diff --git a/Preprocess Stock Data.ipynb b/Preprocess Stock Data.ipynb
new file mode 100644
index 0000000..729226f
--- /dev/null
+++ b/Preprocess Stock Data.ipynb	
@@ -0,0 +1,129 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Import to Postgres\n",
+    "\n",
+    "Load the prices and fundamentals into postgres."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "```shell\n",
+    "DB_NAME=stocks pgfutter csv prices.csv\n",
+    "DB_NAME=stocks pgfutter csv  --fields \"ticker_indicator,date,value\" fundamentals.csv\n",
+    "```\n",
+    "\n",
+    "Setup the prices table.\n",
+    "\n",
+    "```sql\n",
+    "ALTER TABLE import.prices\n",
+    "SET SCHEMA public;\n",
+    "\n",
+    "ALTER TABLE prices ALTER COLUMN \"date\" \n",
+    "SET DATA TYPE date \n",
+    "USING date::date;\n",
+    "\n",
+    "CREATE INDEX idx_prices_date \n",
+    "ON prices(date);\n",
+    "```\n",
+    "\n",
+    "Setup the fundamentals materialized view.\n",
+    "\n",
+    "```sql\n",
+    "ALTER TABLE import.fundamentals ALTER COLUMN \"date\" \n",
+    "SET DATA TYPE date \n",
+    "USING date::date;\n",
+    "\n",
+    "CREATE MATERIALIZED VIEW fundamentals AS\n",
+    "SELECT split_part(ticker_indicator, '_', 1) as ticker, date, value as epsdil \n",
+    "FROM import.fundamentals\n",
+    "WHERE ticker_indicator LIKE '%EPSDIL_MRT';\n",
+    "\n",
+    "CREATE INDEX idx_fundamentals_date \n",
+    "ON fundamentals(date);\n",
+    "\n",
+    "CREATE INDEX idx_fundamentals_ticker \n",
+    "ON fundamentals(ticker);\n",
+    "```"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Query and Export\n",
+    "\n",
+    "Export the prices with fundamentals. This took about 40 minutes."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "```sql\n",
+    "COPY \n",
+    "\n",
+    "(SELECT p.*, \n",
+    "\n",
+    "CASE p.epsdil::numeric\n",
+    "\tWHEN 0 THEN 0\n",
+    "\tELSE p.adj_close::numeric / p.epsdil::numeric\n",
+    "END as pe\n",
+    "\n",
+    "FROM (\n",
+    "\tSELECT prices.adj_close, prices.date, prices.ticker,\n",
+    "\t(\n",
+    "\t\tSELECT fundamentals.epsdil\n",
+    "\t\tFROM fundamentals\n",
+    "\t\tWHERE prices.ticker = fundamentals.ticker\n",
+    "\t\tAND prices.date >= fundamentals.date \n",
+    "\t\tORDER BY fundamentals.date DESC\n",
+    "\t\tLIMIT 1\n",
+    "\t)\n",
+    "\tFROM prices\n",
+    "\tWHERE prices.adj_close IS NOT NULL AND prices.adj_close != ''\n",
+    ") p\n",
+    "WHERE p.epsdil IS NOT NULL AND p.epsdil != ''\n",
+    ")\n",
+    "\n",
+    "TO 'prices.csv' (format CSV);\n",
+    "```"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "collapsed": true
+   },
+   "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": "3.6.0"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}
diff --git a/README.md b/README.md
new file mode 100644
index 0000000..61cac3d
--- /dev/null
+++ b/README.md
@@ -0,0 +1,5 @@
+## Run on floydhub
+
+```shell
+floyd run --gpu --mode jupyter --env tensorflow-1.0
+```
diff --git a/floyd_requirements.txt b/floyd_requirements.txt
new file mode 100644
index 0000000..78620c4
--- /dev/null
+++ b/floyd_requirements.txt
@@ -0,0 +1 @@
+tqdm
diff --git a/stock2vec.ipynb b/stock2vec.ipynb
index 7a9fe66..2889a18 100644
--- a/stock2vec.ipynb
+++ b/stock2vec.ipynb
@@ -18,7 +18,6 @@
    "outputs": [],
    "source": [
     "import csv\n",
-    "# from dateutil.relativedelta import relativedelta\n",
     "import datetime\n",
     "import math\n",
     "import numpy as np\n",
@@ -39,7 +38,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": 56,
    "metadata": {
     "collapsed": false
    },
@@ -48,7 +47,64 @@
      "name": "stderr",
      "output_type": "stream",
      "text": [
-      "rows: 95000000it [02:14, 703865.57it/s]                              \n"
+      "Prices:   3%|▎         | 13.5M/485M [00:01<00:59, 7.95MB/s]  \n"
+     ]
+    },
+    {
+     "ename": "KeyboardInterrupt",
+     "evalue": "",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
+      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
+      "\u001b[0;32m<ipython-input-56-a7b8195dc040>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     19\u001b[0m             \u001b[0;34m'https://s3.amazonaws.com/perl-ml/prices.csv?response-content-disposition=attachment&X-Amz-Security-Token=FQoDYXdzEN3%2F%2F%2F%2F%2F%2F%2F%2F%2F%2FwEaDDGDXIQxfhjhlalnoyKqAiqteedReEObibGFinGZUTbCNLqOsBrBfhb3m%2B9WSc202KdlXdoi8bxYATvctErFAeNF%2FlVgdPlu%2BRy8dLOHw5a%2BvTNM92V8V1XiJnuYgpe69GI914L1xceQGmcJ9qQ1Fg2iSi5cGj2%2FNL26CHIOmdblBGp6VUFUqtu0ZoRb18XXYBlSGQIGk4kxGfwiN5%2BbnQNB%2FInBx0YkDI5XFOIOXa1HzF4anoHgoSSjwdq8FXLQh8LXD5mYvqkTLokIssfZeJrc4TyPy9gZW4hewwbI4NAauQvJfde2Z%2BA%2B5iV4%2B%2B8wFFcDMeM%2Fg%2BYyrTVhaRVZ%2FIU033J6CXshjaL0uHwFleXw%2FHlzMjQst2YZmQu0EqxNCowwwxcugVsKcMaPdMq%2BWJ66qWxN5DZcC3oo%2FJvbxgU%3D&X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Date=20170325T201144Z&X-Amz-SignedHeaders=host&X-Amz-Expires=300&X-Amz-Credential=ASIAIZGCD6XQ355X2AMA%2F20170325%2Fus-east-1%2Fs3%2Faws4_request&X-Amz-Signature=0c2703d3dbef5f58006a3b7e89ff85b2b86e67542b861784fbff2da48434e0df'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     20\u001b[0m             \u001b[0mdataset_filename\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 21\u001b[0;31m             pbar.hook)\n\u001b[0m",
+      "\u001b[0;32m/Users/jperl/anaconda/envs/dlnd/lib/python3.6/urllib/request.py\u001b[0m in \u001b[0;36murlretrieve\u001b[0;34m(url, filename, reporthook, data)\u001b[0m\n\u001b[1;32m    275\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    276\u001b[0m             \u001b[0;32mwhile\u001b[0m \u001b[0;32mTrue\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 277\u001b[0;31m                 \u001b[0mblock\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mread\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    278\u001b[0m                 \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mblock\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    279\u001b[0m                     \u001b[0;32mbreak\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+      "\u001b[0;32m/Users/jperl/anaconda/envs/dlnd/lib/python3.6/http/client.py\u001b[0m in \u001b[0;36mread\u001b[0;34m(self, amt)\u001b[0m\n\u001b[1;32m    447\u001b[0m             \u001b[0;31m# Amount is given, implement using readinto\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    448\u001b[0m             \u001b[0mb\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mbytearray\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mamt\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 449\u001b[0;31m             \u001b[0mn\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mreadinto\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    450\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mmemoryview\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0mn\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtobytes\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    451\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+      "\u001b[0;32m/Users/jperl/anaconda/envs/dlnd/lib/python3.6/http/client.py\u001b[0m in \u001b[0;36mreadinto\u001b[0;34m(self, b)\u001b[0m\n\u001b[1;32m    491\u001b[0m         \u001b[0;31m# connection, and the user is reading more bytes than will be provided\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    492\u001b[0m         \u001b[0;31m# (for example, reading in 1k chunks)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 493\u001b[0;31m         \u001b[0mn\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mreadinto\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    494\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mn\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    495\u001b[0m             \u001b[0;31m# Ideally, we would raise IncompleteRead if the content-length\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+      "\u001b[0;32m/Users/jperl/anaconda/envs/dlnd/lib/python3.6/socket.py\u001b[0m in \u001b[0;36mreadinto\u001b[0;34m(self, b)\u001b[0m\n\u001b[1;32m    584\u001b[0m         \u001b[0;32mwhile\u001b[0m \u001b[0;32mTrue\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    585\u001b[0m             \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 586\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_sock\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrecv_into\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    587\u001b[0m             \u001b[0;32mexcept\u001b[0m \u001b[0mtimeout\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    588\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_timeout_occurred\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mTrue\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+      "\u001b[0;32m/Users/jperl/anaconda/envs/dlnd/lib/python3.6/ssl.py\u001b[0m in \u001b[0;36mrecv_into\u001b[0;34m(self, buffer, nbytes, flags)\u001b[0m\n\u001b[1;32m   1000\u001b[0m                   \u001b[0;34m\"non-zero flags not allowed in calls to recv_into() on %s\"\u001b[0m \u001b[0;34m%\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1001\u001b[0m                   self.__class__)\n\u001b[0;32m-> 1002\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mread\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnbytes\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbuffer\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1003\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1004\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0msocket\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrecv_into\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbuffer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnbytes\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mflags\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+      "\u001b[0;32m/Users/jperl/anaconda/envs/dlnd/lib/python3.6/ssl.py\u001b[0m in \u001b[0;36mread\u001b[0;34m(self, len, buffer)\u001b[0m\n\u001b[1;32m    863\u001b[0m             \u001b[0;32mraise\u001b[0m \u001b[0mValueError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"Read on closed or unwrapped SSL socket.\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    864\u001b[0m         \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 865\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_sslobj\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mread\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbuffer\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    866\u001b[0m         \u001b[0;32mexcept\u001b[0m \u001b[0mSSLError\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    867\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0mSSL_ERROR_EOF\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msuppress_ragged_eofs\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+      "\u001b[0;32m/Users/jperl/anaconda/envs/dlnd/lib/python3.6/ssl.py\u001b[0m in \u001b[0;36mread\u001b[0;34m(self, len, buffer)\u001b[0m\n\u001b[1;32m    623\u001b[0m         \"\"\"\n\u001b[1;32m    624\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mbuffer\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 625\u001b[0;31m             \u001b[0mv\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_sslobj\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mread\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbuffer\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    626\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    627\u001b[0m             \u001b[0mv\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_sslobj\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mread\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
+     ]
+    }
+   ],
+   "source": [
+    "from urllib.request import urlretrieve\n",
+    "from os.path import isfile, isdir\n",
+    "\n",
+    "dataset_folder_path = 'input'\n",
+    "dataset_filename = 'input/prices.csv'\n",
+    "dataset_name = 'Prices'\n",
+    "\n",
+    "class DLProgress(tqdm):\n",
+    "    last_block = 0\n",
+    "\n",
+    "    def hook(self, block_num=1, block_size=1, total_size=None):\n",
+    "        self.total = total_size\n",
+    "        self.update((block_num - self.last_block) * block_size)\n",
+    "        self.last_block = block_num\n",
+    "\n",
+    "if not isfile(dataset_filename):\n",
+    "    with DLProgress(unit='B', unit_scale=True, miniters=1, desc=dataset_name) as pbar:\n",
+    "        urlretrieve(\n",
+    "            'https://s3.amazonaws.com/perl-ml/prices.csv?response-content-disposition=attachment&X-Amz-Security-Token=FQoDYXdzEN3%2F%2F%2F%2F%2F%2F%2F%2F%2F%2FwEaDDGDXIQxfhjhlalnoyKqAiqteedReEObibGFinGZUTbCNLqOsBrBfhb3m%2B9WSc202KdlXdoi8bxYATvctErFAeNF%2FlVgdPlu%2BRy8dLOHw5a%2BvTNM92V8V1XiJnuYgpe69GI914L1xceQGmcJ9qQ1Fg2iSi5cGj2%2FNL26CHIOmdblBGp6VUFUqtu0ZoRb18XXYBlSGQIGk4kxGfwiN5%2BbnQNB%2FInBx0YkDI5XFOIOXa1HzF4anoHgoSSjwdq8FXLQh8LXD5mYvqkTLokIssfZeJrc4TyPy9gZW4hewwbI4NAauQvJfde2Z%2BA%2B5iV4%2B%2B8wFFcDMeM%2Fg%2BYyrTVhaRVZ%2FIU033J6CXshjaL0uHwFleXw%2FHlzMjQst2YZmQu0EqxNCowwwxcugVsKcMaPdMq%2BWJ66qWxN5DZcC3oo%2FJvbxgU%3D&X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Date=20170325T201144Z&X-Amz-SignedHeaders=host&X-Amz-Expires=300&X-Amz-Credential=ASIAIZGCD6XQ355X2AMA%2F20170325%2Fus-east-1%2Fs3%2Faws4_request&X-Amz-Signature=0c2703d3dbef5f58006a3b7e89ff85b2b86e67542b861784fbff2da48434e0df',\n",
+    "            dataset_filename,\n",
+    "            pbar.hook)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "metadata": {
+    "collapsed": false
+   },
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "rows: 10000000it [00:15, 607491.88it/s]                            \n"
      ]
     },
     {
@@ -59,277 +115,128 @@
        "  <thead>\n",
        "    <tr style=\"text-align: right;\">\n",
        "      <th></th>\n",
-       "      <th>0</th>\n",
-       "      <th>1</th>\n",
-       "      <th>2</th>\n",
+       "      <th>adj_close</th>\n",
+       "      <th>date</th>\n",
+       "      <th>ticker</th>\n",
+       "      <th>epsdil</th>\n",
+       "      <th>pe</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
-       "      <th>65171973</th>\n",
-       "      <td>PEP</td>\n",
-       "      <td>2001-06-16</td>\n",
-       "      <td>1.50</td>\n",
+       "      <th>45526</th>\n",
+       "      <td>13.745073</td>\n",
+       "      <td>2001-06-18</td>\n",
+       "      <td>OLED</td>\n",
+       "      <td>-0.87</td>\n",
+       "      <td>-15.798935</td>\n",
        "    </tr>\n",
        "    <tr>\n",
-       "      <th>80257284</th>\n",
-       "      <td>SVU</td>\n",
-       "      <td>2001-06-16</td>\n",
-       "      <td>0.47</td>\n",
+       "      <th>3046709</th>\n",
+       "      <td>6.147269</td>\n",
+       "      <td>2001-06-18</td>\n",
+       "      <td>YUM</td>\n",
+       "      <td>0.69</td>\n",
+       "      <td>8.909085</td>\n",
        "    </tr>\n",
        "    <tr>\n",
-       "      <th>80572836</th>\n",
+       "      <th>1778402</th>\n",
+       "      <td>43.419875</td>\n",
+       "      <td>2001-06-18</td>\n",
        "      <td>SWY</td>\n",
-       "      <td>2001-06-16</td>\n",
        "      <td>2.31</td>\n",
+       "      <td>18.796483</td>\n",
        "    </tr>\n",
        "    <tr>\n",
-       "      <th>93428968</th>\n",
-       "      <td>YUM</td>\n",
-       "      <td>2001-06-16</td>\n",
-       "      <td>0.69</td>\n",
+       "      <th>315864</th>\n",
+       "      <td>29.645033</td>\n",
+       "      <td>2001-06-18</td>\n",
+       "      <td>PEP</td>\n",
+       "      <td>1.50</td>\n",
+       "      <td>19.763356</td>\n",
        "    </tr>\n",
        "    <tr>\n",
-       "      <th>62307830</th>\n",
-       "      <td>OLED</td>\n",
-       "      <td>2001-06-17</td>\n",
-       "      <td>-0.87</td>\n",
+       "      <th>1743245</th>\n",
+       "      <td>11.213455</td>\n",
+       "      <td>2001-06-18</td>\n",
+       "      <td>SVU</td>\n",
+       "      <td>0.47</td>\n",
+       "      <td>23.858415</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
        "</div>"
       ],
       "text/plain": [
-       "             0          1     2\n",
-       "65171973   PEP 2001-06-16  1.50\n",
-       "80257284   SVU 2001-06-16  0.47\n",
-       "80572836   SWY 2001-06-16  2.31\n",
-       "93428968   YUM 2001-06-16  0.69\n",
-       "62307830  OLED 2001-06-17 -0.87"
+       "         adj_close       date ticker  epsdil         pe\n",
+       "45526    13.745073 2001-06-18   OLED   -0.87 -15.798935\n",
+       "3046709   6.147269 2001-06-18    YUM    0.69   8.909085\n",
+       "1778402  43.419875 2001-06-18    SWY    2.31  18.796483\n",
+       "315864   29.645033 2001-06-18    PEP    1.50  19.763356\n",
+       "1743245  11.213455 2001-06-18    SVU    0.47  23.858415"
       ]
      },
-     "execution_count": 2,
+     "execution_count": 11,
      "metadata": {},
      "output_type": "execute_result"
     }
    ],
    "source": [
     "chunksize = 1000000\n",
-    "fund_rows = 94011143\n",
+    "price_rows = 9191528\n",
     "\n",
-    "fund_reader = pd.read_csv('input/sharadar_fundamentals.csv', \n",
+    "price_reader = pd.read_csv('input/prices.csv', \n",
     "                      header=None,\n",
     "                      parse_dates=[1],\n",
     "                      chunksize=chunksize, \n",
     "                      iterator=True)\n",
     "\n",
-    "df_fund = pd.DataFrame()\n",
-    "\n",
-    "with tqdm(total=fund_rows, desc='rows') as pbar:\n",
-    "    for chunk in fund_reader:\n",
-    "        # Select diluted earnings per share data\n",
-    "        chunk = chunk[chunk[0].str.contains('EPSDIL_MRT')]\n",
-    "        # Extract the ticker\n",
-    "        chunk[0] = chunk[0].str.extract('(.+?(?=_))', expand=False)\n",
-    "        df_fund = df_fund.append(chunk)\n",
-    "        pbar.update(chunksize);\n",
-    "\n",
-    "# Sort by date, then ticker\n",
-    "df_fund.sort_values([1, 0], inplace=True)\n",
-    "\n",
-    "earliest_date = df_fund.values[0][1]\n",
-    "\n",
-    "df_fund.head()"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "Load the ticker prices."
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 3,
-   "metadata": {
-    "collapsed": false
-   },
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "rows: 15000000it [00:21, 693364.20it/s]                              \n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "      ticker       date  adj_close\n",
-      "397        A 2001-06-18  28.510446\n",
-      "12636    AAN 2001-06-18   4.669972\n",
-      "19603   AAON 2001-06-18   1.883050\n",
-      "32593   AAPL 2001-06-18   1.323831\n",
-      "43324   ABAX 2001-06-18   5.458700\n"
-     ]
-    }
-   ],
-   "source": [
-    "price_rows = 14684263\n",
-    "\n",
-    "# Import the prices\n",
-    "\n",
-    "prices_reader = pd.read_csv('input/wiki_prices.csv', \n",
-    "                        chunksize=chunksize,\n",
-    "                        parse_dates=['date'],\n",
-    "                        usecols=['adj_close', 'date', 'ticker'])\n",
-    "\n",
     "df_prices = pd.DataFrame()\n",
     "\n",
     "with tqdm(total=price_rows, desc='rows') as pbar:\n",
-    "    for chunk in prices_reader:\n",
-    "        chunk = chunk[pd.notnull(chunk['adj_close'])]  \n",
-    "        chunk = chunk[chunk['date'] >= earliest_date]\n",
+    "    for chunk in price_reader:\n",
     "        df_prices = df_prices.append(chunk)\n",
     "        pbar.update(chunksize);\n",
     "\n",
-    "df_prices.sort_values(['date', 'ticker'], inplace=True)\n",
+    "df_prices.columns = ['adj_close', 'date', 'ticker', 'epsdil', 'pe']\n",
     "\n",
-    "print(df_prices.head())"
+    "# Sort by date, then ticker\n",
+    "df_prices.sort_values(['date', 'pe'], inplace=True)\n",
+    "\n",
+    "df_prices.head()"
    ]
   },
   {
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Append diluted earnings per share."
+    "## Build context\n",
+    "\n",
+    "For each stock, find C stocks that have the closest volatility to that ticker for that day."
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 46,
    "metadata": {
     "collapsed": false
    },
    "outputs": [
     {
-     "name": "stderr",
+     "name": "stdout",
      "output_type": "stream",
      "text": [
-      "rows:   0%|          | 1631/9969557 [01:13<106:16:34, 26.05it/s]\n"
-     ]
-    },
-    {
-     "ename": "KeyboardInterrupt",
-     "evalue": "",
-     "output_type": "error",
-     "traceback": [
-      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
-      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
-      "\u001b[0;32m<ipython-input-4-d80d4dc7c05d>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     28\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     29\u001b[0m \u001b[0;32mwith\u001b[0m \u001b[0mtqdm\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtotal\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdf_prices\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdesc\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'rows'\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mpbar\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 30\u001b[0;31m     \u001b[0mfund_cols\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdf_prices\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mapply\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;32mlambda\u001b[0m \u001b[0mr\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mget_fund_cols\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mr\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpbar\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     31\u001b[0m     \u001b[0mfund_cols\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcolumns\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m'fund_date'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'epsdil'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'pe'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     32\u001b[0m     \u001b[0mdf_prices\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdf_prices\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mjoin\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfund_cols\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
-      "\u001b[0;32m/Users/jperl/anaconda/envs/dlnd/lib/python3.6/site-packages/pandas/core/frame.py\u001b[0m in \u001b[0;36mapply\u001b[0;34m(self, func, axis, broadcast, raw, reduce, args, **kwds)\u001b[0m\n\u001b[1;32m   4150\u001b[0m                     \u001b[0;32mif\u001b[0m \u001b[0mreduce\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   4151\u001b[0m                         \u001b[0mreduce\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mTrue\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 4152\u001b[0;31m                     \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_apply_standard\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mf\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mreduce\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mreduce\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   4153\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   4154\u001b[0m                 \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_apply_broadcast\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mf\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
-      "\u001b[0;32m/Users/jperl/anaconda/envs/dlnd/lib/python3.6/site-packages/pandas/core/frame.py\u001b[0m in \u001b[0;36m_apply_standard\u001b[0;34m(self, func, axis, ignore_failures, reduce)\u001b[0m\n\u001b[1;32m   4246\u001b[0m             \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   4247\u001b[0m                 \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mv\u001b[0m \u001b[0;32min\u001b[0m \u001b[0menumerate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mseries_gen\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 4248\u001b[0;31m                     \u001b[0mresults\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mv\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   4249\u001b[0m                     \u001b[0mkeys\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mv\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mname\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   4250\u001b[0m             \u001b[0;32mexcept\u001b[0m \u001b[0mException\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
-      "\u001b[0;32m<ipython-input-4-d80d4dc7c05d>\u001b[0m in \u001b[0;36m<lambda>\u001b[0;34m(r)\u001b[0m\n\u001b[1;32m     28\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     29\u001b[0m \u001b[0;32mwith\u001b[0m \u001b[0mtqdm\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtotal\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdf_prices\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdesc\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'rows'\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mpbar\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 30\u001b[0;31m     \u001b[0mfund_cols\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdf_prices\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mapply\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;32mlambda\u001b[0m \u001b[0mr\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mget_fund_cols\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mr\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpbar\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     31\u001b[0m     \u001b[0mfund_cols\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcolumns\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m'fund_date'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'epsdil'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'pe'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     32\u001b[0m     \u001b[0mdf_prices\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdf_prices\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mjoin\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfund_cols\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
-      "\u001b[0;32m<ipython-input-4-d80d4dc7c05d>\u001b[0m in \u001b[0;36mget_fund_cols\u001b[0;34m(row, pbar)\u001b[0m\n\u001b[1;32m     16\u001b[0m     \u001b[0mpbar\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mupdate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     17\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 18\u001b[0;31m     \u001b[0mfund\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mget_fund\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrow\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdate\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrow\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mticker\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     19\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     20\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mfund\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
-      "\u001b[0;32m<ipython-input-4-d80d4dc7c05d>\u001b[0m in \u001b[0;36mget_fund\u001b[0;34m(date, ticker)\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mget_fund\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdate\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mticker\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      2\u001b[0m     \u001b[0;31m# Find the most recent fundamentals before the date\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 3\u001b[0;31m     \u001b[0mfund\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdf_fund\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdf_fund\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0mticker\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m&\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mdf_fund\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m<=\u001b[0m \u001b[0mdate\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtail\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      4\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mfund\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m<\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
-      "\u001b[0;32m/Users/jperl/anaconda/envs/dlnd/lib/python3.6/site-packages/pandas/core/ops.py\u001b[0m in \u001b[0;36mwrapper\u001b[0;34m(self, other, axis)\u001b[0m\n\u001b[1;32m    853\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    854\u001b[0m             \u001b[0;32mwith\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0merrstate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mall\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'ignore'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 855\u001b[0;31m                 \u001b[0mres\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mna_op\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mother\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    856\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0misscalar\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mres\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    857\u001b[0m                 raise TypeError('Could not compare %s type with Series' %\n",
-      "\u001b[0;32m/Users/jperl/anaconda/envs/dlnd/lib/python3.6/site-packages/pandas/core/ops.py\u001b[0m in \u001b[0;36mna_op\u001b[0;34m(x, y)\u001b[0m\n\u001b[1;32m    757\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    758\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mis_object_dtype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdtype\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 759\u001b[0;31m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_comp_method_OBJECT_ARRAY\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mop\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    760\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    761\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
-      "\u001b[0;32m/Users/jperl/anaconda/envs/dlnd/lib/python3.6/site-packages/pandas/core/ops.py\u001b[0m in \u001b[0;36m_comp_method_OBJECT_ARRAY\u001b[0;34m(op, x, y)\u001b[0m\n\u001b[1;32m    737\u001b[0m         \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlib\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvec_compare\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mop\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    738\u001b[0m     \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 739\u001b[0;31m         \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlib\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mscalar_compare\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mop\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    740\u001b[0m     \u001b[0;32mreturn\u001b[0m \u001b[0mresult\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    741\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
-      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
+      "window for 0 OLED 0\n",
+      "1 YUM\n",
+      "window for 5 OLED 0\n",
+      "1 YUM\n",
+      "2 SWY\n",
+      "3 PEP\n",
+      "4 SVU\n",
+      "3019\n"
      ]
     }
    ],
-   "source": [
-    "def get_fund(date, ticker):\n",
-    "    # Find the most recent fundamentals before the date\n",
-    "    fund = df_fund[(df_fund[0] == ticker) & (df_fund[1] <= date)].tail(1)    \n",
-    "    if fund.shape[0] < 1:\n",
-    "        return None\n",
-    "    return fund.values[0]\n",
-    "\n",
-    "def price_to_earnings(price, epsdil):\n",
-    "    if epsdil is None:\n",
-    "        return None\n",
-    "    if epsdil == 0:\n",
-    "        return 0\n",
-    "    return price / epsdil\n",
-    "\n",
-    "def get_fund_cols(row, pbar):\n",
-    "    pbar.update()\n",
-    "    \n",
-    "    fund = get_fund(row.date, row.ticker)\n",
-    "    \n",
-    "    if fund is None:\n",
-    "        return pd.Series([None, None, None])\n",
-    "    \n",
-    "    date = fund[1]\n",
-    "    epsdil = fund[2]\n",
-    "    pe = price_to_earnings(row.adj_close, epsdil)\n",
-    "    \n",
-    "    return pd.Series([date, epsdil, pe])\n",
-    "\n",
-    "with tqdm(total=df_prices.shape[0], desc='rows') as pbar:\n",
-    "    fund_cols = df_prices.apply(lambda r: get_fund_cols(r, pbar), axis=1)\n",
-    "    fund_cols.columns = ['fund_date', 'epsdil', 'pe']\n",
-    "    df_prices = df_prices.join(fund_cols)\n",
-    "\n",
-    "print(df_prices.head())"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "## Preprocessing\n",
-    "\n",
-    "Filter ticks for the past ~15 years of stocks with volume > 10000 and volatility > 0."
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {
-    "collapsed": false
-   },
-   "outputs": [],
-   "source": [
-    "df = df[df['date'] >= datetime.date(2016,1,1)]\n",
-    "df = df[df['adj_volume'] > 10000]\n",
-    "del df['adj_volume']\n",
-    "\n",
-    "df['volt'] = (df['adj_high'] - df['adj_low']) / df['adj_close']\n",
-    "del df['adj_high']\n",
-    "del df['adj_low']\n",
-    "del df['adj_close']\n",
-    "\n",
-    "df = df[df['volt'] > 0]\n",
-    "\n",
-    "df.sort_values(['date', 'volt'], inplace=True)\n",
-    "\n",
-    "n_stocks = len(df['ticker'].unique())\n",
-    "print('n_stocks', n_stocks)\n",
-    "\n",
-    "print(df.head())"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "**Get Nearby Stocks**\n",
-    "\n",
-    "For each stock, find C stocks that have the closest volatility to that ticker for that day."
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {
-    "collapsed": false
-   },
-   "outputs": [],
    "source": [
     "ticker_to_int = {}\n",
     "int_to_ticker = {}\n",
@@ -342,10 +249,10 @@
     "    return key\n",
     "\n",
     "def get_stock_date(stocks, idx):\n",
-    "    return stocks.iloc[[idx], 1].values[0]\n",
+    "    return stocks.iloc[idx][1]\n",
     "\n",
     "def get_stock_ticker(stocks, idx):\n",
-    "    return stocks.iloc[[idx], 0].values[0]\n",
+    "    return stocks.iloc[idx][2]\n",
     "\n",
     "def get_stock_int(stocks, idx):\n",
     "    return get_ticker_int(get_stock_ticker(stocks, idx))\n",
@@ -368,15 +275,15 @@
     "    \n",
     "    return window\n",
     "\n",
-    "for idx in range(0, 10, 5):\n",
-    "    print('window for', get_stock_ticker(df, idx), get_stock_int(df, idx))\n",
-    "    for nearby_int in get_window(df, idx, 5):\n",
+    "for idx in range(0, 20, 10):\n",
+    "    print('window for', idx, get_stock_ticker(df_prices, idx), get_stock_int(df_prices, idx))\n",
+    "    for nearby_int in get_window(df_prices, idx, 5):\n",
     "        print(nearby_int, int_to_ticker[nearby_int])"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 44,
    "metadata": {
     "collapsed": false
    },
@@ -405,12 +312,14 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 47,
    "metadata": {
     "collapsed": false
    },
    "outputs": [],
    "source": [
+    "n_stocks = len(df_prices['ticker'].unique())\n",
+    "\n",
     "train_graph = tf.Graph()\n",
     "with train_graph.as_default():\n",
     "    inputs = tf.placeholder(tf.int32, [None], name='inputs')\n",
@@ -432,7 +341,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 48,
    "metadata": {
     "collapsed": false
    },
@@ -456,7 +365,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 49,
    "metadata": {
     "collapsed": false
    },
@@ -489,15 +398,500 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 52,
    "metadata": {
     "collapsed": false
    },
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 1/10 Iteration: 10 Avg. Training loss: 5.2061 1.8950 sec/batch\n",
+      "Epoch 1/10 Iteration: 20 Avg. Training loss: 5.0781 1.9572 sec/batch\n",
+      "Epoch 1/10 Iteration: 30 Avg. Training loss: 4.9133 2.0237 sec/batch\n",
+      "Epoch 1/10 Iteration: 40 Avg. Training loss: 4.7580 2.0461 sec/batch\n",
+      "Epoch 1/10 Iteration: 50 Avg. Training loss: 4.7297 1.9665 sec/batch\n",
+      "Epoch 1/10 Iteration: 60 Avg. Training loss: 4.5833 2.0258 sec/batch\n",
+      "Epoch 1/10 Iteration: 70 Avg. Training loss: 4.6288 2.0546 sec/batch\n",
+      "Epoch 1/10 Iteration: 80 Avg. Training loss: 4.6524 2.1314 sec/batch\n",
+      "Epoch 1/10 Iteration: 90 Avg. Training loss: 4.7750 1.9557 sec/batch\n",
+      "Epoch 1/10 Iteration: 100 Avg. Training loss: 4.6779 1.9280 sec/batch\n",
+      "nearest[k] 2444\n",
+      "nearest[k] 2899\n",
+      "nearest[k] 1547\n",
+      "Nearest to ACAS: CATO, FAST, VLGEA, LM, MYE,\n",
+      "nearest[k] 1840\n",
+      "nearest[k] 2568\n",
+      "nearest[k] 2191\n",
+      "Nearest to RTN: GLF, ADS, HOLX, DDR, CAC,\n",
+      "nearest[k] 1922\n",
+      "nearest[k] 1977\n",
+      "nearest[k] 2228\n",
+      "nearest[k] 2757\n",
+      "nearest[k] 2825\n",
+      "Nearest to SYKE: GNCMA, PPL, BCOR,\n",
+      "nearest[k] 2439\n",
+      "nearest[k] 2354\n",
+      "nearest[k] 1745\n",
+      "Nearest to USAK: SENEA, CBB, MSFT, MSI, SPNC,\n",
+      "nearest[k] 2088\n",
+      "nearest[k] 2785\n",
+      "nearest[k] 2370\n",
+      "nearest[k] 1793\n",
+      "nearest[k] 1818\n",
+      "nearest[k] 2747\n",
+      "nearest[k] 2705\n",
+      "Nearest to IMMU: STT,\n",
+      "nearest[k] 1603\n",
+      "nearest[k] 1657\n",
+      "nearest[k] 2324\n",
+      "nearest[k] 2458\n",
+      "Nearest to UFPT: PHX, CINF, CHS, RMTI,\n",
+      "nearest[k] 2562\n",
+      "nearest[k] 2991\n",
+      "nearest[k] 1769\n",
+      "nearest[k] 1900\n",
+      "Nearest to SENEA: USAK, RE, COLB, ALR,\n",
+      "nearest[k] 1679\n",
+      "nearest[k] 2406\n",
+      "Nearest to MOD: PHM, ENZN, CTG, DST, FBC, VICR,\n",
+      "nearest[k] 2220\n",
+      "nearest[k] 2278\n",
+      "nearest[k] 1632\n",
+      "nearest[k] 2772\n",
+      "nearest[k] 2374\n",
+      "nearest[k] 1547\n",
+      "Nearest to AMG: LAMR, AAN,\n",
+      "nearest[k] 2098\n",
+      "nearest[k] 2526\n",
+      "Nearest to ORLY: MIND, GAIA, STAR, XEL, RGR, LCI,\n",
+      "nearest[k] 2316\n",
+      "nearest[k] 2583\n",
+      "nearest[k] 2135\n",
+      "Nearest to MTD: POOL, CNMD, RGEN, WINA, AIRM,\n",
+      "nearest[k] 2202\n",
+      "nearest[k] 2396\n",
+      "nearest[k] 2634\n",
+      "nearest[k] 1902\n",
+      "Nearest to GE: UFI, RE, MCS, TTWO,\n",
+      "nearest[k] 2874\n",
+      "nearest[k] 1797\n",
+      "nearest[k] 2969\n",
+      "nearest[k] 2726\n",
+      "nearest[k] 1990\n",
+      "nearest[k] 2155\n",
+      "Nearest to ICUI: IOSP, NFX,\n",
+      "nearest[k] 2448\n",
+      "nearest[k] 1594\n",
+      "nearest[k] 2647\n",
+      "Nearest to ECL: NDSN, VSAT, UTX, TILE, INT,\n",
+      "nearest[k] 3018\n",
+      "nearest[k] 2708\n",
+      "nearest[k] 2041\n",
+      "nearest[k] 1619\n",
+      "nearest[k] 2158\n",
+      "nearest[k] 2146\n",
+      "nearest[k] 1747\n",
+      "Nearest to INTC: C,\n",
+      "nearest[k] 1966\n",
+      "nearest[k] 2427\n",
+      "nearest[k] 1967\n",
+      "nearest[k] 1760\n",
+      "nearest[k] 2207\n",
+      "Nearest to FAST: CATY, ACAS, DIN,\n",
+      "Epoch 1/10 Iteration: 110 Avg. Training loss: 4.6291 2.0415 sec/batch\n",
+      "Epoch 1/10 Iteration: 120 Avg. Training loss: 4.5290 2.1442 sec/batch\n",
+      "Epoch 1/10 Iteration: 130 Avg. Training loss: 4.5225 2.1808 sec/batch\n",
+      "Epoch 1/10 Iteration: 140 Avg. Training loss: 4.4390 2.1799 sec/batch\n",
+      "Epoch 1/10 Iteration: 150 Avg. Training loss: 4.3236 2.2162 sec/batch\n",
+      "Epoch 1/10 Iteration: 160 Avg. Training loss: 4.3258 1.9534 sec/batch\n",
+      "Epoch 1/10 Iteration: 170 Avg. Training loss: 4.2205 1.9416 sec/batch\n",
+      "Epoch 1/10 Iteration: 180 Avg. Training loss: 4.3843 2.0543 sec/batch\n",
+      "Epoch 1/10 Iteration: 190 Avg. Training loss: 4.5850 2.2644 sec/batch\n",
+      "Epoch 1/10 Iteration: 200 Avg. Training loss: 4.4068 2.0691 sec/batch\n",
+      "nearest[k] 2444\n",
+      "nearest[k] 2899\n",
+      "nearest[k] 2569\n",
+      "nearest[k] 2975\n",
+      "Nearest to ACAS: CATO, FAST, VLGEA, MYE,\n",
+      "nearest[k] 1840\n",
+      "nearest[k] 2568\n",
+      "nearest[k] 2191\n",
+      "Nearest to RTN: ADS, GLF, CAC, HOLX, DDR,\n",
+      "nearest[k] 1922\n",
+      "nearest[k] 1977\n",
+      "nearest[k] 2757\n",
+      "nearest[k] 2228\n",
+      "nearest[k] 2825\n",
+      "Nearest to SYKE: GNCMA, PPL, BCOR,\n",
+      "nearest[k] 2439\n",
+      "nearest[k] 2354\n",
+      "nearest[k] 1745\n",
+      "Nearest to USAK: SENEA, MSFT, CBB, MSI, WWD,\n",
+      "nearest[k] 2088\n",
+      "nearest[k] 2785\n",
+      "nearest[k] 2370\n",
+      "nearest[k] 1818\n",
+      "nearest[k] 2747\n",
+      "nearest[k] 1793\n",
+      "nearest[k] 2705\n",
+      "Nearest to IMMU: STT,\n",
+      "nearest[k] 1603\n",
+      "nearest[k] 1657\n",
+      "nearest[k] 2324\n",
+      "nearest[k] 2458\n",
+      "Nearest to UFPT: PHX, CINF, RMTI, CHS,\n",
+      "nearest[k] 2562\n",
+      "nearest[k] 1769\n",
+      "nearest[k] 2991\n",
+      "nearest[k] 1900\n",
+      "Nearest to SENEA: USAK, RE, COLB, CPHD,\n",
+      "nearest[k] 1679\n",
+      "nearest[k] 2406\n",
+      "Nearest to MOD: PHM, ENZN, CTG, DST, INGR, GHC,\n",
+      "nearest[k] 2220\n",
+      "nearest[k] 2278\n",
+      "nearest[k] 1632\n",
+      "nearest[k] 2772\n",
+      "nearest[k] 2719\n",
+      "nearest[k] 2374\n",
+      "Nearest to AMG: AAN, BHB,\n",
+      "nearest[k] 2098\n",
+      "nearest[k] 2526\n",
+      "Nearest to ORLY: MIND, GAIA, STAR, XEL, RGR, LCI,\n",
+      "nearest[k] 2316\n",
+      "nearest[k] 2583\n",
+      "nearest[k] 2135\n",
+      "nearest[k] 2462\n",
+      "Nearest to MTD: POOL, CNMD, AIRM, AVA,\n",
+      "nearest[k] 2202\n",
+      "nearest[k] 2396\n",
+      "nearest[k] 2634\n",
+      "nearest[k] 1902\n",
+      "nearest[k] 2728\n",
+      "Nearest to GE: UFI, RE, ARCW,\n",
+      "nearest[k] 2874\n",
+      "nearest[k] 2969\n",
+      "nearest[k] 1797\n",
+      "nearest[k] 2726\n",
+      "nearest[k] 1990\n",
+      "nearest[k] 2155\n",
+      "Nearest to ICUI: IOSP, UBA,\n",
+      "nearest[k] 2448\n",
+      "nearest[k] 1594\n",
+      "nearest[k] 2647\n",
+      "nearest[k] 2177\n",
+      "Nearest to ECL: NDSN, VSAT, TILE, MXWL,\n",
+      "nearest[k] 3018\n",
+      "nearest[k] 2708\n",
+      "nearest[k] 2041\n",
+      "nearest[k] 2158\n",
+      "nearest[k] 2146\n",
+      "nearest[k] 1619\n",
+      "Nearest to INTC: C, BBOX,\n",
+      "nearest[k] 1966\n",
+      "nearest[k] 2427\n",
+      "nearest[k] 1967\n",
+      "nearest[k] 1760\n",
+      "nearest[k] 2207\n",
+      "Nearest to FAST: CATY, ACAS, DIN,\n",
+      "Epoch 1/10 Iteration: 210 Avg. Training loss: 4.3871 2.0840 sec/batch\n",
+      "Epoch 1/10 Iteration: 220 Avg. Training loss: 4.3398 2.1501 sec/batch\n",
+      "Epoch 1/10 Iteration: 230 Avg. Training loss: 4.2177 2.3706 sec/batch\n",
+      "Epoch 1/10 Iteration: 240 Avg. Training loss: 4.1333 2.0591 sec/batch\n",
+      "Epoch 1/10 Iteration: 250 Avg. Training loss: 4.0676 1.9890 sec/batch\n",
+      "Epoch 1/10 Iteration: 260 Avg. Training loss: 3.9750 2.0815 sec/batch\n",
+      "Epoch 1/10 Iteration: 270 Avg. Training loss: 3.9287 1.9713 sec/batch\n",
+      "Epoch 1/10 Iteration: 280 Avg. Training loss: 4.3463 2.8065 sec/batch\n",
+      "Epoch 1/10 Iteration: 290 Avg. Training loss: 4.2628 1.9577 sec/batch\n",
+      "Epoch 1/10 Iteration: 300 Avg. Training loss: 4.1398 2.0271 sec/batch\n",
+      "nearest[k] 2444\n",
+      "nearest[k] 2569\n",
+      "nearest[k] 2975\n",
+      "nearest[k] 2899\n",
+      "Nearest to ACAS: CATO, FAST, VLGEA, MYE,\n",
+      "nearest[k] 1840\n",
+      "nearest[k] 2568\n",
+      "nearest[k] 2191\n",
+      "Nearest to RTN: HOLX, ADS, GLF, CAC, PRKR,\n",
+      "nearest[k] 1922\n",
+      "nearest[k] 1977\n",
+      "nearest[k] 2757\n",
+      "nearest[k] 2228\n",
+      "Nearest to SYKE: GNCMA, MANH, PPL, QCOM,\n",
+      "nearest[k] 2439\n",
+      "nearest[k] 2354\n",
+      "nearest[k] 1745\n",
+      "nearest[k] 2025\n",
+      "Nearest to USAK: SENEA, MSFT, CBB, SPNC,\n",
+      "nearest[k] 2088\n",
+      "nearest[k] 2785\n",
+      "nearest[k] 2370\n",
+      "nearest[k] 1818\n",
+      "nearest[k] 2747\n",
+      "nearest[k] 2323\n",
+      "Nearest to IMMU: STT, HUBG,\n",
+      "nearest[k] 1603\n",
+      "nearest[k] 1657\n",
+      "nearest[k] 2324\n",
+      "nearest[k] 2638\n",
+      "nearest[k] 2458\n",
+      "Nearest to UFPT: PHX, RMTI, CINF,\n",
+      "nearest[k] 2562\n",
+      "nearest[k] 1769\n",
+      "nearest[k] 1635\n",
+      "nearest[k] 1900\n",
+      "nearest[k] 2991\n",
+      "Nearest to SENEA: USAK, RE, NNBR,\n",
+      "nearest[k] 2406\n",
+      "nearest[k] 1679\n",
+      "Nearest to MOD: PHM, ENZN, DST, CTG, GHC, INGR,\n",
+      "nearest[k] 2220\n",
+      "nearest[k] 1632\n",
+      "nearest[k] 2278\n",
+      "nearest[k] 2772\n",
+      "nearest[k] 2719\n",
+      "Nearest to AMG: AAN, BHB, LAMR,\n",
+      "nearest[k] 2098\n",
+      "Nearest to ORLY: MIND, GAIA, STAR, RGR, VAL, XEL, LCI,\n",
+      "nearest[k] 2316\n",
+      "nearest[k] 2583\n",
+      "nearest[k] 2135\n",
+      "nearest[k] 2462\n",
+      "Nearest to MTD: POOL, AIRM, CNMD, WINA,\n",
+      "nearest[k] 2202\n",
+      "nearest[k] 2396\n",
+      "nearest[k] 1902\n",
+      "nearest[k] 2634\n",
+      "nearest[k] 2728\n",
+      "Nearest to GE: UFI, RE, ARCW,\n",
+      "nearest[k] 2874\n",
+      "nearest[k] 2969\n",
+      "nearest[k] 2726\n",
+      "nearest[k] 1990\n",
+      "nearest[k] 1797\n",
+      "Nearest to ICUI: IOSP, UBA, NFX,\n",
+      "nearest[k] 1594\n",
+      "nearest[k] 2448\n",
+      "nearest[k] 2647\n",
+      "nearest[k] 2177\n",
+      "Nearest to ECL: NDSN, MXWL, VSAT, TILE,\n",
+      "nearest[k] 3018\n",
+      "nearest[k] 2708\n",
+      "nearest[k] 2041\n",
+      "nearest[k] 2146\n",
+      "nearest[k] 2158\n",
+      "nearest[k] 1619\n",
+      "Nearest to INTC: DIS, BBOX,\n",
+      "nearest[k] 2427\n",
+      "nearest[k] 1966\n",
+      "nearest[k] 2207\n",
+      "nearest[k] 1967\n",
+      "nearest[k] 1760\n",
+      "Nearest to FAST: ACAS, CATY, DIN,\n",
+      "Epoch 1/10 Iteration: 310 Avg. Training loss: 4.1017 2.0536 sec/batch\n",
+      "Epoch 1/10 Iteration: 320 Avg. Training loss: 4.0373 1.9826 sec/batch\n",
+      "Epoch 1/10 Iteration: 330 Avg. Training loss: 3.9007 2.1121 sec/batch\n",
+      "Epoch 1/10 Iteration: 340 Avg. Training loss: 3.7818 2.1092 sec/batch\n",
+      "Epoch 1/10 Iteration: 350 Avg. Training loss: 3.8070 2.0725 sec/batch\n",
+      "Epoch 1/10 Iteration: 360 Avg. Training loss: 3.7260 2.0439 sec/batch\n",
+      "Epoch 1/10 Iteration: 370 Avg. Training loss: 3.6276 1.9681 sec/batch\n",
+      "Epoch 1/10 Iteration: 380 Avg. Training loss: 4.0391 1.9432 sec/batch\n",
+      "Epoch 1/10 Iteration: 390 Avg. Training loss: 4.0613 2.0991 sec/batch\n",
+      "Epoch 1/10 Iteration: 400 Avg. Training loss: 3.8643 1.9802 sec/batch\n",
+      "nearest[k] 2444\n",
+      "nearest[k] 2569\n",
+      "nearest[k] 2975\n",
+      "nearest[k] 2899\n",
+      "Nearest to ACAS: CATO, FAST, MYE, LM,\n",
+      "nearest[k] 1840\n",
+      "nearest[k] 2191\n",
+      "nearest[k] 2568\n",
+      "Nearest to RTN: HOLX, OLED, CLFD, PRKR, ADS,\n",
+      "nearest[k] 1922\n",
+      "nearest[k] 1977\n",
+      "nearest[k] 2228\n",
+      "Nearest to SYKE: GNCMA, QCOM, MANH, PPL, BCOR,\n",
+      "nearest[k] 2439\n",
+      "nearest[k] 2354\n",
+      "nearest[k] 1745\n",
+      "nearest[k] 1713\n",
+      "Nearest to USAK: SENEA, SPNC, MSFT, JBL,\n",
+      "nearest[k] 2088\n",
+      "nearest[k] 2785\n",
+      "nearest[k] 2370\n",
+      "nearest[k] 1818\n",
+      "nearest[k] 2747\n",
+      "nearest[k] 2323\n",
+      "Nearest to IMMU: STT, HUBG,\n",
+      "nearest[k] 1603\n",
+      "nearest[k] 2324\n",
+      "nearest[k] 1657\n",
+      "nearest[k] 2638\n",
+      "Nearest to UFPT: PHX, RMTI, IDRA, CWST,\n",
+      "nearest[k] 1635\n",
+      "nearest[k] 2562\n",
+      "nearest[k] 1769\n",
+      "Nearest to SENEA: USAK, NBL, COLB, RE, NNBR,\n",
+      "nearest[k] 2406\n",
+      "nearest[k] 1679\n",
+      "nearest[k] 2185\n",
+      "Nearest to MOD: ENZN, PHM, DST, CTG, INGR,\n",
+      "nearest[k] 2220\n",
+      "nearest[k] 2772\n",
+      "nearest[k] 1632\n",
+      "nearest[k] 2719\n",
+      "nearest[k] 2278\n",
+      "Nearest to AMG: AAN, MATW, BHB,\n",
+      "nearest[k] 2098\n",
+      "Nearest to ORLY: MIND, GAIA, LCI, VAL, RGR, GSOL, MIDD,\n",
+      "nearest[k] 2316\n",
+      "nearest[k] 2583\n",
+      "nearest[k] 2462\n",
+      "nearest[k] 2135\n",
+      "Nearest to MTD: POOL, CNMD, AIRM, WINA,\n",
+      "nearest[k] 2202\n",
+      "nearest[k] 2396\n",
+      "nearest[k] 1902\n",
+      "nearest[k] 2634\n",
+      "nearest[k] 2728\n",
+      "Nearest to GE: UFI, RE, ARCW,\n",
+      "nearest[k] 2874\n",
+      "nearest[k] 2969\n",
+      "nearest[k] 2726\n",
+      "Nearest to ICUI: IOSP, AZO, NFX, UBA, OSIS,\n",
+      "nearest[k] 2647\n",
+      "nearest[k] 2448\n",
+      "nearest[k] 2177\n",
+      "Nearest to ECL: NDSN, TRI, VSAT, IDXX, MXWL,\n",
+      "nearest[k] 2708\n",
+      "nearest[k] 3018\n",
+      "nearest[k] 2146\n",
+      "nearest[k] 2041\n",
+      "nearest[k] 2158\n",
+      "nearest[k] 1619\n",
+      "Nearest to INTC: DIS, BBOX,\n",
+      "nearest[k] 2427\n",
+      "nearest[k] 2207\n",
+      "nearest[k] 1712\n",
+      "nearest[k] 1760\n",
+      "nearest[k] 1967\n",
+      "nearest[k] 1966\n",
+      "Nearest to FAST: ACAS, CATY,\n",
+      "Epoch 1/10 Iteration: 410 Avg. Training loss: 3.7915 1.9852 sec/batch\n",
+      "Epoch 1/10 Iteration: 420 Avg. Training loss: 3.8297 1.9980 sec/batch\n",
+      "Epoch 1/10 Iteration: 430 Avg. Training loss: 3.7000 2.1386 sec/batch\n",
+      "Epoch 1/10 Iteration: 440 Avg. Training loss: 3.5404 2.0681 sec/batch\n",
+      "Epoch 1/10 Iteration: 450 Avg. Training loss: 3.4214 2.0249 sec/batch\n",
+      "Epoch 1/10 Iteration: 460 Avg. Training loss: 3.3465 2.0138 sec/batch\n",
+      "Epoch 1/10 Iteration: 470 Avg. Training loss: 3.4769 2.0312 sec/batch\n",
+      "Epoch 1/10 Iteration: 480 Avg. Training loss: 3.6643 2.0105 sec/batch\n",
+      "Epoch 1/10 Iteration: 490 Avg. Training loss: 3.8713 2.0370 sec/batch\n",
+      "Epoch 1/10 Iteration: 500 Avg. Training loss: 3.7673 2.0179 sec/batch\n",
+      "nearest[k] 2444\n",
+      "nearest[k] 2569\n",
+      "Nearest to ACAS: FAST, CATO, MYE, LM, BHB, VMI,\n",
+      "nearest[k] 1840\n",
+      "nearest[k] 2191\n",
+      "Nearest to RTN: HOLX, OLED, CLFD, PRKR, ARCW, ADS,\n",
+      "nearest[k] 1922\n",
+      "nearest[k] 1977\n",
+      "nearest[k] 2388\n",
+      "Nearest to SYKE: GNCMA, MANH, QCOM, PPL, BCOR,\n",
+      "nearest[k] 2439\n",
+      "nearest[k] 2354\n",
+      "nearest[k] 1745\n",
+      "Nearest to USAK: SENEA, MSFT, SPNC, JBL, ORCL,\n",
+      "nearest[k] 2088\n",
+      "nearest[k] 2785\n",
+      "nearest[k] 2370\n",
+      "nearest[k] 1818\n",
+      "nearest[k] 2747\n",
+      "nearest[k] 2323\n",
+      "Nearest to IMMU: STT, HUBG,\n",
+      "nearest[k] 2324\n",
+      "nearest[k] 1657\n",
+      "nearest[k] 2638\n",
+      "Nearest to UFPT: AKRX, PHX, RMTI, CWST, IDRA,\n",
+      "nearest[k] 1635\n",
+      "nearest[k] 2562\n",
+      "nearest[k] 1769\n",
+      "nearest[k] 2991\n",
+      "Nearest to SENEA: USAK, NNBR, RE, COLB,\n",
+      "nearest[k] 2406\n",
+      "nearest[k] 1679\n",
+      "Nearest to MOD: ENZN, PHM, DST, GHC, CTG, TDY,\n",
+      "nearest[k] 2220\n",
+      "nearest[k] 2719\n",
+      "nearest[k] 2772\n",
+      "nearest[k] 1632\n",
+      "nearest[k] 2278\n",
+      "Nearest to AMG: AAN, MATW, BHB,\n",
+      "Nearest to ORLY: GAIA, LCI, VAL, MIND, MNI, GSOL, RGR, MIDD,\n",
+      "nearest[k] 2316\n",
+      "nearest[k] 2583\n",
+      "nearest[k] 2135\n",
+      "Nearest to MTD: POOL, CNMD, WINA, AIRM, SLG,\n",
+      "nearest[k] 2202\n",
+      "nearest[k] 1902\n",
+      "nearest[k] 2396\n",
+      "nearest[k] 2728\n",
+      "Nearest to GE: RE, UFI, MCS, ARCW,\n",
+      "nearest[k] 2874\n",
+      "nearest[k] 2969\n",
+      "nearest[k] 2726\n",
+      "Nearest to ICUI: AZO, IOSP, OSIS, NFX, UBA,\n",
+      "nearest[k] 2647\n",
+      "nearest[k] 2177\n",
+      "Nearest to ECL: NDSN, TRI, IDXX, HOG, VSAT, ANSS,\n",
+      "nearest[k] 2708\n",
+      "nearest[k] 3018\n",
+      "nearest[k] 2041\n",
+      "nearest[k] 2146\n",
+      "nearest[k] 2158\n",
+      "nearest[k] 1619\n",
+      "Nearest to INTC: DIS, NUE,\n",
+      "nearest[k] 1712\n",
+      "nearest[k] 1760\n",
+      "nearest[k] 2207\n",
+      "nearest[k] 2427\n",
+      "Nearest to FAST: ACAS, FRED, CATY, LNN,\n",
+      "Epoch 1/10 Iteration: 510 Avg. Training loss: 3.6362 2.0846 sec/batch\n",
+      "Epoch 1/10 Iteration: 520 Avg. Training loss: 3.6206 2.1565 sec/batch\n",
+      "Epoch 1/10 Iteration: 530 Avg. Training loss: 3.5182 2.0636 sec/batch\n",
+      "Epoch 1/10 Iteration: 540 Avg. Training loss: 3.4995 2.0273 sec/batch\n",
+      "Epoch 1/10 Iteration: 550 Avg. Training loss: 3.2319 2.0127 sec/batch\n",
+      "Epoch 1/10 Iteration: 560 Avg. Training loss: 3.2393 1.9948 sec/batch\n",
+      "Epoch 1/10 Iteration: 570 Avg. Training loss: 3.2264 2.0412 sec/batch\n",
+      "Epoch 1/10 Iteration: 580 Avg. Training loss: 3.3506 2.0951 sec/batch\n"
+     ]
+    },
+    {
+     "ename": "KeyboardInterrupt",
+     "evalue": "",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
+      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
+      "\u001b[0;32m<ipython-input-52-d00979298fdd>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     14\u001b[0m         \u001b[0mbatches\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mget_batches\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdf_prices\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_size\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mwindow_size\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     15\u001b[0m         \u001b[0mstart\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtime\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtime\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 16\u001b[0;31m         \u001b[0;32mfor\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mbatches\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     17\u001b[0m             feed = {inputs: x,\n\u001b[1;32m     18\u001b[0m                     labels: np.array(y)[:, None]}\n",
+      "\u001b[0;32m<ipython-input-44-ce89646d590b>\u001b[0m in \u001b[0;36mget_batches\u001b[0;34m(stocks, batch_size, window_size)\u001b[0m\n\u001b[1;32m      6\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstart\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mstop\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      7\u001b[0m             \u001b[0mbatch_x\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mget_stock_int\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstocks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mi\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 8\u001b[0;31m             \u001b[0mbatch_y\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mget_window\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstocks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mi\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mwindow_size\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      9\u001b[0m             \u001b[0my\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mextend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbatch_y\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     10\u001b[0m             \u001b[0mx\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mextend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mbatch_x\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbatch_y\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+      "\u001b[0;32m<ipython-input-46-959e2d122317>\u001b[0m in \u001b[0;36mget_window\u001b[0;34m(stocks, idx, window_size)\u001b[0m\n\u001b[1;32m     30\u001b[0m     \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstart\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mstop\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     31\u001b[0m         \u001b[0mnearby_stock_int\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mget_stock_int\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstocks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mi\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 32\u001b[0;31m         \u001b[0mnearby_stock_date\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mget_stock_date\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstocks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mi\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     33\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mnearby_stock_int\u001b[0m \u001b[0;34m!=\u001b[0m \u001b[0mstock_int\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0mnearby_stock_date\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0mstock_date\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     34\u001b[0m             \u001b[0mwindow\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnearby_stock_int\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+      "\u001b[0;32m<ipython-input-46-959e2d122317>\u001b[0m in \u001b[0;36mget_stock_date\u001b[0;34m(stocks, idx)\u001b[0m\n\u001b[1;32m     10\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     11\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mget_stock_date\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstocks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0midx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 12\u001b[0;31m     \u001b[0;32mreturn\u001b[0m \u001b[0mstocks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0miloc\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0midx\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     13\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     14\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mget_stock_ticker\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstocks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0midx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+      "\u001b[0;32m/Users/jperl/anaconda/envs/dlnd/lib/python3.6/site-packages/pandas/core/indexing.py\u001b[0m in \u001b[0;36m__getitem__\u001b[0;34m(self, key)\u001b[0m\n\u001b[1;32m   1310\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_getitem_tuple\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1311\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1312\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_getitem_axis\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1313\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1314\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_getitem_axis\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkey\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+      "\u001b[0;32m/Users/jperl/anaconda/envs/dlnd/lib/python3.6/site-packages/pandas/core/indexing.py\u001b[0m in \u001b[0;36m_getitem_axis\u001b[0;34m(self, key, axis)\u001b[0m\n\u001b[1;32m   1628\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_is_valid_integer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1629\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1630\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_get_loc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0maxis\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1631\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1632\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_convert_to_indexer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mis_setter\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+      "\u001b[0;32m/Users/jperl/anaconda/envs/dlnd/lib/python3.6/site-packages/pandas/core/indexing.py\u001b[0m in \u001b[0;36m_get_loc\u001b[0;34m(self, key, axis)\u001b[0m\n\u001b[1;32m    103\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    104\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_get_loc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkey\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 105\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mobj\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_ixs\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkey\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0maxis\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    106\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    107\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_slice\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mobj\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkind\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+      "\u001b[0;32m/Users/jperl/anaconda/envs/dlnd/lib/python3.6/site-packages/pandas/core/frame.py\u001b[0m in \u001b[0;36m_ixs\u001b[0;34m(self, i, axis)\u001b[0m\n\u001b[1;32m   1974\u001b[0m                     \u001b[0mcopy\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mTrue\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1975\u001b[0m                 \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1976\u001b[0;31m                     \u001b[0mnew_values\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_data\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfast_xs\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1977\u001b[0m                     \u001b[0;32mif\u001b[0m \u001b[0mis_scalar\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnew_values\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1978\u001b[0m                         \u001b[0;32mreturn\u001b[0m \u001b[0mnew_values\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+      "\u001b[0;32m/Users/jperl/anaconda/envs/dlnd/lib/python3.6/site-packages/pandas/core/internals.py\u001b[0m in \u001b[0;36mfast_xs\u001b[0;34m(self, loc)\u001b[0m\n\u001b[1;32m   3500\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3501\u001b[0m         \u001b[0;31m# unique\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 3502\u001b[0;31m         \u001b[0mdtype\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_interleaved_dtype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mblocks\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   3503\u001b[0m         \u001b[0mn\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mitems\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3504\u001b[0m         \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mempty\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mn\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdtype\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdtype\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+      "\u001b[0;32m/Users/jperl/anaconda/envs/dlnd/lib/python3.6/site-packages/pandas/core/internals.py\u001b[0m in \u001b[0;36m_interleaved_dtype\u001b[0;34m(blocks)\u001b[0m\n\u001b[1;32m   4462\u001b[0m     \u001b[0mcounts\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdefaultdict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlist\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   4463\u001b[0m     \u001b[0;32mfor\u001b[0m \u001b[0mx\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mblocks\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 4464\u001b[0;31m         \u001b[0mcounts\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mtype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   4465\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   4466\u001b[0m     \u001b[0mhave_int\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcounts\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mIntBlock\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m>\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
+     ]
+    }
+   ],
    "source": [
-    "epochs = 50\n",
+    "epochs = 10\n",
     "batch_size = 1000\n",
-    "window_size = 5\n",
+    "window_size = 10\n",
     "\n",
     "with train_graph.as_default():\n",
     "    saver = tf.train.Saver()\n",
@@ -508,7 +902,7 @@
     "    sess.run(tf.global_variables_initializer())\n",
     "\n",
     "    for e in range(1, epochs+1):\n",
-    "        batches = get_batches(df, batch_size, window_size)\n",
+    "        batches = get_batches(df_prices, batch_size, window_size)\n",
     "        start = time.time()\n",
     "        for x, y in batches:           \n",
     "            feed = {inputs: x,\n",