k1mnSVLL0NG1\ns+iUldYfPrKOVMbbb7/tHJLmr55/1npYEEAAAQQQyEog5Ze5+AtZ7RDPbZlNy+oFBs40sdYXPq2n\nt3WkmdloU88FP/7447J8+XJzcZZeyKSJOv05OZ3iLVy4cKg72m09F6lTeumX8PrDt6UvI3ybc/vF\nF1+UGjVqSJs2bZxVUf11pqD14q30ix/PAftlhiCzx0r6GNpwX2eLdMrWD9O2OovlpylovTiTc8De\nfBbpRYt6XVIsS0KmoJ2GZfaiG558dV8dDTsJVG9nlnx1X72qWa941jL0Smf95xyr250lPPnqOt0n\nUvLVbeH1631nSV+Gs975qy/GOj2Y0+TrHM9fBJIpoNdYhJ/SSWZbqBuBIAgkZAraTUi9kEmv3swq\nSbtZf3jZmrjfeOON8FXcRsAaAb1OQ0/BPPbYY9a0mYYiYLNAQkfAbkF5Ifk6fcvNJelOGfxFAAEE\nEPC/gC8SsP/DRA8RQAABBPwmQAL2W0TpDwIxCuhMkpdmk2LsBochYI2A9eeArZGmoQh4XGD06NG+\nuALa48w0D4GQACPgEAU3EEAAAQQQSJwACThx1tSEAAIIIIBASIAEHKLgBgIIIIAAAokTIAEnzpqa\nEPC0gH6LW2ZfserphtM4BCwV4CIsSwNHsxGIt4D+HnD671uPdx2UhwACvwowAv7VglsIIIAAAggk\nTIAEnDBqKkIAAQQQQOBXAaagf7XgFgKBFrj55ptFv1udBQEEEiNAAk6MM7Ug4HmB1q1b80Ucno8S\nDfSTAFPQfoomfUEAAQQQsEaABGxNqGgoAggggICfBEjAfoomfUEgFwLr1q2T7du356IEDkUAgZwI\nkIBzosW+CPhYYOTIkfLuu+/6uId0DQFvCZCAvRUPWoMAAgggEBABEnBAAk03EUAAAQS8JUAC9lY8\naA0CSRNISUmRPHl4SUhaAKg4cAJ8DjhwIafDCEQWGDt2LJ8DjkzDWgRcEeDtriusFIoAAggggEDW\nAiTgrH3YigACCCCAgCsCJGBXWCkUAQQQQACBrAVIwFn7sBWBwAgMHz5cZs2aFZj+0lEEki3ARVjJ\njgD1I+ARgU2bNkmhQoU80hqagYD/BRgB+z/G9BABBBBAwIMCJGAPBoUmIYAAAgj4X4ApaP/HmB4i\nEJXAzTffLBUrVoxqX3ZCAIHcC5CAc29ICQj4QqB169Z8EYcvIkknbBFgCtqWSNFOBBBAAAFfCZCA\nfRVOOoMAAgggYIsACdiWSNFOBFwWWL9+vezYscPlWigeAQQcARKwI8FfBAIuMGLECJk9e3bAFeg+\nAokTIAEnzpqaEEAAAQQQCAmQgEMU3EAAAQQQQCBxAiTgxFlTEwKeF0hJSfF8G2kgAn4R4HPAfokk\n/UAglwLjxo3jc8C5NORwBHIiwAg4J1rsiwACCCCAQJwESMBxgqQYBBBAAAEEciJAAs6JFvsigAAC\nCCAQJwEScJwgKQYB2wVeeukleeedd2zvBu1HwBoBLsKyJlQ0FAF3BTZu3CgFCxZ0txJKRwCBkAAj\n4BAFNxBAAAEEEEicAAk4cdbUhAACCCCAQEiAKegQBTcQCLbAjTfeKJUqVQo2Ar1HIIECJOAEYlMV\nAl4WuOmmm/giDi8HiLb5ToApaN+FlA4hgAACCNggQAK2IUq0EQEEEEDAdwIkYN+FlA4hEJvAhg0b\nZOfOnbEdzFEIIJBjARJwjsk4AAF/Crz88st8EYc/Q0uvPCpAAvZoYGgWAggggIC/BUjA/o4vvUMA\nAQQQ8KgACdijgaFZCCCAAAL+FuBzwP6OL71DIGqBcePG8TngqLXYEYHcCzACzr0hJSCAAAIIIJBj\nARJwjsk4AAEEEEAAgdwLkIBzb0gJCCCAAAII5FiABJxjMg5AwJ8C+jng2bNn+7Nz9AoBDwpwEZYH\ng0KTEEiGgH4TVoECBZJRNXUiEEgBRsCBDDudRgABBBBItgAJONkRoH4EEEAAgUAKMAUdyLDTaQQy\nCrRu3VoqVaqUcQNrEEDAFQESsCusFIqAfQI333wzX8RhX9hoscUCTEFbHDyajgACCCBgrwAJ2N7Y\n0XIEEEAAAYsFSMAWB4+mIxBPgU2bNsmuXbviWSRlIYBAFgIk4Cxw2IRAkASGDx8us2bNClKX6SsC\nSRUgASeVn8oRQAABBIIqQAIOauTpNwIIIIBAUgVIwEnlp3IEEEAAgaAK8DngoEaefiOQTmDs2LF8\nDjidCXcRcFOAEbCbupSNgEUCKSkpov9YEEAgMQIk4MQ4UwsCCCCAAAJpBEjAaTi4gwACCCCAQGIE\nSMCJcaYWBDwvMGLECJk9e7bn20kDEfCLABdh+SWS9AOBXAqsX7+ei7ByacjhCOREgBFwTrTYFwEE\nEEAAgTgJkIDjBEkxCCCAAAII5ESAKeicaLEvAj4WaNWqlVx00UU+7iFdQ8BbAiRgb8WD1iCQNIE2\nbdpwDjhp+lQcRAGmoIMYdfqMAAIIIJB0ARJw0kNAAxBAAAEEgihAAg5i1OkzAhEENm/eLD/++GOE\nLaxCAAE3BEjAbqhSJgIWCgwbNkzefvttC1tOkxGwU4AEbGfcaDUCCCCAgOUCJGDLA0jzEUAAAQTs\nFCAB2xk3Wo0AAgggYLkAnwO2PIA0H4F4CYwePVoKFCgQr+IoBwEEshFgBJwNEJsRCIpA3rx5JU8e\nXhKCEm/6mXwBnm3JjwEtQAABBBAIoAAJOIBBp8sIIIAAAskX8EUC3rdvn/Tv399oNm/eXI4ePZqp\n7IQJEzLdFs8Na9eulaVLl5oiH3nkkSzbFM96KQuBWAVGjhwp7777bqyHc5wPBQ4cOCBLliyRn3/+\n2Ye9S36XfJGABw0aJL169TKaO3fulAsXLoRkT5w4EbqtN4YPH57m/pkzZ9Lc1zunT58OrdPb4eXp\nhnPnzmVYp+tPnjypf8zywQcfyIYNG8zt7t27y4svvvjvLfxBwJsC69atk+3bt3uzcbQq4QIbN26U\nSy+9VFq0aCHVq1eXVatWJbwNfq/Q+qug9+/fLzrarFSpUppYPfDAA3Lo0CEz8ty2bZt89NFH8vHH\nH8uuXbvkhRdekE6dOskzzzwjmqBLlCgho0aNEi1L1+sI+tFHH5VNmzbJ559/LoUKFZKHHnpI9Ofa\nRowYYdadOnVKunbtKu3atZMGDRpIqVKlJF++fFKxYkV5+eWXZdq0aVKwYEFp1KiR1KhRw9Sto3Tn\nKtNPP/1UBg8eHGpz8eLFpVu3bnLkyJHQOueGJny9OCYlJcVZZe1ffTOjdn5Zzp49K3rxkh9io33R\nN5yRHoO2xUsfZ+fPnzexcavt+vqRqKSkrwH6OEvkoglYZxd1OXjwoLRu3Vpq1qyZ6yY4AxovPGeK\nFSsmkyZNylWf9HEW62J9AtYnQKQHhT5g69WrJ3369JHnnntO5s+fL5qUdSTat29fk2D1fpMmTUSn\npd9//32TLPfu3Sv6nbia1MePHy9ffPGF6DTM9OnTzd958+bJggULzJP7hhtuMAlYH6R//etf5aqr\nrpIOHTrIV199JXfccYeUL19eLrnkEhObatWqyfr166VOnTrmftWqVaVLly6huOmDUZOzPiDSLzpK\n98sVqvpg1b5G6mf6fttwXxNW/vz5fZGA9TGW2WPQhliEt1EfZ/oaoLFxa7npppvk6quvdqv4NOXq\nQKFw4cJp1rl9R1/rJk+eHKrmmmuukbZt24bux3pDY6NJONFvKCK1Nx6P99x8csD6BLx161YpV65c\nJFtp2LChWZ+ampph1LV48WKTEHVKWh8Qmoh10SkXfWDo6Ld+/fpmXenSpaVHjx4yZ84c2bJli9xy\nyy1mvT659+zZY54YV155pVmnSV/fOaZfNBlrW50EXKVKFdF/zqJPsGXLlkV8IdeE5fxz9rf1r/ZD\nF+evrf1w2u3ExQ/90Rmeiy66yBexSURc9HqTRC2HDx82M3WJqk/r0dk9ndVbuHChXHfddfLaa69J\n0aJFc90EfWOkCVjLDvpi/TlgTZiZnbeK9A7LeaG89tpr5eGHH5a5c+fKY489JnXr1jWPBecY3b5y\n5UqzTqdf/uu//kv0HWDlypXNMe+9955J8JrcNXnq+TNdli9fLpdddplJ4prYnUWnwbWtLAh4VUDf\nWDpvWr3aRtqVOAE9hTZx4kTR62qmTp0al+SbuNbbUZP1b0E0cerUbrRL2bJl5c9//rP85S9/MVdO\nv/rqq+bc75QpU0widcrRxKrvcPVdoF5cpftXqFBBWrZsKbfeequZ3mratKmZ4tKk3bNnT3NbL1bQ\n5H38+HHp3bu31K5d24ykd+zYQQJ2cPmLAAIIICApv0wF/HrJsKUg9913nzz++ONRJzi9CEjf3emS\n3bkVPcenUyXh8/x6sYouzhRKrVq15J///KdJuuFTNM75QT0nPGPGjAxXYJtC/v2fMwXdrFmz8NXm\ntt/OAR87dkz0ojM/LE6MnZkVm/ukbxr1tIqb500T5ZOIc8CJ6ovWk4wpaLf657cp6FmzZknHjh1j\n4rJ+Clp7PXDgQBk3blzUAE7y1QOyu7BBT9KHJ189RhOvk3z1vo6q9QU4PPnqej1W1+tFXk899ZSu\nYkHAswI//PCD7N6927Pto2EI+E3AFwlYP4KkPyaerOWzzz7LsuoxY8aYjylluRMbEUiygH5CYObM\nmUluBdUjEBwBXyRgDZdz8ZQXQ+fltnnRizYhgAACQRDwTQIOQrDoIwIIIICAfwRIwP6JJT1BAAEE\nELBIwPqPIVlkTVMR8LTAK6+8EvqqVE83lMYh4BMBRsA+CSTdQCC3AnplP9cr5FaR4xGIXoAEHL0V\neyKAAAIIIBA3ARJw3CgpCAEEEEAAgegFSMDRW7EnAr4W0HPA+h3nLAggkBgBLsJKjDO1IOB5Af06\n1fTf+ub5RtNABCwWYARscfBoOgIIIICAvQIkYHtjR8sRQAABBCwWYAra4uDRdATiKXDDDTeY37uO\nZ5mUhQACmQuQgDO3YQsCgRJo27atL36KMFBBo7NWCzAFbXX4aDwCCCCAgK0CJGBbI0e7EUAAAQSs\nFiABWx0+Go9A/AS2bNkie/bsiV+BlIQAAlkKkICz5GEjAsERGDp0qMyYMSM4HaanCCRZgASc5ABQ\nPQIIIIBAMAVIwMGMO71GAAEEEEiyAAk4yQGgegQQQACBYArwOeBgxp1eI5BBYNSoUVKgQIEM61mB\nAALuCDACdseVUhGwTiB//vySLx/vya0LHA22VoAEbG3oaDgCCCCAgM0CJGCbo0fbEUAAAQSsFSAB\nWxs6Go5AfAVGjx4tc+bMiW+hlIYAApkKcMInUxo2IBAsgTVr1khKSkqwOk1vEUiiACPgJOJTNQII\nIIBAcAVIwMGNPT1HAAEEEEiiAFPQScSnagS8JNCyZUupXLmyl5pEWxDwtQAJ2NfhpXMIRC/Qrl07\n0c8CsyCAQGIEmIJOjDO1IIAAAgggkEaABJyGgzsIIIAAAggkRoAEnBhnakHA8wLbtm2TvXv3er6d\nNBABvwiQgP0SSfqBQC4FBg8eLNOnT89lKRyOAALRCpCAo5ViPwQQQAABBOIoQAKOIyZFIYAAAggg\nEK0ACThaKfZDAAEEEEAgjgJ8DjiOmBSFgM0CI0eOlIIFC9rcBdqOgFUCjICtCheNRcA9gQIFCki+\nfLwnd0+YkhFIK0ACTuvBPQQQQAABBBIiQAJOCDOVIIAAAgggkFaABJzWg3sIBFZgzJgxMnfu3MD2\nn44jkGgBTvgkWpz6EPCowOrVqz3aMpqFgD8FGAH7M670CgEEEEDA4wIkYI8HiOYhgAACCPhTgClo\nf8aVXiGQY4Hrr79eKleunOPjOAABBGITIAHH5sZRCPhOoH379pI/f37f9YsOIeBVAaagvRoZ2oUA\nAggg4GsBErCvw0vnEEAAAQS8KkAC9mpkaBcCCRbYvn27/PTTTwmuleoQCK4ACTi4safnCKQReOGF\nF2TatGlp1nEHAQTcEyABu2dLyQgggAACCGQqQALOlIYNCCCAAAIIuCfAx5Dcs81xyevWrZPTp09n\nOO78+fOSkpJi/mXYaNmKCxcumD765Xdnz507J3nz5rUsCpGbqz9HeOjQIVmwYEHkHSxaq48z/Zcn\njz/GGKdOnfLNbzX7LTb6GhDrkvILxoVYD+a4+AqcOXMmvgV6sLSZM2fK+++/L1OnTvVg64LdpAcf\nfFBq1qwpPXr0CDaEB3tfr149mThxotStW9eDrQtuk3bt2iUNGzaUH3/8MSYERsAxsblzUBC+BEHf\n7508eZIvfHDnIZSrUnX2Rd/NB+FxmCuoJBx87NgxM5onNknAz6JKnf06evRoFntkvckf8zNZ95Gt\nCCCAAAIIeE6ABOy5kNAgBBBAAIEgCDAFHYQoe6iPzZo1M+cZPdQkmvJvAT0HXLx4cTw8KPD666/L\n7373Ow+2LNhNKleunOh1LbEuXIQVqxzHIYAAAgggkAsBpqBzgcehCCCAAAIIxCpAAo5VjuMQQAAB\nBBDIhQAJOBd4HBqdwIgRI6R58+Zyxx13yJEjRyIetHfvXrnmmmsibmNl/AWyi0l22+PfIkpUAf0o\nWPfu3aVJkyby7LPPZoqyZMkS+dOf/pTpdjbEX+CHH34Q/c3spk2byrvvvpuhgrVr18pdd90l7dq1\nkzlz5mTYHmkFCTiSCuviJvCPf/xDvvjiC1m4cKH5wPrgwYMjlv2Xv/xFtm3bFnEbK+MrkF1Mstse\n39ZQWrjASy+9JLVq1ZKPP/5Y1q9fL4sWLQrfbG4fOHBA+vTpI/qmlSVxAj179pQnn3xSZs+eLQMH\nDpSDBw+mqfzpp5+WRx55RGbMmGHePOn3HWS3kICzE2J7rgQ+//xzadu2reTLl8+MgD/55JMM5Y0b\nN05atGghJUuWzLCNFfEXyC4m2W2Pf4so0RH47LPPpFOnTubLUDp27CiRni/6ZrVfv37OIfxNkMDu\n3bvlqquuktTUVGncuLGsWLEiTc2lSpWSZcuWyZo1a0S/1VBf87JbSMDZCbE9VwI7d+6UMmXKmDL0\nAfrzzz+nKU/f5f/v//4v02lpVNy9k11MstvubuuCXXq4fenSpWXfvn1pQCZPnmy+jrJOnTpp1nPH\nXQGddShUqFCokkix0annl19+2ZxCaN26dVTfQ04CDpFyI14C+r21NWrUkEcffdSMap2vajt+/Ljo\n5+bClwceeMDsO2bMGNEHub7AsLgroDMNWcUku+3uti7YpYfb69dPli9fPgSiU546Bao/mjFlyhTZ\nunWr6LlgFvcF9PPx+vrlLOljc/bsWenVq5esWrXKjIz1dNqnn37q7J7pXxJwpjRsiFXg66+/Nuev\nhg0bZqZsli9fborSc4tXXHFFmmL1nIn+AIC+o9RfrilRokSa7dyJv4BOo2UVk+y2x79FlOgIZGWv\n3wP9wgsvmClQfZ7o/SJFijiH8tdFAZ1OLly4sJnB0++z19c4PVcfvlSoUCH0y2gaF90/u4Uv4shO\niO25EtAv9+/du7fo1NqePXvMLyFVrFjRJF0911i2bNlQ+fqA1isJWdwVyCwmOhvRqlUradOmTcSY\nudsqSlcBfZ7oOd4TJ06Y5Dpr1izzM6T6JjX8+bJx40b57//+b/N8Qi4xAjqi1TdAenGVTjHrhXA6\n0tUro7/66iuZPn26uTpafzpWX+NGjRqVbcNIwNkSsUM8BPQFJZp3hPGoizKiE8guJtltj64W9opF\nAPtY1BJzjCbg8PPB4bXq6Fh/uzmz7eH76m0ScHoR7iOAAAIIIJAAAc4BJwCZKhBAAAEEEEgvQAJO\nL8J9BBBAAAEEEiBAAk4AMlUggAACCCCQXoAEnF6E+wggEBcB/YjZ+PHjQ2XpF67cc889ofvcQCDo\nAiTgoD8C6D8CLgk0aNBA3nzzzVDp06ZNE/2SFhYEEPh/Aa6C5pGAAAKuCOg3B+nnIfXrRvUbnS66\n6CL59ttvzV9XKqRQBCwTyP7boi3rEM1FAAFvCOi3Ad14443mp9kuueQS8+UrmoRZEEDg/wVIwDwS\nEEDANYHbbrtNxo4dK5qAO3fu7Fo9FIyAjQJMQdsYNdqMgCUC+q1Av/nNb6RYsWKydOnSND8uYEkX\naCYCrgkwAnaNloIRQKBgwYLm+6X1O47Df9kHGQQQEOEqaB4FCCDgqoD++EOXLl1crYPCEbBRgClo\nG6NGmxGwQOC7774znwNetGiRrFy5UnQ0zIIAAr8KMAL+1YJbCCAQR4HU1FSpUqWKLFiwgOQbR1eK\n8o8AI2D/xJKeIIAAAghYJMAI2KJg0VQEEEAAAf8IkID9E0t6ggACCCBgkQAJ2KJg0VQEEEAAAf8I\nkID9E0t6ggACCCBgkQAJ2KJg0VQEEEAAAf8IkID9E0t6ggACCCBgkcD/Aewx2GOlw/iZAAAAAElF\nTkSuQmCC\n"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%R\n",
    "\n",
    "twoway.replicates %>%\n",
    "  group_by(term) %>%\n",
    "  summarise(y = mean(estimate), lcl = quantile(estimate, 0.025), ucl = quantile(estimate, 0.975)) %>%\n",
    "  ggplot(aes(x = term, y = y, ymin = lcl, ymax = ucl)) +\n",
    "  geom_pointrange() +\n",
    "  geom_hline(yintercept = 0.0, linetype = 'dashed') +\n",
    "  coord_flip() +\n",
    "  theme_bw()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "celltoolbar": "Slideshow",
  "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": "2.7.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}