diff --git a/main.py b/main.py
index 4e4f453..219050d 100644
--- a/main.py
+++ b/main.py
@@ -98,7 +98,7 @@ def _extract(element):
     logger.info('Extracting titles and urls...')
     headlines_df = pd.DataFrame(list(map(_extract, elements)))
     res = headlines_df.dropna().drop_duplicates()
-    logger.info(f"Scraped {len(res)} from NYTimes.com.")
+    logger.info(f"Scraped {len(res)} from NY Times.")
     return res
 
 
@@ -136,7 +136,7 @@ def _extract(element):
     logger.info('Extracting titles and urls...')
     headlines_df = pd.DataFrame(list(map(_extract, elements)))
     res = headlines_df.dropna().drop_duplicates()
-    logger.info(f'Scraped {len(res)} from NYTimes.com.')
+    logger.info(f'Scraped {len(res)} from Washington Post.')
     return res
 
 
@@ -168,12 +168,11 @@ def main():
         headlines.to_csv(os.path.join(args.dir, args.file).format(NOW),
                          index=False, encoding='utf-8')
     else:
-        db_url = os.environ['DATABASE_URL']
+        db_url = os.environ['HEROKU_POSTGRESQL_CHARCOAL_URL']
         engine = create_engine(db_url)
         current = pd.read_sql_query('SELECT * FROM headlines;', engine)
-        mask = ~(headlines['headline'].isin(current['headline']) | ~(
-            headlines['url'].isin(current['url'])))
-        toappend = headlines.loc[mask]
+        in_current = headlines['headline'].isin(current['headline'])
+        toappend = headlines.loc[~in_current]
         if len(current) == 0:
             toappend = headlines
         toappend['timestamp'] = NOW
diff --git a/requirements.txt b/requirements.txt
index 6f484cf..cab2514 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -9,3 +9,9 @@ psycopg2
 requests
 bs4
 lxml
+dash
+dash-renderer
+dash-html-components
+dash-core-components
+plotly
+networkx==2.1
diff --git a/viz.ipynb b/viz.ipynb
new file mode 100644
index 0000000..755a5c6
--- /dev/null
+++ b/viz.ipynb
@@ -0,0 +1,3111 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "***Get headlines from Heroku:***"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1257,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "2,859 raw headlines\n",
+      "0 duplicates\n",
+      "(2859, 3)\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>text</th>\n",
+       "      <th>pub</th>\n",
+       "      <th>url</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>As China Plans Big Concessions on Trade, Some ...</td>\n",
+       "      <td>nyt</td>\n",
+       "      <td>https://www.nytimes.com/2018/05/17/us/politics...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>China, Signaling Thaw With U.S. Over Trade, Ap...</td>\n",
+       "      <td>nyt</td>\n",
+       "      <td>https://www.nytimes.com/2018/05/17/business/de...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>Qatar-Linked Firm Nears Deal to Bail Out Kushn...</td>\n",
+       "      <td>nyt</td>\n",
+       "      <td>https://www.nytimes.com/2018/05/17/nyregion/ku...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>Filing Shows Trump Paid Cohen, Who Paid Stormy...</td>\n",
+       "      <td>nyt</td>\n",
+       "      <td>https://www.nytimes.com/2018/05/16/us/politics...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>Trump Rants, Calling Some Unauthorized Immigra...</td>\n",
+       "      <td>nyt</td>\n",
+       "      <td>https://www.nytimes.com/2018/05/16/us/politics...</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "                                                text  pub  \\\n",
+       "0  As China Plans Big Concessions on Trade, Some ...  nyt   \n",
+       "1  China, Signaling Thaw With U.S. Over Trade, Ap...  nyt   \n",
+       "2  Qatar-Linked Firm Nears Deal to Bail Out Kushn...  nyt   \n",
+       "3  Filing Shows Trump Paid Cohen, Who Paid Stormy...  nyt   \n",
+       "4  Trump Rants, Calling Some Unauthorized Immigra...  nyt   \n",
+       "\n",
+       "                                                 url  \n",
+       "0  https://www.nytimes.com/2018/05/17/us/politics...  \n",
+       "1  https://www.nytimes.com/2018/05/17/business/de...  \n",
+       "2  https://www.nytimes.com/2018/05/17/nyregion/ku...  \n",
+       "3  https://www.nytimes.com/2018/05/16/us/politics...  \n",
+       "4  https://www.nytimes.com/2018/05/16/us/politics...  "
+      ]
+     },
+     "execution_count": 1257,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "import os\n",
+    "import pandas as pd\n",
+    "from sqlalchemy import create_engine\n",
+    "import matplotlib.pyplot as plt\n",
+    "import seaborn as sns\n",
+    "%matplotlib inline\n",
+    "import string\n",
+    "sns.set_style('white')\n",
+    "from IPython.display import display\n",
+    "\n",
+    "\n",
+    "db_url = os.environ['DATABASE_URL']\n",
+    "headlines = pd.read_sql_query('SELECT index, headline AS text, pub, url FROM headlines;', create_engine(db_url), index_col='index')\n",
+    "print(f'{len(headlines):,} raw headlines')\n",
+    "headlines['text'].sample(5, random_state=42).head().values.tolist()\n",
+    "\n",
+    "# Deduplication:\n",
+    "print(f'{len(headlines) - len(headlines.drop_duplicates()):,} duplicates')\n",
+    "headlines.drop_duplicates(inplace=True)\n",
+    "print(headlines.shape)\n",
+    "headlines.reset_index(inplace=True, drop=True)\n",
+    "headlines.head()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1258,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "['Trump’s next round of trade limits could hurt the tech industry he wants to help',\n",
+       " 'Mimi Swartz: Things Have Changed Since Sandy Hook',\n",
+       " 'Dusting Off an Art Deco Feminist Treasure',\n",
+       " 'Why bring politics into Armed Forces Day? Because this year, our servicemen and women deserve it',\n",
+       " 'Meet the donors pouring millions into the 2018 elections']"
+      ]
+     },
+     "execution_count": 1258,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "headlines['text'].sample(5, random_state=0).values.tolist()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Do different sources have different vocabularies?"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Some patterns:\n",
+    "\n",
+    "* Source:paraphrase|quote, e.g., \"Ex-FBI official: Trump was right to fire Comey\"\n",
+    "    * Can we extract quotes/paraphrases and their source?\n",
+    "* State abbreviations, e.g., \"Md.\" for Maryland\n",
+    "* Lots of long apposatives, e.g., \"Md. Democratic candidate for governor Valerie Ervin ...\""
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "***Data cleaning:***"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1259,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>text</th>\n",
+       "      <th>pub</th>\n",
+       "      <th>url</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>2722</th>\n",
+       "      <td>Special Report With Bret Baier - Monday, May 21</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>//video.foxnews.com/v/5787858151001/?playlist_...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2747</th>\n",
+       "      <td>The Story w/ Martha MacCallum - Monday, May 21</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>//video.foxnews.com/v/5787865611001/?playlist_...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>590</th>\n",
+       "      <td>Fox News @ Night - Thursday, May 17</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>//video.foxnews.com/v/5786474684001/?playlist_...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>591</th>\n",
+       "      <td>The Ingraham Angle - Thursday, May 17</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>//video.foxnews.com/v/5786457745001/?playlist_...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1768</th>\n",
+       "      <td>The Next Revolution w/Steve Hilton – Sunday, M...</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>//video.foxnews.com/v/5787483762001/?playlist_...</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "                                                   text  pub  \\\n",
+       "2722    Special Report With Bret Baier - Monday, May 21  fox   \n",
+       "2747     The Story w/ Martha MacCallum - Monday, May 21  fox   \n",
+       "590                 Fox News @ Night - Thursday, May 17  fox   \n",
+       "591               The Ingraham Angle - Thursday, May 17  fox   \n",
+       "1768  The Next Revolution w/Steve Hilton – Sunday, M...  fox   \n",
+       "\n",
+       "                                                    url  \n",
+       "2722  //video.foxnews.com/v/5787858151001/?playlist_...  \n",
+       "2747  //video.foxnews.com/v/5787865611001/?playlist_...  \n",
+       "590   //video.foxnews.com/v/5786474684001/?playlist_...  \n",
+       "591   //video.foxnews.com/v/5786457745001/?playlist_...  \n",
+       "1768  //video.foxnews.com/v/5787483762001/?playlist_...  "
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/plain": [
+       "(2828, 3)"
+      ]
+     },
+     "execution_count": 1259,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# Drop headlines about TV shows, e.g., \"Hannity - Monday, May 16\"\n",
+    "display(headlines.loc[headlines.text.str.contains(r'\\S+\\s* [-–—] \\S+\\s*, May .+')].sample(5))\n",
+    "headlines = headlines.loc[~headlines.text.str.contains(r'\\S+\\s* [-–—] \\S+\\s*, May .+')]\n",
+    "headlines.shape"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1260,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "C:\\Anaconda3\\envs\\media\\lib\\site-packages\\ipykernel_launcher.py:3: UserWarning: This pattern has match groups. To actually get the groups, use str.extract.\n",
+      "  This is separate from the ipykernel package so we can avoid doing imports until\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>text</th>\n",
+       "      <th>pub</th>\n",
+       "      <th>url</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>2026</th>\n",
+       "      <td>Monarchs in My Garden, at Last 5:45 AM ET</td>\n",
+       "      <td>nyt</td>\n",
+       "      <td>https://www.nytimes.com/2018/05/21/opinion/pol...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2590</th>\n",
+       "      <td>Dozens Die in Karachi from Relentless Heat 5:2...</td>\n",
+       "      <td>nyt</td>\n",
+       "      <td>https://www.nytimes.com/2018/05/21/world/karac...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2183</th>\n",
+       "      <td>Monarchs in My Garden, at Last 10:13 AM ET</td>\n",
+       "      <td>nyt</td>\n",
+       "      <td>https://www.nytimes.com/2018/05/21/opinion/pol...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2833</th>\n",
+       "      <td>Dozens Die in Karachi From Relentless Heat 12:...</td>\n",
+       "      <td>nyt</td>\n",
+       "      <td>https://www.nytimes.com/2018/05/21/world/karac...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2506</th>\n",
+       "      <td>How to Vote in Venezuela’s (Non) Election 3:20...</td>\n",
+       "      <td>nyt</td>\n",
+       "      <td>https://www.nytimes.com/2018/05/21/opinion/ven...</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "                                                   text  pub  \\\n",
+       "2026          Monarchs in My Garden, at Last 5:45 AM ET  nyt   \n",
+       "2590  Dozens Die in Karachi from Relentless Heat 5:2...  nyt   \n",
+       "2183         Monarchs in My Garden, at Last 10:13 AM ET  nyt   \n",
+       "2833  Dozens Die in Karachi From Relentless Heat 12:...  nyt   \n",
+       "2506  How to Vote in Venezuela’s (Non) Election 3:20...  nyt   \n",
+       "\n",
+       "                                                    url  \n",
+       "2026  https://www.nytimes.com/2018/05/21/opinion/pol...  \n",
+       "2590  https://www.nytimes.com/2018/05/21/world/karac...  \n",
+       "2183  https://www.nytimes.com/2018/05/21/opinion/pol...  \n",
+       "2833  https://www.nytimes.com/2018/05/21/world/karac...  \n",
+       "2506  https://www.nytimes.com/2018/05/21/opinion/ven...  "
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/plain": [
+       "(2770, 3)"
+      ]
+     },
+     "execution_count": 1260,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# Strip a common time pattern on video headlines\n",
+    "time_pat = f'(\\s\\d+:\\d+ (PM|AM) \\w+[{string.punctuation}]*\\s*)'\n",
+    "display(headlines[headlines.text.str.contains(time_pat)].sample(5))\n",
+    "headlines['text'] = headlines['text'].str.replace(pat, '')\n",
+    "strip_pat(time_pat)\n",
+    "headlines.drop_duplicates(inplace=True)\n",
+    "headlines.shape"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1261,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "(2770, 3)"
+      ]
+     },
+     "execution_count": 1261,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# This could interfere with pos_tag\n",
+    "# Remove punctuation, except for the hyphen\n",
+    "# headlines['text'] = headlines['text'].str.replace(f'[{string.punctuation.replace(\"-\", \"\")}]' , '')\n",
+    "# headlines.drop_duplicates(inplace=True)\n",
+    "headlines.shape"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1262,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "(2768, 3)"
+      ]
+     },
+     "execution_count": 1262,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# Standardize quotation marks:\n",
+    "headlines['text'] = headlines['text'].str.replace(r'[”’\\'\"“‘]', '\\'')\n",
+    "headlines.drop_duplicates(inplace=True)\n",
+    "headlines.shape"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1263,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>text</th>\n",
+       "      <th>pub</th>\n",
+       "      <th>url</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>776</th>\n",
+       "      <td>Follow us on Twitter »</td>\n",
+       "      <td>nyt</td>\n",
+       "      <td>https://twitter.com/nytopinion</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "                       text  pub                             url\n",
+       "776  Follow us on Twitter »  nyt  https://twitter.com/nytopinion"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/plain": [
+       "(2768, 3)"
+      ]
+     },
+     "execution_count": 1263,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "display(headlines[headlines.text.str.contains('»')])\n",
+    "headlines['text'] = headlines['text'].str.replace('»', '')\n",
+    "headlines.drop_duplicates(inplace=True)\n",
+    "headlines.shape"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1264,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>text</th>\n",
+       "      <th>pub</th>\n",
+       "      <th>url</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>1329</th>\n",
+       "      <td>`: Dusting Off an Art Deco Feminist Treasure</td>\n",
+       "      <td>nyt</td>\n",
+       "      <td>https://www.nytimes.com/2018/05/16/nyregion/du...</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "                                              text  pub  \\\n",
+       "1329  `: Dusting Off an Art Deco Feminist Treasure  nyt   \n",
+       "\n",
+       "                                                    url  \n",
+       "1329  https://www.nytimes.com/2018/05/16/nyregion/du...  "
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "display(headlines[headlines.text.str.contains('`')])\n",
+    "headlines = headlines[~headlines.text.str.contains('`')]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1265,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>text</th>\n",
+       "      <th>pub</th>\n",
+       "      <th>url</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>1160</th>\n",
+       "      <td>'Just saying ¯\\_(ツ)_/¯' — BBC Three compares r...</td>\n",
+       "      <td>wapo</td>\n",
+       "      <td>https://www.washingtonpost.com/news/worldviews...</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "                                                   text   pub  \\\n",
+       "1160  'Just saying ¯\\_(ツ)_/¯' — BBC Three compares r...  wapo   \n",
+       "\n",
+       "                                                    url  \n",
+       "1160  https://www.washingtonpost.com/news/worldviews...  "
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "display(headlines[headlines.text.str.contains('ツ')])\n",
+    "headlines = headlines[~headlines.text.str.contains('ツ')]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1266,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>text</th>\n",
+       "      <th>pub</th>\n",
+       "      <th>url</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>1285</th>\n",
+       "      <td>Hits &amp; Misses: 5/19/18</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>//video.foxnews.com/v/5787182621001/</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "                        text  pub                                   url\n",
+       "1285  Hits & Misses: 5/19/18  fox  //video.foxnews.com/v/5787182621001/"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "display(headlines[headlines.text.str.contains('/')])\n",
+    "headlines = headlines[~headlines.text.str.contains('/')]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1267,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>text</th>\n",
+       "      <th>pub</th>\n",
+       "      <th>url</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>2805</th>\n",
+       "      <td>Not at the Game*: Courtside With Run TMC, Watc...</td>\n",
+       "      <td>nyt</td>\n",
+       "      <td>https://www.nytimes.com/2018/05/21/sports/gold...</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "                                                   text  pub  \\\n",
+       "2805  Not at the Game*: Courtside With Run TMC, Watc...  nyt   \n",
+       "\n",
+       "                                                    url  \n",
+       "2805  https://www.nytimes.com/2018/05/21/sports/gold...  "
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "display(headlines[headlines.text.str.contains('\\*')])\n",
+    "headlines = headlines[~headlines.text.str.contains('\\*')]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1268,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>text</th>\n",
+       "      <th>pub</th>\n",
+       "      <th>url</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>142</th>\n",
+       "      <td>Lincoln vs. Obama - The incredible tale of two...</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>http://www.foxnews.com/opinion/2018/05/16/tamm...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>143</th>\n",
+       "      <td>America - We must remember how to respectfully...</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>http://www.foxnews.com/opinion/2018/05/17/cal-...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1005</th>\n",
+       "      <td>GOP pours cash into bid to repeal California g...</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>//www.foxnews.com/politics/2018/05/19/gop-pour...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1118</th>\n",
+       "      <td>Royal caught lounging at wedding - and Twitter...</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>http://www.foxnews.com/entertainment/2018/05/1...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2196</th>\n",
+       "      <td>'Legends &amp; Lies – Abraham Lincoln: The Campaign'</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>//video.foxnews.com/v/5787444706001/?playlist_...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2242</th>\n",
+       "      <td>Jesus banned - School censors valedictorian's ...</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>http://www.foxnews.com/opinion/2018/05/21/todd...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2326</th>\n",
+       "      <td>Disgraced Dem Conyers endorsed son for his own...</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>http://www.foxnews.com/politics/2018/05/21/dis...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2334</th>\n",
+       "      <td>Disgraced Dem Conyers endorsed son for his own...</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>//www.foxnews.com/politics/2018/05/21/disgrace...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2362</th>\n",
+       "      <td>Disgraced Democrat forced out of House endorse...</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>http://www.foxnews.com/politics/2018/05/21/dis...</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "                                                   text  pub  \\\n",
+       "142   Lincoln vs. Obama - The incredible tale of two...  fox   \n",
+       "143   America - We must remember how to respectfully...  fox   \n",
+       "1005  GOP pours cash into bid to repeal California g...  fox   \n",
+       "1118  Royal caught lounging at wedding - and Twitter...  fox   \n",
+       "2196   'Legends & Lies – Abraham Lincoln: The Campaign'  fox   \n",
+       "2242  Jesus banned - School censors valedictorian's ...  fox   \n",
+       "2326  Disgraced Dem Conyers endorsed son for his own...  fox   \n",
+       "2334  Disgraced Dem Conyers endorsed son for his own...  fox   \n",
+       "2362  Disgraced Democrat forced out of House endorse...  fox   \n",
+       "\n",
+       "                                                    url  \n",
+       "142   http://www.foxnews.com/opinion/2018/05/16/tamm...  \n",
+       "143   http://www.foxnews.com/opinion/2018/05/17/cal-...  \n",
+       "1005  //www.foxnews.com/politics/2018/05/19/gop-pour...  \n",
+       "1118  http://www.foxnews.com/entertainment/2018/05/1...  \n",
+       "2196  //video.foxnews.com/v/5787444706001/?playlist_...  \n",
+       "2242  http://www.foxnews.com/opinion/2018/05/21/todd...  \n",
+       "2326  http://www.foxnews.com/politics/2018/05/21/dis...  \n",
+       "2334  //www.foxnews.com/politics/2018/05/21/disgrace...  \n",
+       "2362  http://www.foxnews.com/politics/2018/05/21/dis...  "
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/plain": [
+       "(2764, 3)"
+      ]
+     },
+     "execution_count": 1268,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# Replace incorrectly-used hyphens with an m-dash\n",
+    "display(headlines[headlines.text.str.contains(' [-–] ')])\n",
+    "headlines['text'] = headlines.text.str.replace(' [-–] ', ' — ')\n",
+    "headlines.drop_duplicates(inplace=True)\n",
+    "headlines.shape"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1269,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "(2763, 3)"
+      ]
+     },
+     "execution_count": 1269,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "headlines['text'] = headlines.text.str.replace('\\s+', ' ')\n",
+    "headlines.drop_duplicates(inplace=True)\n",
+    "headlines.shape"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1270,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "(2763, 3)"
+      ]
+     },
+     "execution_count": 1270,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "headlines['text'] = headlines['text'].str.strip()\n",
+    "headlines.drop_duplicates(inplace=True)\n",
+    "headlines.shape"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1271,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "{'!',\n",
+       " '#',\n",
+       " '$',\n",
+       " '%',\n",
+       " '&',\n",
+       " \"'\",\n",
+       " '(',\n",
+       " ')',\n",
+       " '+',\n",
+       " ',',\n",
+       " '-',\n",
+       " '.',\n",
+       " ':',\n",
+       " ';',\n",
+       " '=',\n",
+       " '?',\n",
+       " '|',\n",
+       " 'á',\n",
+       " 'í',\n",
+       " '—'}"
+      ]
+     },
+     "execution_count": 1271,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "set(re.sub(pattern=f'[{string.ascii_letters}\\s\\d]', repl='', string=' '.join(headlines.text)))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1272,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>text</th>\n",
+       "      <th>pub</th>\n",
+       "      <th>url</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>1556</th>\n",
+       "      <td>California man wins more than $6 million from ...</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>//www.foxnews.com/us/2018/05/20/california-man...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1555</th>\n",
+       "      <td>California man wins 4 lottery prizes worth $6M...</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>http://www.foxnews.com/us/2018/05/20/californi...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2153</th>\n",
+       "      <td>Wrecked $250,000 Ferrari found without a driver</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>http://video.foxnews.com/v/5787683904001/</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>739</th>\n",
+       "      <td>Steven Tyler reveals he once hit on Cameron Di...</td>\n",
+       "      <td>fox</td>\n",
+       "      <td>http://www.foxnews.com/entertainment/2018/05/1...</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>98</th>\n",
+       "      <td>$800,000 Homes in Missouri, Georgia and Pennsy...</td>\n",
+       "      <td>nyt</td>\n",
+       "      <td>https://www.nytimes.com/2018/05/16/realestate/...</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "                                                   text  pub  \\\n",
+       "1556  California man wins more than $6 million from ...  fox   \n",
+       "1555  California man wins 4 lottery prizes worth $6M...  fox   \n",
+       "2153    Wrecked $250,000 Ferrari found without a driver  fox   \n",
+       "739   Steven Tyler reveals he once hit on Cameron Di...  fox   \n",
+       "98    $800,000 Homes in Missouri, Georgia and Pennsy...  nyt   \n",
+       "\n",
+       "                                                    url  \n",
+       "1556  //www.foxnews.com/us/2018/05/20/california-man...  \n",
+       "1555  http://www.foxnews.com/us/2018/05/20/californi...  \n",
+       "2153          http://video.foxnews.com/v/5787683904001/  \n",
+       "739   http://www.foxnews.com/entertainment/2018/05/1...  \n",
+       "98    https://www.nytimes.com/2018/05/16/realestate/...  "
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/plain": [
+       "(43, 3)"
+      ]
+     },
+     "execution_count": 1272,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# References to money\n",
+    "display(headlines[headlines.text.str.contains('\\$')].sample(5))\n",
+    "headlines[headlines.text.str.contains('\\$')].shape"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1273,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "array([[\"It's the 50th anniversary of the day Trump left college and (briefly) faced the draft\",\n",
+       "        'wapo',\n",
+       "        'https://www.washingtonpost.com/news/retropolis/wp/2018/05/20/its-the-50th-anniversary-of-the-day-trump-left-college-and-nearly-had-to-go-to-war/'],\n",
+       "       ['Lawyers for Michael Cohen and Stormy Daniels Sling More (Legal) Mud',\n",
+       "        'nyt',\n",
+       "        'https://www.nytimes.com/2018/05/18/nyregion/michael-cohen-stormy-daniels-legal-fight.html'],\n",
+       "       ['Bitcoin Goes Mainstream (in a Hilton Ballroom)', 'nyt',\n",
+       "        'https://www.nytimes.com/2018/05/18/technology/consensus-bitcoin-conference-photos.html'],\n",
+       "       ['We dropped $18,000 worth of phones, and these are the toughest (and weakest)',\n",
+       "        'fox',\n",
+       "        '//www.foxnews.com/tech/2018/05/21/dropped-18000-worth-phones-and-these-are-toughest-and-weakest.html'],\n",
+       "       ['Step-by-step | How to decorate a layer cake (like a royal)',\n",
+       "        'wapo',\n",
+       "        'http://www.washingtonpost.com/graphics/2018/voraciously/amp-stories/decorating-a-layer-cake-like-a-royal-step-by-step/']],\n",
+       "      dtype=object)"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/plain": [
+       "(21, 3)"
+      ]
+     },
+     "execution_count": 1273,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# Parentheticals\n",
+    "display(headlines[headlines.text.str.contains('\\(')].sample(5).values)\n",
+    "headlines[headlines.text.str.contains('\\(')].shape"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1274,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "array([['Smug Seattle to mom and pop landlords: Criminals are welcome! Your rights not so much',\n",
+       "        'fox',\n",
+       "        '//www.foxnews.com/opinion/2018/05/21/smug-seattle-to-mom-and-pop-landlords-criminals-are-welcome-your-rights-not-so-much.html'],\n",
+       "       ['Here comes the bride! Meghan, Harry arrive', 'fox',\n",
+       "        'http://www.foxnews.com/entertainment/2018/05/19/meghan-markle-prince-harrys-royal-wedding-guests-begin-arriving-couples-new-titles-revealed.html'],\n",
+       "       [\"15-foot great white shark shocks scuba divers above Florida Keys shipwreck: 'Get on the boat!'\",\n",
+       "        'fox',\n",
+       "        'http://www.foxnews.com/science/2018/05/21/15-foot-great-white-shark-shocks-scuba-divers-above-florida-keys-shipwreck-get-on-boat.html'],\n",
+       "       [\"Trump Says Russia Inquiry's Overreach Is a Scandal 'Bigger Than Watergate!'\",\n",
+       "        'nyt',\n",
+       "        'https://www.nytimes.com/2018/05/17/us/politics/trump-watergate-russia-investigation.html'],\n",
+       "       [\"'My next call is to ICE!': Man's rant against workers who spoke Spanish at a Manhattan deli caught on camera\",\n",
+       "        'wapo',\n",
+       "        'https://www.washingtonpost.com/news/business/wp/2018/05/16/my-next-call-is-to-ice-watch-a-man-wig-out-because-workers-spoke-spanish-at-a-manhattan-deli/']],\n",
+       "      dtype=object)"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/plain": [
+       "(12, 3)"
+      ]
+     },
+     "execution_count": 1274,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# Excalamations\n",
+    "display(headlines[headlines.text.str.contains('!')].sample(5).values)\n",
+    "headlines[headlines.text.str.contains('!')].shape"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "***Visualizations:***"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1276,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 432x288 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "ax = headlines.groupby('pub').apply(lambda grp: grp['text'].unique().shape[0])[['nyt', 'fox', 'wapo']].plot(kind='bar', title='Count unique headlines by pub')\n",
+    "ax.set(ylabel='count')\n",
+    "[tick.set_rotation(0) for tick in ax.get_xticklabels()]\n",
+    "plt.tight_layout()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1277,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 432x288 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "ax = headlines.groupby('pub').apply(lambda grp: grp['url'].unique().shape[0])[['nyt', 'fox', 'wapo']].plot(kind='bar', title='Count unique headlines by pub')\n",
+    "ax.set(ylabel='count')\n",
+    "[tick.set_rotation(0) for tick in ax.get_xticklabels()]\n",
+    "plt.tight_layout()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1279,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "pub\n",
+       "fox     1.073657\n",
+       "nyt     1.506726\n",
+       "wapo    1.379061\n",
+       "dtype: float64"
+      ]
+     },
+     "execution_count": 1279,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "ct_headlines = headlines.groupby('pub').apply(lambda grp: grp['text'].unique().shape[0])\n",
+    "ct_urls = headlines.groupby('pub').apply(lambda grp: grp['url'].unique().shape[0])\n",
+    "ct_headlines / ct_urls"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Does Fox have this many real headlines, or do they just swap out headlines more often? Interestingly, Fox actually has _fewer_ headlines per URL, suggesting that WaPo and NYT change headlines mid-publication more often."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "ax = headlines['pub'].value_counts()[['nyt', 'fox', 'wapo']].plot(kind='bar', title='Count unique headlines by pub')\n",
+    "ax.set(ylabel='count')\n",
+    "[tick.set_rotation(0) for tick in ax.get_xticklabels()]\n",
+    "plt.tight_layout()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Headline length:\n",
+    "\n",
+    "WaPo tends to have longer headlines, but only because they have more words (rather than using longer words)."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1101,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "headlines['len_words'] = headlines['text'].str.split('\\s').apply(len)\n",
+    "headlines['len_chars'] = headlines['text'].apply(len)\n",
+    "headlines['len_avg_word'] = headlines['text'].str.replace(f'[{string.punctuation}]' , '').str.split('\\s').apply(lambda l: sum(map(len, l)) / len(l))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1215,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 720x360 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "sns.distplot(headlines.loc[headlines['pub']=='nyt', 'len_words'], label='nyt', hist=False, kde_kws={'shade': True})\n",
+    "sns.distplot(headlines.loc[headlines['pub']=='fox', 'len_words'], label='fox', hist=False, kde_kws={'shade': True})\n",
+    "ax = sns.distplot(headlines.loc[headlines['pub']=='wapo', 'len_words'], label='wapo', hist=False, kde_kws={'shade': True})\n",
+    "\n",
+    "ax.axvline(headlines.loc[headlines['pub']=='nyt', 'len_words'].mean(), color='blue')\n",
+    "ax.axvline(headlines.loc[headlines['pub']=='fox', 'len_words'].mean(), color='green')\n",
+    "ax.axvline(headlines.loc[headlines['pub']=='wapo', 'len_words'].mean(), color='red')\n",
+    "\n",
+    "ax.legend()\n",
+    "ax.axes.get_yaxis().set_ticks([])\n",
+    "ax.set(title='headline length in words by publication', ylabel='density')\n",
+    "ax.figure.set_size_inches((10,5))\n",
+    "plt.tight_layout()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1108,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Examples of headlines with many words:\n",
+      "['Woman who helped concoct fake story about Muslim men groping women at German bar found dead'\n",
+      " 'Two sons of Rev. Moon have split from his church — and their followers are armed'\n",
+      " \"Critic's Notebook: On That 'Barry' Finale and Why Some Shows Are So Good, They Need to End\"\n",
+      " \"Once Hated by U.S. and Tied to Iran, Is Sadr Now 'Face of Reform' in Iraq?\"\n",
+      " 'The truth is out there: The real story of Area 51 may be weirder than the science fiction']\n",
+      "\n",
+      "Examples of headlines with average words:\n",
+      "[\"A father's secret wish for mothers that get little credit\"\n",
+      " 'D.C. United shakes Earthquakes with early goals in 3-1 victory'\n",
+      " 'Judge in Stanford rape case is subject of recall vote'\n",
+      " 'Bernard Lewis, Influential Scholar of Islam, Is Dead at 101'\n",
+      " \"Texans defensive end J.J. Watt to pay for victims' funerals\"]\n",
+      "\n",
+      "Examples of headlines with few words:\n",
+      "['Mortgage Calculator' \"Venezuela's sham election\" 'Death of a Biohacker'\n",
+      " 'Trump demands DOJ probe' 'The Royal Wedding']\n"
+     ]
+    }
+   ],
+   "source": [
+    "print('Examples of headlines with many words:')\n",
+    "print(headlines.loc[headlines['len_words'] > 15, 'text'].sample(5).values)\n",
+    "\n",
+    "print()\n",
+    "print('Examples of headlines with average words:')\n",
+    "print(headlines.loc[headlines['len_words'] == 10, 'text'].sample(5).head().values)\n",
+    "\n",
+    "print()\n",
+    "print('Examples of headlines with few words:')\n",
+    "print(headlines.loc[headlines['len_words'] < 5, 'text'].sample(5).head().values)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1103,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 720x360 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "sns.distplot(headlines.loc[headlines['pub']=='nyt', 'len_chars'], label='nyt', hist=False, kde_kws={'shade': True})\n",
+    "sns.distplot(headlines.loc[headlines['pub']=='fox', 'len_chars'], label='fox', hist=False, kde_kws={'shade': True})\n",
+    "ax = sns.distplot(headlines.loc[headlines['pub']=='wapo', 'len_chars'], label='wapo', hist=False, kde_kws={'shade': True})\n",
+    "\n",
+    "ax.axvline(headlines.loc[headlines['pub']=='nyt', 'len_chars'].mean(), color='blue')\n",
+    "ax.axvline(headlines.loc[headlines['pub']=='fox', 'len_chars'].mean(), color='green')\n",
+    "ax.axvline(headlines.loc[headlines['pub']=='wapo', 'len_chars'].mean(), color='red')\n",
+    "\n",
+    "ax.legend()\n",
+    "ax.axes.get_yaxis().set_ticks([])\n",
+    "ax.set(title='headline length in chars by publication', ylabel='density')\n",
+    "ax.figure.set_size_inches((10,5))\n",
+    "plt.tight_layout()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1109,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Examples of headlines with many chars:\n",
+      "['MUST SEE: Nail salon worker jumps onto moving car to stop customer from leaving without paying her bill'\n",
+      " \"At Texas vigil, another school mourns another mass shooting: 'It is hope in the face of tragedy that brings us together today'\"\n",
+      " \"HBO star Shannon Woodward makes shocking return to 'Westworld' after storming social media with anti-Trump insults\"\n",
+      " \"Meghan Markle, Prince Harry's royal wedding: Groom arrives with brother Prince William, waves to crowds\"\n",
+      " 'Cops focus on teen boy previously caught with Pennsylvania teacher in steamed-up car in murder investigation']\n",
+      "\n",
+      "Examples of headlines with average chars:\n",
+      "[\"Don't let the Trump administration derail criminal-justice reform\"\n",
+      " 'Heisman winner Manziel signs, will play with new team this summer'\n",
+      " 'Camping without the hassle: Rent the gear you need — even the car'\n",
+      " 'Loose Ends: The Only Houseplants You Can Be Trusted With: A Guide'\n",
+      " 'Trump will nominate the acting secretary to head Veterans Affairs']\n",
+      "\n",
+      "Examples of headlines with few chars:\n",
+      "['Mortgage Calculator' 'The Royal Wedding' 'Enough Is Enough'\n",
+      " \"It's Spring, Right?\" 'A.I. Is Hard']\n"
+     ]
+    }
+   ],
+   "source": [
+    "print('Examples of headlines with many chars:')\n",
+    "print(headlines.loc[headlines['len_chars'] > 100, 'text'].sample(5).values)\n",
+    "\n",
+    "print()\n",
+    "print('Examples of headlines with average chars:')\n",
+    "print(headlines.loc[headlines['len_chars'] == 65, 'text'].sample(5).head().values)\n",
+    "\n",
+    "print()\n",
+    "print('Examples of headlines with few chars:')\n",
+    "print(headlines.loc[headlines['len_chars'] < 20, 'text'].sample(5).head().values)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1113,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "array([['A.I. Is Hard', 'nyt', 3, 12, 2.6666666666666665]], dtype=object)"
+      ]
+     },
+     "execution_count": 1113,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "headlines[headlines['len_chars'] == headlines['len_chars'].min()].values"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1112,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "array([['Prisons are allowing mothers to raise their babies behind bars. But is the radical experiment in parenting and punishment a good idea?',\n",
+       "        'wapo', 22, 134, 5.045454545454546]], dtype=object)"
+      ]
+     },
+     "execution_count": 1112,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "headlines[headlines['len_chars'] == headlines['len_chars'].max()].values"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1104,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 720x360 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "sns.distplot(headlines.loc[headlines['pub']=='nyt', 'len_avg_word'], label='nyt', hist=False, kde_kws={'shade': True})\n",
+    "sns.distplot(headlines.loc[headlines['pub']=='fox', 'len_avg_word'], label='fox', hist=False, kde_kws={'shade': True})\n",
+    "ax = sns.distplot(headlines.loc[headlines['pub']=='wapo', 'len_avg_word'], label='wapo', hist=False, kde_kws={'shade': True})\n",
+    "\n",
+    "ax.axvline(headlines.loc[headlines['pub']=='nyt', 'len_avg_word'].mean(), color='blue')\n",
+    "ax.axvline(headlines.loc[headlines['pub']=='fox', 'len_avg_word'].mean(), color='green')\n",
+    "ax.axvline(headlines.loc[headlines['pub']=='wapo', 'len_avg_word'].mean(), color='red')\n",
+    "\n",
+    "ax.legend()\n",
+    "ax.axes.get_yaxis().set_ticks([])\n",
+    "ax.set(title='avg word length by publication', ylabel='density')\n",
+    "ax.figure.set_size_inches((10,5))\n",
+    "plt.tight_layout()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1114,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Examples of headlines with long avg word length:\n",
+      "['State regulators unveil nationwide crackdown on suspicious cryptocurrency investment schemes'\n",
+      " 'Progressive Democrats outperforming moderates in primaries'\n",
+      " 'Mortgage Calculator' 'Monarchy Meets African-American Experience'\n",
+      " \"Researchers answer long-running rumors surrounding Hitler's death\"]\n",
+      "\n",
+      "Examples of headlines with avg avg word length:\n",
+      "[\"Meghan Markle, Prince Harry marry at St. George's Chapel\"\n",
+      " 'Trump pushes for prison reform bill'\n",
+      " 'Prince Charles Walks a Radiant Meghan Markle Down the Aisle'\n",
+      " 'T.J. Oshie, Braden Holtby carry Capitals to Game 7 against Lightning'\n",
+      " 'Fans at the Royal Wedding, and Their Accessories']\n",
+      "\n",
+      "Examples of headlines with short avg word length:\n",
+      "['A Staten Island Man Found a Safe of Cash in His Backyard. Then Things Got Weird.'\n",
+      " 'Why the GOP needs someone — anyone — to challenge Trump in 2020'\n",
+      " 'Hawaii lava flow ramps up as new magma mixes with old'\n",
+      " 'Veteran: I had five hour wait at VA last week'\n",
+      " 'How a Digital Rabbit Hole Gave a 1983 Album a New Life']\n"
+     ]
+    }
+   ],
+   "source": [
+    "print('Examples of headlines with long avg word length:')\n",
+    "print(headlines.loc[headlines['len_avg_word'] > 8, 'text'].sample(5).values)\n",
+    "\n",
+    "print()\n",
+    "print('Examples of headlines with avg avg word length:')\n",
+    "print(headlines.loc[headlines['len_avg_word'] == 5, 'text'].sample(5).values)\n",
+    "\n",
+    "print()\n",
+    "print('Examples of headlines with short avg word length:')\n",
+    "print(headlines.loc[headlines['len_avg_word'] < 4, 'text'].sample(5).head().values)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1117,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Ten random headlines:\n"
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "['Controversial armored police vehicle factory expands production to meet demand',\n",
+       " 'The Latest: Hawaii lava flow may cross highway within hours',\n",
+       " 'T.J. Oshie, Braden Holtby carry Capitals to Game 7 against Lightning',\n",
+       " \"Woman claims Mario Batali drugged, sexually assaulted her inside NYC restaurant's 'rape room'; NYPD investigating allegations\",\n",
+       " \"'World's most expensive Witch Hunt': Trump lashes out at New York Times, Democrats\",\n",
+       " 'What to Pack in Your New York City Survival Kit',\n",
+       " 'Venezuelan President Nicolás Maduro wins reelection amid charges of irregularities',\n",
+       " 'Lawmakers are nudging Twitter CEO Jack Dorsey to testify before Congress on privacy',\n",
+       " 'Baby hit by softball opening eyes, moving legs',\n",
+       " 'Crisis-weary Venezuelans are voting in election decried as a Maduro power grab']"
+      ]
+     },
+     "execution_count": 1117,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "print('Ten random headlines:')\n",
+    "headlines.sample(10, random_state=0)['text'].values.tolist()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "***Play with NLTK and TextBlob:***"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1193,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<Text: As China Plans Big Concessions on Trade, Some...>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/plain": [
+       "TextBlob(\"As China P\")"
+      ]
+     },
+     "execution_count": 1193,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "from nltk import Text\n",
+    "from textblob import TextBlob\n",
+    "\n",
+    "txt = Text('. '.join(headlines['text']).split())\n",
+    "display(txt)\n",
+    "\n",
+    "blob = TextBlob('. '.join(headlines['text']))\n",
+    "blob[:10]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1221,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Meghan Markle; Prince Harry; Meghan Markle,; school shooting; Prince\n",
+      "Harry's; Meghan Markle's; royal wedding; Markle, Prince; NBA\n",
+      "playoffs:; royal wedding.; high school; Texas school; White House; New\n",
+      "York; Naval Academy; Royal Wedding; Lightning-Capitals Game; North\n",
+      "Korea; Academy freshmen; school shooting.\n"
+     ]
+    }
+   ],
+   "source": [
+    "txt.collocations()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1237,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "NOT REAL NEWS: False comments attributed to Zimbabwe leader\n"
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "['NOT',\n",
+       " 'REAL',\n",
+       " 'NEWS',\n",
+       " ':',\n",
+       " 'False',\n",
+       " 'comments',\n",
+       " 'attributed',\n",
+       " 'to',\n",
+       " 'Zimbabwe',\n",
+       " 'leader']"
+      ]
+     },
+     "execution_count": 1237,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "from nltk.tokenize import RegexpTokenizer\n",
+    "tokenizer = RegexpTokenizer('\\w+|\\$[\\d\\.]+|\\S+')\n",
+    "b\n",
+    "ex = headlines.text.sample(random_state=3).values[0]\n",
+    "print(ex)\n",
+    "tokenizer.tokenize(ex)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1241,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "1142          poses\n",
+       "1044            ohr\n",
+       "1000    negotiating\n",
+       "1121        planned\n",
+       "1212         reagan\n",
+       "dtype: object"
+      ]
+     },
+     "execution_count": 1241,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# from https://www.nltk.org/book/ch02.html#wordlist-corpora\n",
+    "def unusual_words(text):\n",
+    "    text_vocab = set(w.lower() for w in text if w.isalpha())\n",
+    "    english_vocab = set(w.lower() for w in nltk.corpus.words.words())\n",
+    "    unusual = text_vocab - english_vocab\n",
+    "    return sorted(unusual)\n",
+    "\n",
+    "pd.Series(unusual_words(txt)).sample(5, random_state=0)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1246,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "False\n",
+      "True\n"
+     ]
+    }
+   ],
+   "source": [
+    "print('abortions' in nltk.corpus.words.words())\n",
+    "print('abortion' in nltk.corpus.words.words())"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1247,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "0.7330637724338511"
+      ]
+     },
+     "execution_count": 1247,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "def content_fraction(text):\n",
+    "    stopwords = nltk.corpus.stopwords.words('english')\n",
+    "    content = [w for w in text if w.lower() not in stopwords]\n",
+    "    return len(content) / len(text)\n",
+    "\n",
+    "content_fraction(txt)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1253,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 432x288 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "headlines.groupby('pub').apply(lambda grp: content_fraction(Text(' '.join(grp['text']).split())))[['nyt', 'fox', 'wapo']].plot(kind='bar', title='content fraction by publication')\n",
+    "plt.tight_layout()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1238,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "[('NOT', 'NNP'),\n",
+       " ('REAL', 'NNP'),\n",
+       " ('NEWS', 'NNP'),\n",
+       " (':', ':'),\n",
+       " ('False', 'JJ'),\n",
+       " ('comments', 'NNS'),\n",
+       " ('attributed', 'VBD'),\n",
+       " ('to', 'TO'),\n",
+       " ('Zimbabwe', 'NNP'),\n",
+       " ('leader', 'NN')]"
+      ]
+     },
+     "execution_count": 1238,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "from nltk import pos_tag\n",
+    "\n",
+    "pos_tag(tokenizer.tokenize(ex))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1216,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 432x288 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "vocab = headlines.groupby('pub').apply(lambda grp: len(set(' '.join(grp['text']).split())))[['nyt', 'fox', 'wapo']]\n",
+    "ax = vocab.plot(title='vocabulary size by publication', kind='bar')\n",
+    "[tick.set_rotation(0) for tick in ax.get_xticklabels()]\n",
+    "ax.set(ylabel='unique words')\n",
+    "plt.tight_layout()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1254,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<matplotlib.axes._subplots.AxesSubplot at 0x1c908c97b70>"
+      ]
+     },
+     "execution_count": 1254,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWkAAAEiCAYAAAA75CPhAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAGqhJREFUeJzt3XtU1HX+x/EXznBRwQQVuqBtXrfVzPKSimbltbwgaki6eNzNOtkpj22WN1IXb7W6/kTcfrpd1nMIAy8IXkpL0CUv4aU105O6WMcCQTFRQNQBZn5/GPMDlYslzCd5Pv5ymO8M7y+jTz7zne+Mbg6HwyEAgJHquXoAAEDFiDQAGIxIA4DBiDQAGIxIA4DBiDQAGIxI14IXXnhB6enprh7jV4mMjFR0dLSrxyhnyJAhSktLc/UYtyQ/P1/jxo276XXTpk3TBx98UKvzpKWlaciQIVXOExwcrLy8vF/0PXbu3KmoqChJUnJysubNm/fLhq2jrK4eoC547733XD0CDHHx4kV98803rh7jliUlJf3i237zzTe6ePGiJKlv377q27fv7RqrTiDStyAtLU1z587V5s2bb7gcHR2tzMxM5eTkKDMzUwEBAVq0aJH8/f311FNPKSoqSg899JCioqK0adMm+fr6qkuXLjpy5IhiYmI0bdo0tWnTRs8//7wklbt85swZRUZGKisrS0VFRRo8eLBeeumlcrNt375dH374oVavXi1JGjhwoAYPHqxJkyYpOztbo0aNUmpqqlJSUrR8+XLZ7XY1bNhQ06dPV8eOHRUdHa1Dhw7p7NmzateunebMmaOZM2fq2LFj8vf3l8ViUefOnSVJq1evVlxcnNzd3eXp6anIyEi1bt263DzR0dE6deqUsrOzlZOTo9///veaP3++vL29K9yfjIwMjR07Vq1atVJmZqZiYmLk7+/vvM/09HTNmDFDly9fVsuWLVVYWFhu/2+2X8XFxVq0aJF27twpi8WiRx55RLNnz9bKlSuVm5urWbNmOectvRweHq727dvr0KFDOn/+vEJDQ3Xu3Dnt27dPly9f1tKlS9WuXTvl5+dr/vz5OnHihIqKitSjRw+9+eabslqteuihh/Tiiy9q9+7dOnv2rCZMmKAxY8Zo+vTpunLlioKDg5WQkCCLxVLu53bw4EFt27ZNBQUFCgoK0tSpU/XJJ584f+aSdPr0aYWGhiolJUUeHh7O206bNk2enp46duyYfvrpJwUFBSkiIkLu7u5q166d9u7dKz8/P0lyXpakwsJCTZo0SadOnVKjRo0UGRmpBx54oNxcZW+/cuVKbdiwQVarVffff7/efvttWSwWzZkzR6dOndKFCxfUsGFDLV68WPn5+YqLi1NJSYl8fHx0//33a9u2bVq5cqWys7M1Z84cZWZmyuFwaPjw4ZowYYIyMjI0fvx49enTR19//bXy8vL0xhtvqH///tX+t3on4XDHbXTgwAFFRUVp69atql+/vvMfVanPPvtMn332mRITE7V69epqHwJ54403NHLkSCUkJGjdunXas2ePPvnkk3Lb9OrVS8ePH1deXp4yMjJ06dIl7dmzR9K1p5j9+vXT999/r9mzZys6OlobN27UpEmT9PLLL6ugoECSlJmZqQ0bNmjx4sVatmyZvLy8tHXrVkVFRen777+XJJWUlGjBggV6//33tX79eoWGhurgwYM3nXv//v1aunSpPv30U1mtVv3jH/+ocn+ys7P18ssva9u2beUCLUlTpkzRs88+q02bNmncuHE6ffq0JOnkyZMV7tfq1at19OhRJSUlafPmzbp06dINP7ubyczMVFxcnBYtWqRFixapW7duSkhIUO/evfXRRx9JkhYsWKD27dsrISFBiYmJys3N1b/+9S9Jks1mk6+vr+Li4rRs2TItXLhQV69e1cKFC+Xl5aWkpKQbAl26/6tWrVJiYqKOHTumNWvWaNCgQfrhhx/03//+V5K0du1ahYSElAt0qcOHD+vDDz/UJ598opMnTyo+Pr7Kfc3KytL48eOVlJSkIUOG6M0336xw2+TkZCUkJCg+Pl6bN29WYGCgPvroI6WmpqpRo0aKj4/Xtm3b1KFDB8XGxurhhx9WWFiYnnnmGb322mvl7mvKlCl67LHHtGnTJn388cfauHGjtmzZIkn68ccf1atXL61bt06vv/66FixYUOV+3KmI9G3UrVs3eXt7S5L+8Ic/OJ/ilfryyy/Vv39/eXt7y93dXaNHj67yPgsLC7V//35FRUUpODhYoaGhysrK0rFjx8pt5+XlpZ49e2r37t1KTU3V6NGjlZGRofz8fKWkpGjgwIH68ssv1b17dzVv3lyS1KNHD/n5+enIkSOSpE6dOslqvfbkau/evRo+fLjc3Nzk5+fnXMVYLBYNGjRIYWFhioyMVKNGjTRq1Kibzj5o0CA1bdpU9erV06hRo7Rr164q98dqtapTp0433Fdubq6OHz+u4cOHS5I6d+6sNm3aOH+uFe3Xnj17FBwcLC8vL9WrV09Lly513kdlSve39D579+4tSWrRooXzcd25c6fi4+MVHBysESNG6PDhwzpx4oTzPkqf1rdv3142m63cyr8iwcHBatCggTw8PDRs2DDt2bNHHh4eevbZZ7V27VqVlJRow4YNCg0NventQ0JC1LBhQ3l4eCg4OFi7du2q8nu2a9dOjz76qPP2R44cUX5+/k233bt3rwYNGqS77rpLkjR9+nRNnDhRgwYNUkhIiGJiYjRv3jzt27ev0v0tLCzUV199pbFjx0qSfHx8NGLECKWmpkqS3N3d1adPH0nX/i1duHChyv24U3G44xa4ubmp7EedFBUVlbvey8urwm0lydPTs9zX3N3dq7xvu90uh8OhuLg41a9fX5J0/vx5eXp63jBfv379lJqaqry8PE2YMEHfffedtm/frhMnTqhbt25KT0+Xm5tbuds4HA4VFxdLkho0aHDDdaXKrvoWL16sEydOaM+ePfrnP/+ppKQk5wtDZZW9jd1uV7169Srdn9zcXHl4eDh/UdxM2ZlKt7Pb7RXu1/X3de7cOef2lT2W169Syz5WZfcpKipKrVq1kiTl5eWVm6P0MSr9WnU+Jqfsz8zhcDjnDwsL06hRo9StWze1adPG+cujqtvXq3fjOsxms5W7fP02bm5uFT4GFoul3D7m5eUpLy9PqampWrNmjcaOHauhQ4eqcePGysjIqHA/S/8eXP+10r+L7u7uzrmuf2zrGlbSt8DPz0+nT5/WTz/9JIfD4XxqVl1PPPGEtm7dqosXL8putysxMdF5na+vr3NFe+bMGe3bt0+S5O3trU6dOjmfRufl5em5555TcnLyDff/1FNPae/evfr222/VsWNHBQUFKSoqSo8//rgsFot69OihXbt26ccff5R0bVWUlZWlhx9++Ib76t27t9atWye73a6LFy86v9/58+fVp08fNW7cWOPHj9fkyZMrfCEsOTlZ+fn5stvtWrNmjZ588slb2p+yfH191b59e61du1aSdPToUeeqtbL96tGjhzZv3iybzSa73a45c+Zoy5Yt8vX11dGjR+VwOFRQUKAdO3ZU+v1vplevXlq1apUcDodsNpsmTpzoPBRSEavVqpKSkgqDvWXLFtlsNl29elUbNmzQ448/Lkm655571KlTJy1YsEDPPfdchff/6aeflrv9k08+Kena393Sx6n0NZVSx48f17fffitJio+PV+fOnZ2/QK/Xs2dPff75585DZNHR0Vq1apV27dqlkJAQPfvss3rggQeUkpKikpISSdfCXhrfUt7e3nr44YcVGxsr6dpZL4mJierZs2fFP7w6ipX0LWjdurXCwsI0cuRINWvWTE888cQtvVL/2GOPady4cRozZow8PT113333Oa8LDw/XlClTNHDgQAUGBqp79+7O6xYvXqy5c+dq6NChstlsGjJkiIYNG3bD/fv4+KhVq1aqX7++LBaLevfurZkzZ2rAgAHO+WfPnq1XXnlFJSUl8vLy0ooVK+Tj43PDfb366quaPXu2nn76afn5+alt27aSrv1jnzhxosaPHy8vLy9ZLJYKT6lq2rSpXnjhBeXm5qpr167OFzsr2p/KVl6StGTJEk2fPl1xcXFq0aKFWrZsWeV+hYWFKTMzUyNGjJDD4VC3bt0UHh6uy5cv64svvtCAAQMUEBCgbt26VWulW9bMmTM1f/58DR06VEVFRerZs6cmTJhQ6W2aNWumjh07avDgwYqNjZWvr2+56wMDAzVmzBhdunRJ/fv3V0hIiPO6ESNGaO7cuc7DADfj5eWlMWPGKC8vTwMHDtTIkSMlSREREc7DUz179lSzZs2ct2nZsqWWL1+uH3/8UU2aNNHbb79d4f336dNH6enpzl8UrVu31ty5c3Xs2DHNmjVL69atk3Tt0FnpL9Hu3btrypQpmjt3rtq3b++8r8WLFysyMlIJCQmy2WwaOnSoRowYoczMzEp/hnWNGx9V6jpbt25VbGysYmJiXD3KbVf2bAn8ena7XZGRkbr33nv14osv3nSb688Qwp2Bwx2A4QoKCvTYY48pKyurwjfC4M7FShoADMZKGgAMdltfOLxy5YqOHDmiZs2a3fREfQDAjUpKSpSTk6MOHTqUO5VXus2RPnLkiPPkdADArYmNjVWXLl3Kfe22Rrr0tJ7Y2Fjdfffdt/OuAeCOlZ2drbFjx5Y7NbLUbY106SGOu+++W4GBgbfzrgHgjnezw8S8cAgABiPSAGAwIg0ABiPSAGAwIg0ABiPSAGAwIg0ABiPSAGAwPvQfwC+yO3ikq0eoUUFJ6109giRW0gBgNCINAAYj0gBgMCINAAYj0gBgMCINAAYj0gBgMCINAAYj0gBgMCINAAYj0gBgMCINAAYj0gBgMCINAAYj0gBgMCINAAYj0gBgMCINAAYj0gBgMCINAAYj0gBgsGpF+qefflKfPn108uTJmp4HAFBGlZEuKirSrFmz5OXlVRvzAADKqDLS77zzjsLCwuTv718b8wAAyqg00gkJCfLz81Pv3r1rax4AQBmVRnr9+vXas2ePwsPD9e2332rq1KnKycmprdkAoM6zVnZlbGys88/h4eGaM2eOmjVrVuNDAQCu4RQ8ADBYpSvpsmJiYmpyDgDATbCSBgCDEWkAMBiRBgCDEWkAMFi1XzgEbrfQ+ImuHqFGrRn9v64eAXcAVtIAYDAiDQAGI9IAYDAiDQAGI9IAYDAiDQAG+82fgjf09SRXj1CjNv092NUjAHAhVtIAYDAiDQAGI9IAYDAiDQAGI9IAYDAiDQAGI9IAYDAiDQAGI9IAYDAiDQAGI9IAYDAiDQAGI9IAYDAiDQAGI9IAYDAiDQAGI9IAYDAiDQAGI9IAYDAiDQAGI9IAYDAiDQAGI9IAYDAiDQAGI9IAYDAiDQAGs1a1QUlJiSIiIvT999/LYrFo4cKFatGiRW3MBgB1XpUr6R07dkiS4uLiNGnSJC1cuLDGhwIAXFPlSrpfv3564oknJEmnT59W06ZNa3omAMDPqoy0JFmtVk2dOlWff/65li1bVtMzAQB+Vu0XDt955x1t27ZNb731lgoLC2tyJgDAz6qMdGJiolauXClJql+/vtzc3GSxWGp8MABANQ53DBgwQNOnT9fYsWNVXFysGTNmyNPTszZmA4A6r8pIN2jQQFFRUbUxCwDgOryZBQAMRqQBwGBEGgAMRqQBwGBEGgAMRqQBwGBEGgAMRqQBwGBEGgAMRqQBwGBEGgAMRqQBwGBEGgAMRqQBwGBEGgAMRqQBwGBEGgAMRqQBwGBEGgAMRqQBwGBEGgAMRqQBwGBEGgAMRqQBwGBEGgAMRqQBwGBEGgAMRqQBwGBEGgAMRqQBwGBEGgAMRqQBwGBEGgAMRqQBwGBEGgAMRqQBwGBEGgAMZq3syqKiIs2YMUOZmZmy2WyaOHGi+vbtW1uzAUCdV2mkN27cqMaNG2vRokXKzc1VSEgIkQaAWlRppAcNGqSBAwc6L1sslhofCADw/yqNdMOGDSVJBQUFmjRpkiZPnlwrQwEArqnyhcOsrCyNGzdOwcHBGjp0aG3MBAD4WaUr6XPnzunPf/6zZs2apR49etTWTACAn1W6kl6xYoXy8vL07rvvKjw8XOHh4bpy5UptzQYAdV6lK+mIiAhFRETU1iwAgOvwZhYAMBiRBgCDEWkAMBiRBgCDEWkAMBiRBgCDEWkAMBiRBgCDEWkAMBiRBgCDEWkAMBiRBgCDEWkAMBiRBgCDEWkAMBiRBgCDEWkAMBiRBgCDEWkAMBiRBgCDEWkAMBiRBgCDEWkAMBiRBgCDEWkAMBiRBgCDEWkAMBiRBgCDEWkAMBiRBgCDEWkAMBiRBgCDEWkAMBiRBgCDEWkAMBiRBgCDEWkAMBiRBgCDVSvSX3/9tcLDw2t6FgDAdaxVbfDee+9p48aNql+/fm3MAwAoo8qVdIsWLRQdHV0bswAArlNlpAcOHCirtcoFNwCgBvDCIQAYjEgDgMGINAAYrFqRDgwM1Jo1a2p6FgDAdVhJA4DBiDQAGIxIA4DBiDQAGIxIA4DBiDQAGIxIA4DBiDQAGIxIA4DBiDQAGIxIA4DBiDQAGIxIA4DBiDQAGIxIA4DBiDQAGIxIA4DBiDQAGIxIA4DBiDQAGIxIA4DBiDQAGIxIA4DBiDQAGIxIA4DBiDQAGIxIA4DBiDQAGIxIA4DBiDQAGIxIA4DBiDQAGIxIA4DBiDQAGIxIA4DBiDQAGIxIA4DBrFVtYLfbNWfOHB0/flweHh6aN2+e7r///tqYDQDqvCpX0tu3b5fNZlN8fLxef/11vf3227UxFwBA1VhJHzx4UL1795YkderUSUeOHKlw25KSEklSdnb2bRqvakWF52vte7lCRkaGq0eoMbbcy64eoUbdyY+dJJ2z2Vw9Qo2qzcevtJmlDS2rykgXFBTI29vbedlisai4uFhW6403zcnJkSSNHTv2Fw+L8vqm8Mzlt6rv//R19Qj4NfrW/uOXk5Nzw+HkKiPt7e2tS5cuOS/b7fabBlqSOnTooNjYWDVr1kwWi+VXjgsAdUNJSYlycnLUoUOHG66rMtKPPvqoduzYoWeeeUaHDh1S27ZtK9zWy8tLXbp0+XXTAkAdVNEJGW4Oh8NR2Q1Lz+44ceKEHA6HFixYoFatWtXIkACA8qqMNADAdXgzCwAYjEgDgMGINAAYjEgDgMGIdDV888035S7v27fPRZPgVpU9x1+SsrKyXDQJ8MtwdkclDhw4oPT0dK1atUp/+tOfJF076Xz16tXavHmzi6dDdQwbNkyLFi1Su3bttG3bNi1dulSffvqpq8dCNZSUlCguLk7p6en63e9+p+eee04eHh6uHqvWVflmlrqsUaNGOnfunGw2m/Mt725ubnrjjTdcPBmqa8mSJZo5c6aaNGkiq9Wq2NhYV4+Eanrrrbfk4+OjoKAg7du3TxEREfrb3/7m6rFqHZGuRNu2bdW2bVsFBgZqyJAhFb4dHuYqfaJos9nk7u7OxxX8hpw6dcr5S7Vfv34KCwtz8USuwTHpakhPT9fIkSP1zjvv6OTJk64eB7dg8uTJioiI0Pvvv6+nn35aY8aMcfVIqKarV6/q8uVrn5R45cqVm35CXF3AMelqstvtSk1N1fr165WTk6PQ0FANGzaM1bXhSj/F8fz582rcuLGys7N17733unosVMPGjRu1fPlytWnTRunp6Zo0aZIGDx7s6rFqHZGuBofDoS+++EIJCQn64YcfNGzYMBUXF+vAgQNasWKFq8dDJdLS0jRjxgx5e3srPz9fc+fOVVBQkKvHQjVduHBBGRkZCgwMVOPGjV09jkuwDKyGAQMGqEuXLho3bpweffRR59c59GG+pUuXavXq1QoICNCZM2f0yiuvEOnfiK+++kp//etfde7cOQUEBGj+/Pl68MEHXT1WrSPS1RAeHq6PP/643PnRycnJWrhwoQunQnVYLBYFBARIkgICAuTp6eniiVBd8+bN09///ne1bt1aJ06c0KxZsxQXF+fqsWodka6G+Ph4rVixQvfcc4+rR0E15efny8fHR97e3oqJiVHXrl21f/9+3XXXXa4eDdXk4+Oj1q1bS7p2ppWXl5eLJ3INIl0NzZs3539I/4156aWXFBsbK39/f2VlZWnp0qVq2bIlz35+Q5o0aaKZM2eqe/fuOnr0qOx2u+Lj4yVJo0ePdvF0tYcXDqth8uTJKigo0IMPPig3NzdJ0l/+8hcXT4XKPP/887pw4YJOnTpV7j+pcHNzq5NPmX+Lli9fXuF1r7zySi1O4lpEuho2bNhww9dCQkJcMAmqy2636+zZs5o1a5Zmz55d7rr77rvPRVPhVp09e1bFxcVyOBw6e/asHnnkEVePVOuINAAjzZgxQ4cOHdLly5d15coVNW/eXGvWrHH1WLWOdxwCMNJ3332nLVu2qFevXtqyZUudPTOHSAMwUoMGDeTm5qbCwkL5+fmpqKjI1SO5BJEGYKQOHTrogw8+kL+/v1577TUVFxe7eiSX4BQ8AEbatWuXAgMDNXjwYHl4eKhjx46uHskleOEQgLFOnjyp5ORkpaSkqGnTppWelnenYiUNwEjHjh3T7t27lZaWJklq2bKliydyDVbSAIzUuXNnNW/eXK+99pr69Onj6nFchkgDMFJxcbEOHjyoXbt26fDhw2rSpImWLFni6rFqHWd3ADBSXl6ezpw5o9OnT+vKlSt19j9rYCUNwEgjRoxQv3791L9/f7Vp08bV47gMkQYAg3G4AwAMRqQBwGBEGnVWu3btXD0CUCUiDQAG4x2HuGOkpaXp3XffldVqVUZGhjp27KiJEydqwoQJSklJkSRFR0dLkl599VVJ0ltvvaXDhw/L19dXCxYsqLOnecFcrKRxR/nPf/6jmTNnauvWrbp69ar+/e9/V7p9165dlZSUpP79+2v+/Pm1NCVQfUQad5SuXbuqZcuWcnNzU3BwsL788ssKt/Xy8tKwYcMkScHBwdq3b19tjQlUG5HGHcVisTj/7HA4VFhYqLJvBSj7mcT16tUrt63VytE/mIdI445y8OBBnTlzRna7XYmJierXr58uXLig8+fPy2az6YsvvnBuW1hYqOTkZEnS+vXr1bNnT1eNDVSIpQPuKP7+/nrzzTd15swZBQUF6Y9//KMKCgo0atQo3X333XrooYec2zZq1Ejbt29XVFSUAgICtHDhQhdODtwcbwvHHSMtLU3Lly9XTEyMq0cBbhsOdwCAwVhJA4DBWEkDgMGINAAYjEgDgMGINAAYjEgDgMH+Dwbrt2fUMlg9AAAAAElFTkSuQmCC\n",
+      "text/plain": [
+       "<Figure size 432x288 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "(vocab / headlines.groupby('pub').count()['text'])[['nyt', 'fox', 'wapo']].plot(title='unique words per document by publication', kind='bar')"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Fox has a larger overall vocabulary, but only because it publishes more headlines in general. WaPo has the most diverse vocabulary on a per-headline basis because they tend to use more words."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1133,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "245"
+      ]
+     },
+     "execution_count": 1133,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "txt.count('Trump')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1134,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "31"
+      ]
+     },
+     "execution_count": 1134,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "txt.count('Mueller')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1135,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "14"
+      ]
+     },
+     "execution_count": 1135,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "txt.count('America')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1142,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "93"
+      ]
+     },
+     "execution_count": 1142,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "txt.count('wedding')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1143,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "535"
+      ]
+     },
+     "execution_count": 1143,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "txt.count('the')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1144,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "42"
+      ]
+     },
+     "execution_count": 1144,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "txt.count('an')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1145,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "331"
+      ]
+     },
+     "execution_count": 1145,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "txt.count('a')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1146,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "469"
+      ]
+     },
+     "execution_count": 1146,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "txt.count('of')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1147,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "82"
+      ]
+     },
+     "execution_count": 1147,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "txt.count('by')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1148,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "173"
+      ]
+     },
+     "execution_count": 1148,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "txt.count('with')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1157,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "28194"
+      ]
+     },
+     "execution_count": 1157,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "len(txt)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1166,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "his u.s. watch entire russia long alleged resume yanny likely gop's go\n",
+      "official football point surveilled 'westworld'\n"
+     ]
+    }
+   ],
+   "source": [
+    "txt.similar('Trump')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1176,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "the_inquiry\n"
+     ]
+    }
+   ],
+   "source": [
+    "txt.common_contexts(['Trump', 'Russia'])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1177,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "russia mueller's fbi\n"
+     ]
+    }
+   ],
+   "source": [
+    "txt.similar('Mueller')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1178,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "the_probe\n"
+     ]
+    }
+   ],
+   "source": [
+    "txt.common_contexts(['Mueller', 'Russia'])"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "WordNet:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1287,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "[Synset('car.n.01'),\n",
+       " Synset('car.n.02'),\n",
+       " Synset('car.n.03'),\n",
+       " Synset('car.n.04'),\n",
+       " Synset('cable_car.n.01')]"
+      ]
+     },
+     "execution_count": 1287,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "from nltk.corpus import wordnet as wn\n",
+    "\n",
+    "car = wn.synsets('car')\n",
+    "car"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1303,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "[Synset('motor_vehicle.n.01')]"
+      ]
+     },
+     "execution_count": 1303,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "first = car[0]\n",
+    "first.hypernyms()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1306,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "'NOT REAL NEWS: False comments attributed to Zimbabwe leader'"
+      ]
+     },
+     "execution_count": 1306,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "ex"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1313,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "[Synset('news.n.01'),\n",
+       " Synset('news.n.02'),\n",
+       " Synset('news_program.n.01'),\n",
+       " Synset('news.n.04'),\n",
+       " Synset('newsworthiness.n.01')]"
+      ]
+     },
+     "execution_count": 1313,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "wn.synsets('NEWS')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1319,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "'NOT REAL NEWS: False comments attributed to Zimbabwe leader'"
+      ]
+     },
+     "execution_count": 1319,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "ex"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1334,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "'hello'"
+      ]
+     },
+     "execution_count": 1334,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "ex_synset = wn.synsets('hello')[0]\n",
+    "ex_synset.name()[:ex_synset.name().find('.')]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1336,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "['hello', 'hullo', 'hi', 'howdy', 'how-do-you-do']"
+      ]
+     },
+     "execution_count": 1336,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "ex_synset.lemma_names()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1335,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "['not',\n",
+       " 'real_number',\n",
+       " 'news',\n",
+       " 'false',\n",
+       " 'remark',\n",
+       " 'impute',\n",
+       " None,\n",
+       " 'zimbabwe',\n",
+       " 'leader']"
+      ]
+     },
+     "execution_count": 1335,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "list(map(lambda w: wn.synsets(w)[0].name()[:wn.synsets(w)[0].name().find('.')] if len(wn.synsets(w)) else None, re.sub(pattern=f'[{string.punctuation}]', repl='', string=ex).split()))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "***Vectorize the headlines:***"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 798,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "showing info https://raw.githubusercontent.com/nltk/nltk_data/gh-pages/index.xml\n"
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "True"
+      ]
+     },
+     "execution_count": 798,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "nltk.download()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1181,
+   "metadata": {},
+   "outputs": [
+    {
+     "ename": "SyntaxError",
+     "evalue": "invalid syntax (<ipython-input-1181-4822f92aecd0>, line 16)",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[1;36m  File \u001b[1;32m\"<ipython-input-1181-4822f92aecd0>\"\u001b[1;36m, line \u001b[1;32m16\u001b[0m\n\u001b[1;33m    from nltk.tag import ???\u001b[0m\n\u001b[1;37m                         ^\u001b[0m\n\u001b[1;31mSyntaxError\u001b[0m\u001b[1;31m:\u001b[0m invalid syntax\n"
+     ]
+    }
+   ],
+   "source": [
+    "import matplotlib.pyplot as plt\n",
+    "import seaborn as sns\n",
+    "sns.set()\n",
+    "%matplotlib inline\n",
+    "import numpy as np\n",
+    "import pandas as pd\n",
+    "\n",
+    "from sklearn.preprocessing import StandardScaler\n",
+    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
+    "\n",
+    "from nltk.stem.snowball import EnglishStemmer\n",
+    "from nltk.stem.lancaster import LancasterStemmer\n",
+    "from nltk.stem.wordnet import WordNetLemmatizer\n",
+    "\n",
+    "from nltk.tokenize import RegexpTokenizer\n",
+    "from nltk.tag import ???\n",
+    "\n",
+    "from nltk.corpus import stopwords\n",
+    "\n",
+    "analyzer_args = dict(        \n",
+    "    stop_words='english', \n",
+    "    ngram_range=(2,2),\n",
+    "    max_df=1.0,\n",
+    "    min_df=1,\n",
+    "    max_features=None,\n",
+    "    binary=False,\n",
+    "    norm=None,\n",
+    "    use_idf=False,\n",
+    "    smooth_idf=False,\n",
+    "    sublinear_tf=False\n",
+    ")\n",
+    "\n",
+    "tokenizer = RegexpTokenizer('\\w+|\\$[\\d\\.]+|\\S+')\n",
+    "def custom_tokenizer(string):\n",
+    "    stop_words = set(stopwords.words('english'))\n",
+    "    tokens = tokenizer(string) \n",
+    "    tokens = [w for w in tokens if not w in stop_words]\n",
+    "    return map(str, pos_tag(tokens, tagset='universal'))\n",
+    "\n",
+    "def custom_analyzer(doc):\n",
+    "    stemmer = WordNetLemmatizer()\n",
+    "    analyzer = TfidfVectorizer(tokenizer=custom_tokenizer, **analyzer_args).build_analyzer()\n",
+    "    return (stemmer.lemmatize(w) for w in analyzer(doc))\n",
+    "\n",
+    "vect = TfidfVectorizer(analyzer=custom_analyzer, tokenizer=custom_tokenizer, **analyzer_args)\n",
+    "\n",
+    "tdmatrix = vect.fit_transform(headlines['text'])\n",
+    "tdmatrix = pd.DataFrame(tdmatrix.toarray(), columns=vect.get_feature_names(), index=headlines.index)\n",
+    "print(tdmatrix.shape)\n",
+    "print(tdmatrix.sum().sort_values(ascending=False).head(10))\n",
+    "tdmatrix.head()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "***Term counts:***"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 800,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<matplotlib.axes._subplots.AxesSubplot at 0x1c6b2ddb208>"
+      ]
+     },
+     "execution_count": 800,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 432x288 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "sns.boxplot(tdmatrix.sum())"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 801,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<matplotlib.axes._subplots.AxesSubplot at 0x1c6ae2e40f0>"
+      ]
+     },
+     "execution_count": 801,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 432x288 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "tdmatrix.sum().sort_values().tail(15).plot(kind='barh')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 791,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<matplotlib.axes._subplots.AxesSubplot at 0x1c6ac638668>"
+      ]
+     },
+     "execution_count": 791,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 432x288 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "tdmatrix.sum().sort_values(ascending=False).head(100).plot(kind='line')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 792,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "C:\\Anaconda3\\envs\\media\\lib\\site-packages\\matplotlib\\axes\\_axes.py:6462: UserWarning: The 'normed' kwarg is deprecated, and has been replaced by the 'density' kwarg.\n",
+      "  warnings.warn(\"The 'normed' kwarg is deprecated, and has been \"\n"
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "<matplotlib.axes._subplots.AxesSubplot at 0x1c6ac695b70>"
+      ]
+     },
+     "execution_count": 792,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 432x288 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "sns.distplot(tdmatrix.sum().sort_values())"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "***Dimensionality reduction:***"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "PCA:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 793,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 1127.25x360 with 3 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "from sklearn.decomposition import PCA\n",
+    "from sklearn.preprocessing import StandardScaler\n",
+    "\n",
+    "scaler = StandardScaler(with_mean=True, with_std=True)\n",
+    "\n",
+    "pca = PCA(n_components=2)\n",
+    "pca_coords = pca.fit_transform(scaler.fit_transform(tdmatrix))\n",
+    "# pca_coords = pca.fit_transform(tdmatrix)\n",
+    "\n",
+    "chartdata = pd.DataFrame(np.concatenate([pca_coords, headlines['pub'].values.reshape((len(headlines),1))], axis=1), index=headlines.index)\n",
+    "chartdata['len'] = headlines['text'].str.split().apply(len)\n",
+    "chartdata.loc[:, 0] = chartdata[0].astype(float)\n",
+    "chartdata.loc[:, 1] = chartdata[1].astype(float)\n",
+    "\n",
+    "chartdata.columns = ['x', 'y', 'pub', 'len']\n",
+    "\n",
+    "ax = sns.lmplot(data=chartdata, x='x',y='y', fit_reg=False, col='pub', hue='len', palette='Blues')"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Isomap:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 794,
+   "metadata": {},
+   "outputs": [
+    {
+     "ename": "ValueError",
+     "evalue": "cannot reshape array of size 1751 into shape (1471,1)",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
+      "\u001b[1;31mValueError\u001b[0m                                Traceback (most recent call last)",
+      "\u001b[1;32m<ipython-input-794-237a42ccc909>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[0;32m      4\u001b[0m \u001b[0mcoords\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0miso\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfit_transform\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtdmatrix\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      5\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 6\u001b[1;33m \u001b[0mchartdata\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mpd\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mDataFrame\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mconcatenate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mcoords\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mheadlines\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'pub'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mreshape\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m1471\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mindex\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mheadlines\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mindex\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      7\u001b[0m \u001b[0mchartdata\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'len'\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mheadlines\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;34m'headline'\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstr\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msplit\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mapply\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      8\u001b[0m \u001b[0mchartdata\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mloc\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mchartdata\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mastype\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfloat\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
+      "\u001b[1;31mValueError\u001b[0m: cannot reshape array of size 1751 into shape (1471,1)"
+     ]
+    }
+   ],
+   "source": [
+    "from sklearn.manifold import Isomap\n",
+    "\n",
+    "iso = Isomap(n_components=2)\n",
+    "coords = iso.fit_transform(tdmatrix)\n",
+    "\n",
+    "chartdata = pd.DataFrame(np.concatenate([coords, headlines['pub'].values.reshape((1471,1))], axis=1), index=headlines.index)\n",
+    "chartdata['len'] = headlines['headline'].str.split().apply(len)\n",
+    "chartdata.loc[:, 0] = chartdata[0].astype(float)\n",
+    "chartdata.loc[:, 1] = chartdata[1].astype(float)\n",
+    "\n",
+    "chartdata.columns = ['x', 'y', 'pub', 'len']\n",
+    "\n",
+    "ax = sns.lmplot(data=chartdata, x='x',y='y', fit_reg=False, hue='len', col='pub', palette='Blues')\n",
+    "# ax.set(xlim=[-15,5], ylim=[-20,40])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 405,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 1127.25x360 with 3 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "from sklearn.manifold import TSNE\n",
+    "\n",
+    "tsne = TSNE(n_components=2, perplexity=5)\n",
+    "# scaler = StandardScaler(with_mean=True, with_std=True)\n",
+    "# tsne_coords = tsne.fit_transform(scaler.fit_transform(td))\n",
+    "tsne_coords = tsne.fit_transform(tdmatrix)\n",
+    "\n",
+    "chartdata = pd.DataFrame(np.concatenate([tsne_coords, headlines['pub'].values.reshape((1471,1))], axis=1), index=headlines.index)\n",
+    "chartdata['len'] = headlines['headline'].str.split().apply(len)\n",
+    "chartdata.loc[:, 0] = chartdata[0].astype(float)\n",
+    "chartdata.loc[:, 1] = chartdata[1].astype(float)\n",
+    "\n",
+    "chartdata.columns = ['x', 'y', 'pub', 'len']\n",
+    "\n",
+    "ax = sns.lmplot(data=chartdata, x='x',y='y', fit_reg=False, hue='len', col='pub', palette='Blues')"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "parts of speech"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 110,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>#</th>\n",
+       "      <th>$</th>\n",
+       "      <th>''</th>\n",
+       "      <th>(</th>\n",
+       "      <th>)</th>\n",
+       "      <th>,</th>\n",
+       "      <th>.</th>\n",
+       "      <th>:</th>\n",
+       "      <th>CC</th>\n",
+       "      <th>CD</th>\n",
+       "      <th>...</th>\n",
+       "      <th>VB</th>\n",
+       "      <th>VBD</th>\n",
+       "      <th>VBG</th>\n",
+       "      <th>VBN</th>\n",
+       "      <th>VBP</th>\n",
+       "      <th>VBZ</th>\n",
+       "      <th>WDT</th>\n",
+       "      <th>WP</th>\n",
+       "      <th>WRB</th>\n",
+       "      <th>``</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>1.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>2.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>1.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>...</td>\n",
+       "      <td>1.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>1.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>1.0</td>\n",
+       "      <td>1.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>1.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>...</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>1.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>1.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "      <td>0.0</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>5 rows × 42 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "     #    $   ''    (    )    ,    .    :   CC   CD ...    VB  VBD  VBG  VBN  \\\n",
+       "0  0.0  0.0  0.0  0.0  0.0  1.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0  0.0   \n",
+       "1  0.0  0.0  0.0  0.0  0.0  2.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  1.0  0.0   \n",
+       "2  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   1.0  0.0  0.0  0.0   \n",
+       "3  0.0  0.0  0.0  0.0  0.0  1.0  0.0  0.0  0.0  0.0 ...   0.0  1.0  1.0  0.0   \n",
+       "4  0.0  0.0  0.0  0.0  0.0  1.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  1.0  0.0   \n",
+       "\n",
+       "   VBP  VBZ  WDT   WP  WRB   ``  \n",
+       "0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
+       "1  0.0  0.0  0.0  0.0  0.0  0.0  \n",
+       "2  0.0  0.0  0.0  0.0  0.0  0.0  \n",
+       "3  0.0  0.0  0.0  0.0  0.0  0.0  \n",
+       "4  1.0  0.0  0.0  0.0  0.0  0.0  \n",
+       "\n",
+       "[5 rows x 42 columns]"
+      ]
+     },
+     "execution_count": 110,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "import nltk\n",
+    "def pos_tag(doc):\n",
+    "    tokenized = nltk.word_tokenize(doc)\n",
+    "    tagged = nltk.pos_tag(tokenized)\n",
+    "    pos = [x[1] for x in tagged]\n",
+    "    return pd.Series(pos).value_counts()\n",
+    "\n",
+    "pos_counts = headlines['headline'].apply(pos_tag).fillna(0)\n",
+    "pos_counts.head()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 111,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<matplotlib.axes._subplots.AxesSubplot at 0x1c6b8536f28>"
+      ]
+     },
+     "execution_count": 111,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 432x288 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "pos_counts.loc[headlines['pub']=='fox'].sum().sort_values().tail(15).plot(kind='barh')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 112,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<matplotlib.axes._subplots.AxesSubplot at 0x1c6b7b50fd0>"
+      ]
+     },
+     "execution_count": 112,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 432x288 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "pos_counts.loc[headlines['pub']=='nyt'].sum().sort_values().tail(15).plot(kind='barh')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 113,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<matplotlib.axes._subplots.AxesSubplot at 0x1c6b7bdd240>"
+      ]
+     },
+     "execution_count": 113,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 432x288 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "pos_counts.loc[headlines['pub']=='wapo'].sum().sort_values().tail(15).plot(kind='barh')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 114,
+   "metadata": {},
+   "outputs": [
+    {
+     "ename": "NameError",
+     "evalue": "name 'tagged' is not defined",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
+      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
+      "\u001b[1;32m<ipython-input-114-b054ae3ac08c>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[1;33m[\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;32mfor\u001b[0m \u001b[0mx\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mtagged\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
+      "\u001b[1;31mNameError\u001b[0m: name 'tagged' is not defined"
+     ]
+    }
+   ],
+   "source": [
+    "[x[1] for x in tagged]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 262,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "\u001b[1;31mSignature:\u001b[0m \u001b[0mnltk\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpos_tag\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtokens\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtagset\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mNone\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlang\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;34m'eng'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
+       "\u001b[1;31mDocstring:\u001b[0m\n",
+       "Use NLTK's currently recommended part of speech tagger to\n",
+       "tag the given list of tokens.\n",
+       "\n",
+       "    >>> from nltk.tag import pos_tag\n",
+       "    >>> from nltk.tokenize import word_tokenize\n",
+       "    >>> pos_tag(word_tokenize(\"John's big idea isn't all that bad.\"))\n",
+       "    [('John', 'NNP'), (\"'s\", 'POS'), ('big', 'JJ'), ('idea', 'NN'), ('is', 'VBZ'),\n",
+       "    (\"n't\", 'RB'), ('all', 'PDT'), ('that', 'DT'), ('bad', 'JJ'), ('.', '.')]\n",
+       "    >>> pos_tag(word_tokenize(\"John's big idea isn't all that bad.\"), tagset='universal')\n",
+       "    [('John', 'NOUN'), (\"'s\", 'PRT'), ('big', 'ADJ'), ('idea', 'NOUN'), ('is', 'VERB'),\n",
+       "    (\"n't\", 'ADV'), ('all', 'DET'), ('that', 'DET'), ('bad', 'ADJ'), ('.', '.')]\n",
+       "\n",
+       "NB. Use `pos_tag_sents()` for efficient tagging of more than one sentence.\n",
+       "\n",
+       ":param tokens: Sequence of tokens to be tagged\n",
+       ":type tokens: list(str)\n",
+       ":param tagset: the tagset to be used, e.g. universal, wsj, brown\n",
+       ":type tagset: str\n",
+       ":param lang: the ISO 639 code of the language, e.g. 'eng' for English, 'rus' for Russian\n",
+       ":type lang: str\n",
+       ":return: The tagged tokens\n",
+       ":rtype: list(tuple(str, str))\n",
+       "\u001b[1;31mFile:\u001b[0m      c:\\anaconda3\\envs\\media\\lib\\site-packages\\nltk\\tag\\__init__.py\n",
+       "\u001b[1;31mType:\u001b[0m      function\n"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "nltk.pos_tag?"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.6.5"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}