├── ARIMA.ipynb
├── ARIMA_v2.ipynb
├── Continuous_Stream_Data.py
├── Continuous_Stream_Sentiment.py
├── LSTM_FINAL.ipynb
├── LSTM_V1.ipynb
├── README.md
├── bda_data_plot.PNG
├── bitcoinprices.txt
├── engine.py
├── future_virtual_trading.py
├── get_bitcoin_new.py
├── get_bitcoin_price.py
├── get_twitter.py
├── get_twitter_new.py
├── live_bitcoin.csv
├── live_tweet.csv
├── sentiment6.txt
└── tweet_data


/ARIMA_v2.ipynb:
--------------------------------------------------------------------------------
  1 | {
  2 |  "cells": [
  3 |   {
  4 |    "cell_type": "code",
  5 |    "execution_count": 1,
  6 |    "metadata": {},
  7 |    "outputs": [],
  8 |    "source": [
  9 |     "import pyspark\n",
 10 |     "from pyspark import SparkContext,SQLContext\n",
 11 |     "sc = SparkContext.getOrCreate()"
 12 |    ]
 13 |   },
 14 |   {
 15 |    "cell_type": "code",
 16 |    "execution_count": null,
 17 |    "metadata": {
 18 |     "collapsed": true
 19 |    },
 20 |    "outputs": [],
 21 |    "source": [
 22 |     "bitCoinPrices = sc.textFile('bitcoinprices.txt')"
 23 |    ]
 24 |   },
 25 |   {
 26 |    "cell_type": "code",
 27 |    "execution_count": null,
 28 |    "metadata": {
 29 |     "collapsed": true
 30 |    },
 31 |    "outputs": [],
 32 |    "source": [
 33 |     "bitCoinPrices.map(lambda x: x.split(',')).take(10)"
 34 |    ]
 35 |   },
 36 |   {
 37 |    "cell_type": "code",
 38 |    "execution_count": null,
 39 |    "metadata": {
 40 |     "collapsed": true
 41 |    },
 42 |    "outputs": [],
 43 |    "source": [
 44 |     "bitCoinPrices=spark.read.csv('bitcoinprices.txt', mode=\"DROPMALFORMED\",inferSchema=True, header = False)"
 45 |    ]
 46 |   },
 47 |   {
 48 |    "cell_type": "code",
 49 |    "execution_count": null,
 50 |    "metadata": {
 51 |     "collapsed": true
 52 |    },
 53 |    "outputs": [],
 54 |    "source": [
 55 |     "bitCoinPrices = bitCoinPrices.withColumnRenamed('_c0','date').withColumnRenamed('_c1','price')"
 56 |    ]
 57 |   },
 58 |   {
 59 |    "cell_type": "code",
 60 |    "execution_count": null,
 61 |    "metadata": {
 62 |     "collapsed": true
 63 |    },
 64 |    "outputs": [],
 65 |    "source": [
 66 |     "bitCoinPrices.sort('date').head(657)"
 67 |    ]
 68 |   },
 69 |   {
 70 |    "cell_type": "code",
 71 |    "execution_count": null,
 72 |    "metadata": {
 73 |     "collapsed": true
 74 |    },
 75 |    "outputs": [],
 76 |    "source": [
 77 |     "tweetSents=spark.read.csv('sentiment6.txt', mode=\"DROPMALFORMED\",inferSchema=True, header = False)"
 78 |    ]
 79 |   },
 80 |   {
 81 |    "cell_type": "code",
 82 |    "execution_count": null,
 83 |    "metadata": {
 84 |     "collapsed": true
 85 |    },
 86 |    "outputs": [],
 87 |    "source": [
 88 |     "tweetSents = tweetSents.withColumnRenamed('_c0','date').withColumnRenamed('_c1','SentimentProb')"
 89 |    ]
 90 |   },
 91 |   {
 92 |    "cell_type": "code",
 93 |    "execution_count": null,
 94 |    "metadata": {
 95 |     "collapsed": true
 96 |    },
 97 |    "outputs": [],
 98 |    "source": [
 99 |     "tweetSents.sort('date').head(643)"
100 |    ]
101 |   },
102 |   {
103 |    "cell_type": "code",
104 |    "execution_count": null,
105 |    "metadata": {
106 |     "collapsed": true
107 |    },
108 |    "outputs": [],
109 |    "source": [
110 |     "df1 = flintContext.read.pandas(pd.read_csv(bitcoinprices.txt))"
111 |    ]
112 |   },
113 |   {
114 |    "cell_type": "code",
115 |    "execution_count": 3,
116 |    "metadata": {
117 |     "collapsed": true
118 |    },
119 |    "outputs": [],
120 |    "source": [
121 |     "import pandas"
122 |    ]
123 |   },
124 |   {
125 |    "cell_type": "code",
126 |    "execution_count": 2,
127 |    "metadata": {},
128 |    "outputs": [
129 |     {
130 |      "ename": "NameError",
131 |      "evalue": "name 'flintContext' is not defined",
132 |      "output_type": "error",
133 |      "traceback": [
134 |       "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
135 |       "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
136 |       "\u001b[1;32m<ipython-input-2-0572c1bfc892>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mdf1\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mflintContext\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mread\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpandas\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mpd\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mread_csv\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mbitcoinprices\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtxt\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
137 |       "\u001b[1;31mNameError\u001b[0m: name 'flintContext' is not defined"
138 |      ]
139 |     }
140 |    ],
141 |    "source": [
142 |     "df1 = flintContext.read.pandas(pd.read_csv(bitcoinprices.txt))"
143 |    ]
144 |   },
145 |   {
146 |    "cell_type": "code",
147 |    "execution_count": 4,
148 |    "metadata": {},
149 |    "outputs": [
150 |     {
151 |      "ename": "ImportError",
152 |      "evalue": "No module named 'ts'",
153 |      "output_type": "error",
154 |      "traceback": [
155 |       "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
156 |       "\u001b[1;31mImportError\u001b[0m                               Traceback (most recent call last)",
157 |       "\u001b[1;32m<ipython-input-4-db01aeac061b>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[1;32mimport\u001b[0m \u001b[0mts\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mflint\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      2\u001b[0m \u001b[1;32mfrom\u001b[0m \u001b[0mts\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mflint\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0mFlintContext\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
158 |       "\u001b[1;31mImportError\u001b[0m: No module named 'ts'"
159 |      ]
160 |     }
161 |    ],
162 |    "source": [
163 |     "import ts.flint\n",
164 |     "from ts.flint import FlintContext"
165 |    ]
166 |   },
167 |   {
168 |    "cell_type": "code",
169 |    "execution_count": 19,
170 |    "metadata": {
171 |     "scrolled": true
172 |    },
173 |    "outputs": [
174 |     {
175 |      "name": "stdout",
176 |      "output_type": "stream",
177 |      "text": [
178 |       "                             ARIMA Model Results                              \n",
179 |       "==============================================================================\n",
180 |       "Dep. Variable:                D.427.1   No. Observations:                  655\n",
181 |       "Model:                ARIMA(10, 1, 0)   Log Likelihood               -5765.509\n",
182 |       "Method:                       css-mle   S.D. of innovations           1605.919\n",
183 |       "Date:                Wed, 15 Nov 2017   AIC                          11555.018\n",
184 |       "Time:                        21:44:04   BIC                          11608.833\n",
185 |       "Sample:                    06-13-2016   HQIC                         11575.884\n",
186 |       "                         - 05-24-2017                                         \n",
187 |       "==================================================================================\n",
188 |       "                     coef    std err          z      P>|z|      [0.025      0.975]\n",
189 |       "----------------------------------------------------------------------------------\n",
190 |       "const              0.2322      7.744      0.030      0.976     -14.946      15.410\n",
191 |       "ar.L1.D.427.1     -0.9627      0.039    -24.934      0.000      -1.038      -0.887\n",
192 |       "ar.L2.D.427.1     -0.9628      0.052    -18.344      0.000      -1.066      -0.860\n",
193 |       "ar.L3.D.427.1     -0.9123      0.062    -14.804      0.000      -1.033      -0.791\n",
194 |       "ar.L4.D.427.1     -0.9390      0.067    -14.099      0.000      -1.070      -0.808\n",
195 |       "ar.L5.D.427.1     -0.8788      0.069    -12.784      0.000      -1.014      -0.744\n",
196 |       "ar.L6.D.427.1     -0.8319      0.069    -12.123      0.000      -0.966      -0.697\n",
197 |       "ar.L7.D.427.1     -0.6530      0.066     -9.831      0.000      -0.783      -0.523\n",
198 |       "ar.L8.D.427.1     -0.5174      0.061     -8.412      0.000      -0.638      -0.397\n",
199 |       "ar.L9.D.427.1     -0.3347      0.053     -6.375      0.000      -0.438      -0.232\n",
200 |       "ar.L10.D.427.1    -0.1598      0.039     -4.133      0.000      -0.236      -0.084\n",
201 |       "                                    Roots                                     \n",
202 |       "==============================================================================\n",
203 |       "                  Real           Imaginary           Modulus         Frequency\n",
204 |       "------------------------------------------------------------------------------\n",
205 |       "AR.1             0.8465           -0.6878j            1.0907           -0.1086\n",
206 |       "AR.2             0.8465           +0.6878j            1.0907            0.1086\n",
207 |       "AR.3             0.4148           -1.1292j            1.2030           -0.1940\n",
208 |       "AR.4             0.4148           +1.1292j            1.2030            0.1940\n",
209 |       "AR.5            -0.2562           -1.1801j            1.2076           -0.2840\n",
210 |       "AR.6            -0.2562           +1.1801j            1.2076            0.2840\n",
211 |       "AR.7            -1.1750           -0.4087j            1.2441           -0.4467\n",
212 |       "AR.8            -1.1750           +0.4087j            1.2441            0.4467\n",
213 |       "AR.9            -0.8773           -0.9170j            1.2690           -0.3715\n",
214 |       "AR.10           -0.8773           +0.9170j            1.2690            0.3715\n",
215 |       "------------------------------------------------------------------------------\n"
216 |      ]
217 |     },
218 |     {
219 |      "data": {
220 |       "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYkAAAEICAYAAACqMQjAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
221 |       "text/plain": [
222 |        "<matplotlib.figure.Figure at 0x1d9007099e8>"
223 |       ]
224 |      },
225 |      "metadata": {},
226 |      "output_type": "display_data"
227 |     },
228 |     {
229 |      "data": {
230 |       "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZ4AAAD8CAYAAABHN8LqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
231 |       "text/plain": [
232 |        "<matplotlib.figure.Figure at 0x1d9017188d0>"
233 |       ]
234 |      },
235 |      "metadata": {},
236 |      "output_type": "display_data"
237 |     },
238 |     {
239 |      "name": "stdout",
240 |      "output_type": "stream",
241 |      "text": [
242 |       "                 0\n",
243 |       "count   655.000000\n",
244 |       "mean      4.884351\n",
245 |       "std    1608.617675\n",
246 |       "min   -2960.046887\n",
247 |       "25%    -998.792861\n",
248 |       "50%    -468.573260\n",
249 |       "75%     482.996348\n",
250 |       "max    6497.914552\n"
251 |      ]
252 |     }
253 |    ],
254 |    "source": [
255 |     "from pandas import read_csv\n",
256 |     "from pandas import datetime\n",
257 |     "from pandas import DataFrame\n",
258 |     "from statsmodels.tsa.arima_model import ARIMA\n",
259 |     "from matplotlib import pyplot\n",
260 |     "def parser(x):\n",
261 |     "    return datetime.strptime(x, '%Y%m%d')\n",
262 |     "\n",
263 |     "series = read_csv('bitcoinprices.txt', header=0, parse_dates=[0], index_col=0, squeeze=True, date_parser=parser)\n",
264 |     "# fit model\n",
265 |     "model = ARIMA(series, order=(10,1,0))\n",
266 |     "model_fit = model.fit(disp=0)\n",
267 |     "print(model_fit.summary())\n",
268 |     "# plot residual errors\n",
269 |     "residuals = DataFrame(model_fit.resid)\n",
270 |     "residuals.plot()\n",
271 |     "pyplot.show()\n",
272 |     "residuals.plot(kind='kde')\n",
273 |     "pyplot.show()\n",
274 |     "print(residuals.describe())"
275 |    ]
276 |   },
277 |   {
278 |    "cell_type": "code",
279 |    "execution_count": 44,
280 |    "metadata": {},
281 |    "outputs": [
282 |     {
283 |      "name": "stdout",
284 |      "output_type": "stream",
285 |      "text": [
286 |       "predicted=1131.924869, expected=1160.800000\n",
287 |       "predicted=1170.547150, expected=1185.400000\n",
288 |       "predicted=1176.787441, expected=1189.500000\n",
289 |       "predicted=1205.632655, expected=1186.200000\n",
290 |       "predicted=1181.409992, expected=1209.500000\n",
291 |       "predicted=1207.497883, expected=1203.600000\n",
292 |       "predicted=1198.542001, expected=1228.400000\n",
293 |       "predicted=1234.204510, expected=1211.700000\n",
294 |       "predicted=1214.034524, expected=1178.900000\n",
295 |       "predicted=1179.338993, expected=1180.600000\n",
296 |       "predicted=1186.100887, expected=1188.700000\n",
297 |       "predicted=1188.088996, expected=1171.800000\n",
298 |       "predicted=1184.020209, expected=1193.000000\n",
299 |       "predicted=1190.549582, expected=1196.400000\n",
300 |       "predicted=1191.032111, expected=1219.000000\n",
301 |       "predicted=1218.158646, expected=1248.900000\n",
302 |       "predicted=1250.968988, expected=1249.900000\n",
303 |       "predicted=1247.991102, expected=1250.000000\n",
304 |       "predicted=1255.063327, expected=1254.200000\n",
305 |       "predicted=1251.173412, expected=1267.000000\n",
306 |       "predicted=1273.833156, expected=1290.300000\n",
307 |       "predicted=1294.878080, expected=1318.100000\n",
308 |       "predicted=1317.511181, expected=1358.600000\n",
309 |       "predicted=1355.473493, expected=1350.100000\n",
310 |       "predicted=1344.784704, expected=1349.800000\n",
311 |       "predicted=1350.949127, expected=1374.700000\n",
312 |       "predicted=1377.227513, expected=1460.000000\n",
313 |       "predicted=1464.127557, expected=1469.000000\n",
314 |       "predicted=1475.129079, expected=1560.200000\n",
315 |       "predicted=1555.679824, expected=1620.000000\n",
316 |       "predicted=1613.594885, expected=1586.900000\n",
317 |       "predicted=1583.958339, expected=1599.400000\n",
318 |       "predicted=1620.590858, expected=1620.700000\n",
319 |       "predicted=1617.525455, expected=1706.000000\n",
320 |       "predicted=1732.566885, expected=1727.900000\n",
321 |       "predicted=1737.132850, expected=1838.500000\n",
322 |       "predicted=1838.684060, expected=1787.900000\n",
323 |       "predicted=1784.040613, expected=1699.900000\n",
324 |       "predicted=1698.041310, expected=1799.800000\n",
325 |       "predicted=1824.508692, expected=1774.900000\n",
326 |       "predicted=1784.460201, expected=1744.100000\n",
327 |       "predicted=1789.977027, expected=1829.000000\n",
328 |       "predicted=1810.583733, expected=1804.000000\n",
329 |       "predicted=1776.289599, expected=1924.900000\n",
330 |       "predicted=1938.665334, expected=1993.000000\n",
331 |       "predicted=1982.087626, expected=2073.400000\n",
332 |       "predicted=2077.787080, expected=2173.300000\n",
333 |       "predicted=2173.144998, expected=2166.900000\n",
334 |       "predicted=2177.340053, expected=2312.100000\n",
335 |       "predicted=2353.478398, expected=2675.000000\n",
336 |       "predicted=2729.538515, expected=2434.300000\n",
337 |       "predicted=2521.227597, expected=2126.700000\n",
338 |       "predicted=2192.437465, expected=2219.900000\n",
339 |       "predicted=2267.553479, expected=2226.900000\n",
340 |       "predicted=2210.937489, expected=2287.000000\n",
341 |       "predicted=2382.663109, expected=2188.000000\n",
342 |       "predicted=2076.024796, expected=2377.900000\n",
343 |       "predicted=2427.001155, expected=2434.900000\n",
344 |       "predicted=2285.718401, expected=2458.200000\n",
345 |       "predicted=2465.057423, expected=2518.400000\n",
346 |       "predicted=2545.356883, expected=2584.900000\n",
347 |       "predicted=2531.707461, expected=2845.800000\n",
348 |       "predicted=2854.724686, expected=2818.700000\n",
349 |       "predicted=2784.662659, expected=2712.000000\n",
350 |       "predicted=2760.327596, expected=2727.700000\n",
351 |       "predicted=2753.909838, expected=2814.100000\n",
352 |       "predicted=2810.864977, expected=2909.500000\n",
353 |       "predicted=2898.695116, expected=2978.900000\n",
354 |       "predicted=2972.391247, expected=2749.900000\n",
355 |       "predicted=2749.528822, expected=2800.000000\n",
356 |       "predicted=2845.162376, expected=2362.900000\n",
357 |       "predicted=2355.875274, expected=2455.700000\n",
358 |       "predicted=2516.245923, expected=2466.700000\n",
359 |       "predicted=2482.689243, expected=2660.000000\n",
360 |       "predicted=2602.144612, expected=2574.000000\n",
361 |       "predicted=2560.706251, expected=2624.400000\n",
362 |       "predicted=2530.643286, expected=2696.800000\n",
363 |       "predicted=2764.077180, expected=2560.000000\n",
364 |       "predicted=2560.586733, expected=2712.100000\n",
365 |       "predicted=2745.801813, expected=2706.900000\n",
366 |       "predicted=2784.942755, expected=2535.800000\n",
367 |       "predicted=2540.127330, expected=2475.100000\n",
368 |       "predicted=2479.719725, expected=2400.900000\n",
369 |       "predicted=2435.177319, expected=2487.200000\n",
370 |       "predicted=2481.743783, expected=2564.800000\n",
371 |       "predicted=2554.993948, expected=2515.700000\n",
372 |       "predicted=2528.072223, expected=2491.900000\n",
373 |       "predicted=2477.426457, expected=2415.200000\n",
374 |       "predicted=2427.048630, expected=2499.900000\n",
375 |       "predicted=2528.917717, expected=2627.000000\n",
376 |       "predicted=2639.786744, expected=2562.700000\n",
377 |       "predicted=2569.524658, expected=2618.200000\n",
378 |       "predicted=2597.334077, expected=2577.000000\n",
379 |       "predicted=2576.844470, expected=2498.600000\n",
380 |       "predicted=2518.992635, expected=2555.500000\n",
381 |       "predicted=2569.756363, expected=2513.800000\n",
382 |       "predicted=2542.963449, expected=2343.400000\n",
383 |       "predicted=2350.918711, expected=2312.500000\n",
384 |       "predicted=2313.503663, expected=2384.000000\n",
385 |       "predicted=2404.339863, expected=2337.500000\n",
386 |       "predicted=2337.357382, expected=2106.100000\n",
387 |       "predicted=2128.183443, expected=1984.800000\n",
388 |       "predicted=2024.570623, expected=2028.400000\n",
389 |       "predicted=2044.497974, expected=2238.400000\n",
390 |       "predicted=2237.037219, expected=2348.300000\n",
391 |       "predicted=2335.056754, expected=2319.100000\n",
392 |       "predicted=2292.173954, expected=2759.700000\n",
393 |       "predicted=2757.917788, expected=2752.100000\n",
394 |       "predicted=2745.357688, expected=2788.000000\n",
395 |       "predicted=2805.702360, expected=2741.000000\n",
396 |       "predicted=2770.701086, expected=2763.600000\n",
397 |       "predicted=2775.207446, expected=2451.000000\n",
398 |       "predicted=2400.437535, expected=2560.000000\n",
399 |       "predicted=2516.170788, expected=2682.100000\n",
400 |       "predicted=2671.490654, expected=2740.900000\n",
401 |       "predicted=2669.225041, expected=2741.000000\n",
402 |       "predicted=2746.130293, expected=2722.100000\n",
403 |       "predicted=2728.390130, expected=2863.000000\n",
404 |       "predicted=2861.590586, expected=2732.000000\n",
405 |       "predicted=2740.457117, expected=2711.600000\n",
406 |       "predicted=2738.536177, expected=2779.500000\n",
407 |       "predicted=2785.808733, expected=3112.000000\n",
408 |       "predicted=3065.507445, expected=3207.400000\n",
409 |       "predicted=3167.859172, expected=3222.500000\n",
410 |       "predicted=3203.455583, expected=3402.900000\n",
411 |       "predicted=3403.932412, expected=3402.500000\n",
412 |       "predicted=3372.562346, expected=3362.000000\n",
413 |       "predicted=3335.927739, expected=3473.200000\n",
414 |       "predicted=3497.394115, expected=3761.800000\n",
415 |       "predicted=3755.427507, expected=4094.500000\n",
416 |       "predicted=4072.590915, expected=4074.900000\n",
417 |       "predicted=4061.443886, expected=4220.000000\n",
418 |       "predicted=4249.376390, expected=4061.300000\n",
419 |       "predicted=4055.315128, expected=4319.000000\n",
420 |       "predicted=4318.599109, expected=4269.800000\n",
421 |       "predicted=4334.075520, expected=4045.400000\n",
422 |       "predicted=4064.022979, expected=4098.000000\n",
423 |       "predicted=4112.813583, expected=4002.100000\n",
424 |       "predicted=4007.746395, expected=3847.400000\n",
425 |       "predicted=3799.196501, expected=4090.200000\n",
426 |       "predicted=4095.203236, expected=4164.000000\n",
427 |       "predicted=4179.896762, expected=4320.000000\n",
428 |       "predicted=4285.863197, expected=4320.500000\n",
429 |       "predicted=4324.302924, expected=4356.000000\n",
430 |       "predicted=4381.049635, expected=4265.500000\n",
431 |       "predicted=4256.278756, expected=4371.200000\n",
432 |       "predicted=4380.932643, expected=4604.000000\n",
433 |       "predicted=4634.811493, expected=4610.200000\n",
434 |       "predicted=4607.534484, expected=4713.300000\n",
435 |       "predicted=4711.653426, expected=4750.000000\n",
436 |       "predicted=4755.003398, expected=4660.000000\n",
437 |       "predicted=4652.487901, expected=4566.700000\n",
438 |       "predicted=4575.943206, expected=4337.300000\n",
439 |       "predicted=4365.519103, expected=4510.500000\n",
440 |       "predicted=4521.438243, expected=4538.700000\n",
441 |       "predicted=4537.225584, expected=4621.700000\n",
442 |       "predicted=4626.139651, expected=4436.600000\n",
443 |       "predicted=4440.605955, expected=4280.100000\n",
444 |       "predicted=4272.733130, expected=4184.100000\n",
445 |       "predicted=4179.631284, expected=4347.700000\n",
446 |       "predicted=4377.896937, expected=4020.800000\n",
447 |       "predicted=4061.150513, expected=3830.600000\n",
448 |       "predicted=3832.917484, expected=3167.300000\n",
449 |       "predicted=3148.843922, expected=3789.900000\n",
450 |       "predicted=3721.181918, expected=3643.200000\n",
451 |       "predicted=3745.897799, expected=3900.000000\n",
452 |       "predicted=3708.258332, expected=3969.200000\n",
453 |       "predicted=4052.890232, expected=3944.900000\n",
454 |       "predicted=3899.139036, expected=3871.900000\n",
455 |       "predicted=3898.634597, expected=3753.200000\n",
456 |       "predicted=3763.655407, expected=3712.900000\n",
457 |       "predicted=3799.038340, expected=3659.900000\n",
458 |       "predicted=3595.991489, expected=3773.000000\n",
459 |       "predicted=3810.407828, expected=3886.700000\n",
460 |       "predicted=3883.985961, expected=3933.600000\n",
461 |       "predicted=3934.820535, expected=4197.900000\n",
462 |       "predicted=4177.994727, expected=4051.800000\n",
463 |       "predicted=4076.279921, expected=4244.900000\n",
464 |       "predicted=4186.023139, expected=4315.000000\n",
465 |       "predicted=4348.559501, expected=4409.000000\n",
466 |       "predicted=4396.137765, expected=4292.000000\n",
467 |       "predicted=4302.816002, expected=4332.300000\n",
468 |       "predicted=4339.602687, expected=4157.600000\n",
469 |       "predicted=4200.457495, expected=4394.700000\n",
470 |       "predicted=4350.026625, expected=4345.000000\n",
471 |       "predicted=4423.837030, expected=4463.600000\n",
472 |       "predicted=4414.030733, expected=4561.000000\n",
473 |       "predicted=4590.717464, expected=4827.000000\n",
474 |       "predicted=4804.070276, expected=4764.100000\n",
475 |       "predicted=4791.744934, expected=4896.900000\n",
476 |       "predicted=4848.183540, expected=5511.700000\n",
477 |       "predicted=5526.703045, expected=5585.700000\n",
478 |       "predicted=5611.475654, expected=5739.900000\n",
479 |       "predicted=5726.468306, expected=5661.000000\n",
480 |       "predicted=5739.215874, expected=5637.600000\n",
481 |       "predicted=5638.735624, expected=5372.200000\n",
482 |       "predicted=5390.497509, expected=5572.300000\n",
483 |       "predicted=5584.648964, expected=5691.400000\n",
484 |       "predicted=5746.262773, expected=6080.000000\n",
485 |       "predicted=6061.883044, expected=5861.700000\n",
486 |       "predicted=5940.101643, expected=5996.000000\n",
487 |       "predicted=5918.916990, expected=5820.500000\n",
488 |       "predicted=5887.679333, expected=5470.000000\n",
489 |       "predicted=5441.629503, expected=5735.000000\n",
490 |       "predicted=5755.150264, expected=5895.000000\n",
491 |       "predicted=5972.697175, expected=5768.100000\n",
492 |       "predicted=5771.727576, expected=5795.000000\n",
493 |       "predicted=5778.850580, expected=6125.000000\n",
494 |       "predicted=6108.006964, expected=6135.300000\n",
495 |       "predicted=6096.833865, expected=6406.500000\n",
496 |       "predicted=6404.500817, expected=6657.300000\n",
497 |       "predicted=6695.266464, expected=7210.100000\n",
498 |       "predicted=7164.324354, expected=7191.000000\n",
499 |       "predicted=7266.238541, expected=7489.900000\n",
500 |       "predicted=7512.093253, expected=7316.700000\n",
501 |       "predicted=7407.825367, expected=7186.800000\n",
502 |       "predicted=7135.793483, expected=7364.300000\n",
503 |       "predicted=7440.824509, expected=7376.900000\n",
504 |       "predicted=7435.657469, expected=7195.000000\n",
505 |       "predicted=7208.926001, expected=6704.100000\n",
506 |       "predicted=6700.998647, expected=6079.300000\n",
507 |       "predicted=5965.979839, expected=6241.000000\n",
508 |       "predicted=6174.557138, expected=6570.000000\n",
509 |       "predicted=6682.627290, expected=6565.000000\n"
510 |      ]
511 |     }
512 |    ],
513 |    "source": [
514 |     "from pandas import read_csv\n",
515 |     "from pandas import datetime\n",
516 |     "from pandas import DataFrame\n",
517 |     "from statsmodels.tsa.arima_model import ARIMA\n",
518 |     "from matplotlib import pyplot\n",
519 |     "import numpy as np\n",
520 |     "def parser(x):    \n",
521 |     "    return datetime.strptime(x, '%Y-%m-%d')\n",
522 |     "\n",
523 |     "series = read_csv('bitcoinsort.csv', header='infer', dtype={'time':np.str,'price':np.float64},parse_dates=[0], index_col=0, squeeze=True, date_parser=parser)\n",
524 |     "\n",
525 |     "X = series.values\n",
526 |     "size = int(len(X) * 0.66)\n",
527 |     "train, test = X[0:size], X[size:len(X)]\n",
528 |     "history = [x for x in train]\n",
529 |     "predictions = list()\n",
530 |     "for t in range(len(test)):\n",
531 |     "    model = ARIMA(history, order=(10,1,0))\n",
532 |     "    model_fit = model.fit(disp=0)\n",
533 |     "    output = model_fit.forecast()\n",
534 |     "    yhat = output[0]\n",
535 |     "    predictions.append(yhat)\n",
536 |     "    obs = test[t]\n",
537 |     "    history.append(obs)\n",
538 |     "    print('predicted=%f, expected=%f' % (yhat, obs))"
539 |    ]
540 |   },
541 |   {
542 |    "cell_type": "code",
543 |    "execution_count": 30,
544 |    "metadata": {},
545 |    "outputs": [
546 |     {
547 |      "data": {
548 |       "text/html": [
549 |        "<div>\n",
550 |        "<style scoped>\n",
551 |        "    .dataframe tbody tr th:only-of-type {\n",
552 |        "        vertical-align: middle;\n",
553 |        "    }\n",
554 |        "\n",
555 |        "    .dataframe tbody tr th {\n",
556 |        "        vertical-align: top;\n",
557 |        "    }\n",
558 |        "\n",
559 |        "    .dataframe thead th {\n",
560 |        "        text-align: right;\n",
561 |        "    }\n",
562 |        "</style>\n",
563 |        "<table border=\"1\" class=\"dataframe\">\n",
564 |        "  <thead>\n",
565 |        "    <tr style=\"text-align: right;\">\n",
566 |        "      <th></th>\n",
567 |        "      <th>time</th>\n",
568 |        "      <th>price</th>\n",
569 |        "    </tr>\n",
570 |        "  </thead>\n",
571 |        "  <tbody>\n",
572 |        "    <tr>\n",
573 |        "      <th>0</th>\n",
574 |        "      <td>2016-02-27</td>\n",
575 |        "      <td>427.1</td>\n",
576 |        "    </tr>\n",
577 |        "    <tr>\n",
578 |        "      <th>1</th>\n",
579 |        "      <td>2016-11-18</td>\n",
580 |        "      <td>750.9</td>\n",
581 |        "    </tr>\n",
582 |        "    <tr>\n",
583 |        "      <th>2</th>\n",
584 |        "      <td>2016-06-13</td>\n",
585 |        "      <td>690.9</td>\n",
586 |        "    </tr>\n",
587 |        "    <tr>\n",
588 |        "      <th>3</th>\n",
589 |        "      <td>2016-08-08</td>\n",
590 |        "      <td>588.7</td>\n",
591 |        "    </tr>\n",
592 |        "    <tr>\n",
593 |        "      <th>4</th>\n",
594 |        "      <td>2017-02-06</td>\n",
595 |        "      <td>1047.3</td>\n",
596 |        "    </tr>\n",
597 |        "  </tbody>\n",
598 |        "</table>\n",
599 |        "</div>"
600 |       ],
601 |       "text/plain": [
602 |        "        time   price\n",
603 |        "0 2016-02-27   427.1\n",
604 |        "1 2016-11-18   750.9\n",
605 |        "2 2016-06-13   690.9\n",
606 |        "3 2016-08-08   588.7\n",
607 |        "4 2017-02-06  1047.3"
608 |       ]
609 |      },
610 |      "execution_count": 30,
611 |      "metadata": {},
612 |      "output_type": "execute_result"
613 |     }
614 |    ],
615 |    "source": [
616 |     "from pandas import read_csv\n",
617 |     "from pandas import datetime\n",
618 |     "from pandas import DataFrame\n",
619 |     "from statsmodels.tsa.arima_model import ARIMA\n",
620 |     "from matplotlib import pyplot\n",
621 |     "def parser(x):\n",
622 |     "    return datetime.strptime(x, '%Y%m%d')\n",
623 |     "\n",
624 |     "y = read_csv('bitcoinprices.txt', header=None,names=['time','price'], parse_dates=[0], index_col=False, squeeze=True, date_parser=parser)\n",
625 |     "y.head()\n"
626 |    ]
627 |   },
628 |   {
629 |    "cell_type": "code",
630 |    "execution_count": 42,
631 |    "metadata": {},
632 |    "outputs": [],
633 |    "source": [
634 |     "#x=y.sort([\"time\"])\n",
635 |     "x = y.sort_values('time',axis=0)\n",
636 |     "x.to_csv('bitcoinsort.csv', index=False)"
637 |    ]
638 |   },
639 |   {
640 |    "cell_type": "code",
641 |    "execution_count": 45,
642 |    "metadata": {},
643 |    "outputs": [
644 |     {
645 |      "name": "stdout",
646 |      "output_type": "stream",
647 |      "text": [
648 |       "Test MSE: 32409.630\n"
649 |      ]
650 |     },
651 |     {
652 |      "data": {
653 |       "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAD8CAYAAAB+UHOxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n1ORmAAAAAElFTkSuQmCC\n",
654 |       "text/plain": [
655 |        "<matplotlib.figure.Figure at 0x1d97f4ae048>"
656 |       ]
657 |      },
658 |      "metadata": {},
659 |      "output_type": "display_data"
660 |     }
661 |    ],
662 |    "source": [
663 |     "from sklearn.metrics import mean_squared_error\n",
664 |     "error = mean_squared_error(test, predictions)\n",
665 |     "print('Test MSE: %.3f' % error)\n",
666 |     "# plot\n",
667 |     "pyplot.plot(test)\n",
668 |     "pyplot.plot(predictions, color='red')\n",
669 |     "pyplot.show()"
670 |    ]
671 |   },
672 |   {
673 |    "cell_type": "code",
674 |    "execution_count": null,
675 |    "metadata": {
676 |     "collapsed": true
677 |    },
678 |    "outputs": [],
679 |    "source": []
680 |   }
681 |  ],
682 |  "metadata": {
683 |   "kernelspec": {
684 |    "display_name": "Python 3",
685 |    "language": "python",
686 |    "name": "python3"
687 |   },
688 |   "language_info": {
689 |    "codemirror_mode": {
690 |     "name": "ipython",
691 |     "version": 3
692 |    },
693 |    "file_extension": ".py",
694 |    "mimetype": "text/x-python",
695 |    "name": "python",
696 |    "nbconvert_exporter": "python",
697 |    "pygments_lexer": "ipython3",
698 |    "version": "3.6.3"
699 |   }
700 |  },
701 |  "nbformat": 4,
702 |  "nbformat_minor": 2
703 | }
704 | 


--------------------------------------------------------------------------------
/Continuous_Stream_Data.py:
--------------------------------------------------------------------------------
 1 | 
 2 | # coding: utf-8
 3 | 
 4 | # In[1]:
 5 | 
 6 | 
 7 | import requests
 8 | import time
 9 | import datetime
10 | 
11 | f_name = input("dataset name:")
12 | f = open(f_name,"a")
13 | keys = ["price_usd","24h_volume_usd","market_cap_usd","available_supply","total_supply","percent_change_1h","percent_change_24h","percent_change_7d"]
14 | vals = [0]*len(keys)
15 | 
16 | while True:
17 |     data = requests.get("https://api.coinmarketcap.com/v1/ticker/bitcoin/").json()[0]
18 |     #bstamp = requests.get("https://www.bitstamp.net/api/v2/ticker/btcusd/").json() 
19 |     bkc = requests.get("https://blockchain.info/ticker").json()
20 |     for d in data.keys():
21 |         if d in keys:
22 |             indx = keys.index(d)
23 |             vals[indx] = data[d]
24 |     for val in vals:
25 |         f.write(val+",")
26 |       
27 |     #f.write("{},{},".format(bstamp["volume"],bstamp["vwap"]))
28 |     f.write("{},{},{}".format(bkc["USD"]["sell"],bkc["USD"]["buy"],bkc["USD"]["15m"]))
29 |     f.write(","+datetime.datetime.now().strftime("%y-%m-%d-%H-%M"))
30 |     f.write("\n")
31 |     f.flush()
32 |     time.sleep(60)
33 | 
34 | 


--------------------------------------------------------------------------------
/Continuous_Stream_Sentiment.py:
--------------------------------------------------------------------------------
 1 | # import tweepy library for twitter api access and textblob libary for sentiment analysis
 2 | import csv
 3 | import tweepy
 4 | import numpy as np
 5 | from textblob import TextBlob
 6 | import datetime
 7 | import time
 8 | 
 9 | def main():
10 | 
11 |     # set twitter api credentials
12 |     consumer_key= 'MOe623rxcqck6x8y5XhzK8MJT'
13 |     consumer_secret= 'mcBq9Km1f3OYERRD6vKmOfWSgCjsqzXAreIsn8klxAtPIo40E7'
14 |     access_token='913787859630460928-RXF8NVN3gGbxD64NCZ7wBma5M2WPwlv'
15 |     access_token_secret='P4UU9I2DimdnUown2EM6p4WZ0ftdPNsysDNW6xGh0Ts4f'
16 | 
17 |     # set path of csv file to save sentiment stats
18 |     path = 'live_tweet.csv'
19 |     f = open(path,"a")
20 |     f1 = open('tweet_data','a')
21 |     # access twitter api via tweepy methods
22 |     auth = tweepy.OAuthHandler(consumer_key, consumer_secret)
23 |     auth.set_access_token(access_token, access_token_secret)
24 |     twitter_api = tweepy.API(auth)
25 | 
26 |     while True:
27 | 
28 |         # fetch tweets by keywords
29 |         tweets = twitter_api.search(q=['bitcoin, price, crypto'], count=100)
30 | 
31 |         # get polarity
32 |         polarity = get_polarity(tweets,f1)
33 |         sentiment = np.mean(polarity)
34 | 
35 |         # save sentiment data to csv file
36 |         f.write(str(sentiment))
37 |         f.write(","+datetime.datetime.now().strftime("%y-%m-%d-%H-%M"))
38 |         f.write("\n")
39 |         f.flush()
40 |         time.sleep(60)
41 |     
42 | 
43 | def get_polarity(tweets,f):
44 |     # run polarity analysis on tweets
45 | 
46 |     tweet_polarity = []
47 | 
48 |     for tweet in tweets:
49 |         f.write(tweet.text+'\n')
50 |         analysis = TextBlob(tweet.text)
51 |         tweet_polarity.append(analysis.sentiment.polarity)
52 | 
53 |     return tweet_polarity
54 | 
55 | main()
56 | 


--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
 1 | # Overview
 2 | 
 3 | This is the code for [this](https://youtu.be/EqWm8A-dRYg) video on Youtube by Siraj Raval on Bitcoin Prediction. 
 4 | 
 5 | ##### Usage #####
 6 | The entire setup works on Google cloud.
 7 | 
 8 | ##### Part 1: Data Gathering: #####
 9 | In order to capture the real-time data, we run the following two python programs in background to continuously fetch the data. 
10 | a) Continuous_Stream_Data.py 
11 | b) Continuous_Stream_Sentiment.py
12 | 
13 | The two code do the preprocessing of data and store them in “live_tweet.csv” and “live_bitcoin.csv” files.
14 | 
15 | ##### Part 2: Core Engine: #####
16 | From the experiments we found LSTM based model to be performing better than ARIMA (discussed in detail in next section). We have set our best model parameters in engine.py file and once it is run it gathers data from the “live_bitcoin.csv” and “live_tweet.csv”, and generate features in real-time and is fed into the model.
17 | 
18 | The model outputs the next price. It also does a computation based on the threshold set in the code (this is fed from the settings file). 
19 | 
20 | The information about the time stamp, predicted price, current real price and buy/sell decision is then written into a mysql database
21 | 
22 | ##### Part 3: Tableau and Notification system: #####
23 | 
24 | We have used Tableau to generate plots in real-time form the sql-database mentioned in previous section. 
25 | 
26 | 
27 | 
28 | # Credits
29 | 
30 | Credits for this code go to [sapphirine](https://github.com/Sapphirine/BITCOIN-PRICE-PREDICTION-USING-SENTIMENT-ANALYSIS). I've merely created a wrapper to get people started. 
31 | 


--------------------------------------------------------------------------------
/bda_data_plot.PNG:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/llSourcell/bitcoin_prediction/dcba51a72f7e1a33e631703b48d9fea2a2e9b478/bda_data_plot.PNG


--------------------------------------------------------------------------------
/bitcoinprices.txt:
--------------------------------------------------------------------------------
  1 | 20160227,427.1
  2 | 20161118,750.9
  3 | 20160613,690.9
  4 | 20160808,588.7
  5 | 20170206,1047.3
  6 | 20161125,737.3
  7 | 20170724,2763.6
  8 | 20161007,619.5
  9 | 20170422,1250.0
 10 | 20171015,5661.0
 11 | 20161001,613.1
 12 | 20170305,1281.7
 13 | 20160730,636.0
 14 | 20160907,614.8
 15 | 20170123,917.9
 16 | 20170624,2535.8
 17 | 20170501,1460.0
 18 | 20170810,3473.2
 19 | 20170930,4315.0
 20 | 20160326,416.9
 21 | 20160609,577.7
 22 | 20160804,572.8
 23 | 20170226,1184.1
 24 | 20170612,2749.9
 25 | 20170728,2740.9
 26 | 20170418,1196.4
 27 | 20171019,5691.4
 28 | 20161018,636.7
 29 | 20170301,1227.4
 30 | 20160726,655.8
 31 | 20160911,608.7
 32 | 20161209,770.5
 33 | 20170529,2287.0
 34 | 20170814,4220.0
 35 | 20160309,411.5
 36 | 20160219,425.9
 37 | 20160605,583.5
 38 | 20160519,442.4
 39 | 20161024,655.1
 40 | 20170608,2727.7
 41 | 20170716,2028.4
 42 | 20161013,635.1
 43 | 20170901,4750.0
 44 | 20171023,5820.5
 45 | 20160717,681.5
 46 | 20161111,715.0
 47 | 20170313,1241.4
 48 | 20160722,659.0
 49 | 20161115,711.3
 50 | 20170525,2434.3
 51 | 20170818,4045.4
 52 | 20170526,2126.7
 53 | 20160223,421.4
 54 | 20170419,1219.0
 55 | 20160601,538.2
 56 | 20170110,915.8
 57 | 20170218,1066.4
 58 | 20170604,2584.9
 59 | 20170720,2759.7
 60 | 20160221,438.8
 61 | 20161103,688.1
 62 | 20170905,4510.5
 63 | 20171027,5768.1
 64 | 20160831,572.6
 65 | 20160428,450.3
 66 | 20170309,1192.5
 67 | 20160526,473.3
 68 | 20160718,675.6
 69 | 20160919,609.3
 70 | 20170415,1188.7
 71 | 20170521,2173.3
 72 | 20170822,4090.2
 73 | 20160301,432.8
 74 | 20160629,639.7
 75 | 20170114,824.0
 76 | 20170222,1132.2
 77 | 20161114,710.1
 78 | 20160513,457.6
 79 | 20160224,424.0
 80 | 20170909,4280.1
 81 | 20171031,6406.5
 82 | 20160130,381.6
 83 | 20160419,437.7
 84 | 20160827,571.6
 85 | 20161020,629.3
 86 | 20170321,1089.9
 87 | 20170711,2312.5
 88 | 20160923,605.1
 89 | 20170411,1228.4
 90 | 20170517,1804.0
 91 | 20170826,4356.0
 92 | 20160305,399.1
 93 | 20160625,662.2
 94 | 20170102,1018.8
 95 | 20170628,2564.8
 96 | 20161110,714.5
 97 | 20170913,3830.6
 98 | 20160210,375.9
 99 | 20160502,447.0
100 | 20160823,578.7
101 | 20161109,716.9
102 | 20170317,1058.1
103 | 20161003,610.4
104 | 20170715,1984.8
105 | 20160927,603.5
106 | 20170407,1189.5
107 | 20170830,4610.2
108 | 20170520,2073.4
109 | 20160715,669.3
110 | 20160621,651.0
111 | 20170106,846.2
112 | 20171111,6079.3
113 | 20161027,685.2
114 | 20161011,640.1
115 | 20170917,3900.0
116 | 20160214,406.8
117 | 20160819,574.5
118 | 20170211,1012.5
119 | 20170329,1045.6
120 | 20170703,2627.0
121 | 20160523,446.2
122 | 20170403,1155.3
123 | 20171004,4157.6
124 | 20160411,429.4
125 | 20160329,413.0
126 | 20160711,650.1
127 | 20160617,750.0
128 | 20170126,920.7
129 | 20170512,1699.9
130 | 20171107,7364.3
131 | 20160515,458.7
132 | 20161102,738.6
133 | 20170921,3753.2
134 | 20160202,372.9
135 | 20160815,567.5
136 | 20170215,1024.1
137 | 20170325,967.8
138 | 20170707,2498.6
139 | 20161123,738.8
140 | 20161213,780.3
141 | 20171008,4561.0
142 | 20160415,433.2
143 | 20160317,415.6
144 | 20160707,641.4
145 | 20170130,934.9
146 | 20161105,717.7
147 | 20170508,1706.0
148 | 20170616,2466.7
149 | 20160319,410.6
150 | 20170803,2779.5
151 | 20170925,3886.7
152 | 20160206,376.0
153 | 20160528,520.0
154 | 20160811,590.2
155 | 20170203,1014.4
156 | 20170727,2682.1
157 | 20160709,650.9
158 | 20170427,1358.6
159 | 20171012,5511.7
160 | 20160311,416.4
161 | 20160619,763.1
162 | 20160403,419.9
163 | 20170522,2166.9
164 | 20160703,666.1
165 | 20160904,609.3
166 | 20170118,887.9
167 | 20160527,493.2
168 | 20161210,766.9
169 | 20160410,421.8
170 | 20161120,727.4
171 | 20170807,3402.9
172 | 20170929,4244.9
173 | 20160226,426.6
174 | 20160612,680.6
175 | 20160807,591.1
176 | 20170207,1067.9
177 | 20170731,2863.0
178 | 20160517,456.5
179 | 20170423,1254.2
180 | 20171016,5637.6
181 | 20160131,375.3
182 | 20160407,422.1
183 | 20170304,1252.1
184 | 20160731,633.4
185 | 20160908,622.0
186 | 20170122,924.5
187 | 20161214,779.9
188 | 20160509,462.0
189 | 20161017,638.9
190 | 20170811,3761.8
191 | 20160310,418.8
192 | 20160608,580.2
193 | 20160803,582.8
194 | 20170227,1192.6
195 | 20170613,2800.0
196 | 20170719,2319.1
197 | 20160630,684.0
198 | 20161219,789.6
199 | 20171020,6080.0
200 | 20161022,648.3
201 | 20160727,657.0
202 | 20160912,608.7
203 | 20170528,2226.9
204 | 20170815,4061.3
205 | 20160314,414.8
206 | 20160218,417.7
207 | 20160813,586.5
208 | 20170109,904.0
209 | 20161113,705.1
210 | 20170609,2814.1
211 | 20170723,2741.0
212 | 20161026,686.8
213 | 20171024,5470.0
214 | 20170312,1233.6
215 | 20160723,656.7
216 | 20160916,608.2
217 | 20161222,861.0
218 | 20160213,397.7
219 | 20160426,470.1
220 | 20161217,790.7
221 | 20170819,4098.0
222 | 20160302,415.4
223 | 20160222,434.5
224 | 20170113,829.9
225 | 20170219,1060.4
226 | 20170605,2845.8
227 | 20160529,528.8
228 | 20170904,4337.3
229 | 20171028,5795.0
230 | 20160830,572.6
231 | 20161012,638.0
232 | 20170308,1173.4
233 | 20160719,671.9
234 | 20160920,599.3
235 | 20170412,1211.7
236 | 20161226,896.8
237 | 20161218,794.8
238 | 20160511,456.5
239 | 20170823,4164.0
240 | 20160306,404.8
241 | 20160628,637.3
242 | 20170101,997.9
243 | 20170223,1199.9
244 | 20170601,2434.9
245 | 20160611,636.3
246 | 20170908,4436.6
247 | 20160209,374.9
248 | 20161206,753.9
249 | 20160826,575.9
250 | 20160423,458.4
251 | 20170320,1047.0
252 | 20170710,2343.4
253 | 20160924,603.2
254 | 20161002,614.9
255 | 20170408,1186.2
256 | 20161230,958.7
257 | 20170827,4265.5
258 | 20171106,7186.8
259 | 20160624,683.2
260 | 20170105,1001.6
261 | 20170629,2515.7
262 | 20170516,1829.0
263 | 20171114,6565.0
264 | 20160520,444.8
265 | 20170912,4020.8
266 | 20161121,729.0
267 | 20160525,450.6
268 | 20160822,585.9
269 | 20170208,1055.4
270 | 20170518,1924.9
271 | 20160208,370.4
272 | 20170714,2106.1
273 | 20170623,2706.9
274 | 20170404,1144.0
275 | 20171001,4409.0
276 | 20160424,464.9
277 | 20160414,426.1
278 | 20170831,4713.3
279 | 20160330,414.0
280 | 20160712,669.3
281 | 20160620,704.8
282 | 20170125,901.3
283 | 20161204,757.6
284 | 20170625,2475.1
285 | 20170618,2574.0
286 | 20161215,780.4
287 | 20160818,572.9
288 | 20170212,1009.0
289 | 20170328,1015.3
290 | 20170702,2499.9
291 | 20160915,606.0
292 | 20171005,4394.7
293 | 20160505,449.0
294 | 20160512,455.8
295 | 20161104,704.0
296 | 20160318,409.5
297 | 20160708,659.1
298 | 20160616,718.8
299 | 20170129,919.1
300 | 20161028,702.3
301 | 20170515,1744.1
302 | 20170621,2560.0
303 | 20170316,1123.1
304 | 20160402,419.3
305 | 20161223,911.8
306 | 20160205,374.0
307 | 20161009,617.7
308 | 20160814,572.3
309 | 20170324,901.0
310 | 20170706,2577.0
311 | 20161004,610.0
312 | 20170428,1350.1
313 | 20171009,4827.0
314 | 20160321,415.9
315 | 20160928,605.8
316 | 20160322,417.3
317 | 20160704,678.3
318 | 20160901,570.9
319 | 20170117,904.7
320 | 20161117,740.0
321 | 20170511,1787.9
322 | 20170617,2660.0
323 | 20160530,533.8
324 | 20170924,3773.0
325 | 20161216,788.9
326 | 20160313,414.8
327 | 20160810,593.1
328 | 20170204,1030.1
329 | 20170726,2560.0
330 | 20160216,411.9
331 | 20170424,1267.0
332 | 20171013,5585.7
333 | 20160430,454.3
334 | 20161119,751.4
335 | 20161019,628.9
336 | 20170307,1188.0
337 | 20160905,606.8
338 | 20170121,939.7
339 | 20170504,1620.0
340 | 20170507,1620.7
341 | 20170804,3112.0
342 | 20170928,4051.8
343 | 20160229,433.3
344 | 20161014,639.0
345 | 20160615,699.2
346 | 20160806,593.9
347 | 20170224,1184.9
348 | 20170730,2722.1
349 | 20160917,607.5
350 | 20170420,1248.9
351 | 20171017,5372.2
352 | 20160507,462.7
353 | 20161205,755.7
354 | 20170303,1293.1
355 | 20160728,657.5
356 | 20160909,620.1
357 | 20161030,695.0
358 | 20170503,1560.2
359 | 20170808,3402.5
360 | 20170527,2219.9
361 | 20160217,418.4
362 | 20161106,717.8
363 | 20171105,7316.7
364 | 20160802,543.6
365 | 20170228,1202.7
366 | 20170614,2362.9
367 | 20170718,2348.3
368 | 20161006,613.4
369 | 20170416,1171.8
370 | 20171021,5861.7
371 | 20160408,417.6
372 | 20161126,733.9
373 | 20160930,612.3
374 | 20170315,1245.4
375 | 20160724,660.0
376 | 20160913,608.4
377 | 20171110,6704.1
378 | 20170531,2377.9
379 | 20170812,4094.5
380 | 20160315,414.2
381 | 20161023,649.9
382 | 20160607,573.8
383 | 20170108,891.6
384 | 20170216,1039.9
385 | 20170610,2909.5
386 | 20170722,2788.0
387 | 20170903,4566.7
388 | 20171025,5735.0
389 | 20161021,629.8
390 | 20170311,1189.4
391 | 20170916,3643.2
392 | 20160720,664.9
393 | 20171102,7210.1
394 | 20161208,769.2
395 | 20161221,861.7
396 | 20160201,373.3
397 | 20170816,4319.0
398 | 20160303,421.7
399 | 20160603,580.0
400 | 20170112,788.6
401 | 20170220,1093.5
402 | 20170606,2818.7
403 | 20170907,4621.7
404 | 20171029,6125.0
405 | 20160128,377.1
406 | 20161129,730.3
407 | 20160829,575.4
408 | 20161108,742.4
409 | 20170323,1048.8
410 | 20160716,669.4
411 | 20160921,597.0
412 | 20170413,1178.9
413 | 20170523,2312.1
414 | 20170820,4002.1
415 | 20161227,948.9
416 | 20170602,2458.2
417 | 20161008,617.2
418 | 20170911,4347.7
419 | 20160418,431.8
420 | 20160825,576.2
421 | 20161203,765.7
422 | 20170319,1022.4
423 | 20170709,2513.8
424 | 20160925,601.3
425 | 20170409,1209.5
426 | 20170519,1993.0
427 | 20160531,527.2
428 | 20170824,4320.0
429 | 20160327,423.5
430 | 20161116,748.2
431 | 20160627,648.6
432 | 20170104,1169.7
433 | 20161101,726.4
434 | 20170630,2491.9
435 | 20160514,460.0
436 | 20170626,2400.9
437 | 20170915,3789.9
438 | 20160212,389.0
439 | 20160422,447.6
440 | 20160821,580.9
441 | 20170209,985.3
442 | 20170331,1088.5
443 | 20170713,2337.5
444 | 20160929,606.2
445 | 20170405,1160.8
446 | 20171002,4292.0
447 | 20170828,4371.2
448 | 20160331,415.4
449 | 20160713,655.9
450 | 20160623,674.0
451 | 20170124,894.1
452 | 20171109,7195.0
453 | 20160325,416.1
454 | 20170919,3944.9
455 | 20160503,449.0
456 | 20160817,574.2
457 | 20170213,1005.3
458 | 20170327,1039.0
459 | 20170701,2415.2
460 | 20170401,1089.9
461 | 20171006,4345.0
462 | 20160522,442.9
463 | 20160409,420.8
464 | 20161225,900.1
465 | 20161127,734.8
466 | 20161130,742.4
467 | 20161229,972.1
468 | 20170128,923.7
469 | 20170514,1774.9
470 | 20170622,2712.1
471 | 20161010,618.9
472 | 20170923,3659.9
473 | 20160204,384.0
474 | 20161112,694.0
475 | 20170201,988.9
476 | 20170705,2618.2
477 | 20160516,456.5
478 | 20170429,1349.8
479 | 20171010,4764.1
480 | 20160413,425.3
481 | 20160323,415.9
482 | 20160705,671.0
483 | 20160902,575.4
484 | 20170116,892.1
485 | 20170510,1838.5
486 | 20171101,6657.3
487 | 20160508,459.8
488 | 20170801,2732.0
489 | 20170927,4197.9
490 | 20160501,456.9
491 | 20160416,433.9
492 | 20160809,584.6
493 | 20170205,1019.8
494 | 20170920,3871.9
495 | 20170725,2451.0
496 | 20170425,1290.3
497 | 20171014,5739.9
498 | 20160401,419.0
499 | 20170306,1278.5
500 | 20160701,680.6
501 | 20160906,610.6
502 | 20170120,920.4
503 | 20170506,1599.4
504 | 20160316,417.8
505 | 20170805,3207.4
506 | 20160420,440.0
507 | 20160228,434.6
508 | 20160614,683.5
509 | 20160805,579.0
510 | 20170225,1159.9
511 | 20161202,771.3
512 | 20170729,2741.0
513 | 20161124,730.2
514 | 20170421,1249.9
515 | 20171018,5572.3
516 | 20160405,422.0
517 | 20170302,1271.5
518 | 20160729,657.4
519 | 20160910,623.0
520 | 20170502,1469.0
521 | 20160427,443.3
522 | 20170809,3362.0
523 | 20160308,409.9
524 | 20160524,449.1
525 | 20161015,641.9
526 | 20160610,585.8
527 | 20160801,603.5
528 | 20161207,765.8
529 | 20170615,2455.7
530 | 20170717,2238.4
531 | 20160518,455.5
532 | 20160412,428.7
533 | 20170417,1193.0
534 | 20171022,5996.0
535 | 20170627,2487.2
536 | 20160425,468.5
537 | 20170314,1249.5
538 | 20160725,647.0
539 | 20160914,607.5
540 | 20170530,2188.0
541 | 20160510,455.6
542 | 20170813,4074.9
543 | 20160312,413.5
544 | 20160220,439.2
545 | 20160604,580.8
546 | 20160606,586.3
547 | 20170111,758.1
548 | 20170217,1069.5
549 | 20160307,414.0
550 | 20170611,2978.9
551 | 20170721,2752.1
552 | 20170902,4660.0
553 | 20171026,5895.0
554 | 20160404,422.4
555 | 20160429,455.1
556 | 20170310,1196.8
557 | 20161231,964.0
558 | 20160721,669.3
559 | 20160918,610.2
560 | 20171103,7191.0
561 | 20170414,1180.6
562 | 20161220,810.8
563 | 20170817,4269.8
564 | 20160406,421.8
565 | 20160602,536.9
566 | 20170115,836.5
567 | 20170221,1114.2
568 | 20170607,2712.0
569 | 20170906,4538.7
570 | 20171030,6135.3
571 | 20160129,380.2
572 | 20160828,574.5
573 | 20160417,433.3
574 | 20170322,1032.0
575 | 20161025,658.0
576 | 20160922,596.7
577 | 20170410,1203.6
578 | 20161224,875.1
579 | 20170821,3847.4
580 | 20160304,397.5
581 | 20170926,3933.6
582 | 20170103,1064.9
583 | 20161212,779.1
584 | 20170603,2518.4
585 | 20170910,4184.1
586 | 20160211,379.2
587 | 20160824,574.5
588 | 20160421,446.9
589 | 20170318,964.4
590 | 20170708,2555.5
591 | 20160926,606.3
592 | 20170406,1185.4
593 | 20161228,982.1
594 | 20160504,448.5
595 | 20170825,4320.5
596 | 20160324,413.8
597 | 20160714,660.9
598 | 20160626,657.9
599 | 20170107,942.2
600 | 20161029,718.6
601 | 20171112,6241.0
602 | 20170914,3167.3
603 | 20160215,400.9
604 | 20160225,423.9
605 | 20160820,583.9
606 | 20170210,1012.5
607 | 20170330,1049.7
608 | 20170712,2384.0
609 | 20161201,754.0
610 | 20161005,612.7
611 | 20170402,1153.6
612 | 20171003,4332.3
613 | 20170829,4604.0
614 | 20160328,423.5
615 | 20160710,651.4
616 | 20160622,594.9
617 | 20170127,924.0
618 | 20171113,6570.0
619 | 20170513,1799.8
620 | 20171108,7376.9
621 | 20161107,705.3
622 | 20170918,3969.2
623 | 20160203,371.4
624 | 20160816,579.1
625 | 20170214,1009.4
626 | 20170326,972.3
627 | 20161122,745.1
628 | 20170430,1374.7
629 | 20171007,4463.6
630 | 20161016,638.9
631 | 20170620,2696.8
632 | 20160706,650.1
633 | 20160618,757.3
634 | 20170131,977.9
635 | 20170505,1586.9
636 | 20170509,1727.9
637 | 20171104,7489.9
638 | 20170802,2711.6
639 | 20170922,3712.9
640 | 20160207,379.1
641 | 20160521,445.1
642 | 20160812,588.4
643 | 20170202,1015.7
644 | 20170704,2562.7
645 | 20170426,1318.1
646 | 20171011,4896.9
647 | 20160506,460.2
648 | 20161128,730.9
649 | 20160320,408.9
650 | 20160702,690.9
651 | 20160903,598.7
652 | 20170119,898.4
653 | 20161031,712.0
654 | 20161211,772.7
655 | 20170619,2624.4
656 | 20170806,3222.5
657 | 20170524,2675.0
658 | 


--------------------------------------------------------------------------------
/engine.py:
--------------------------------------------------------------------------------
  1 | from math import sqrt
  2 | from numpy import concatenate
  3 | from matplotlib import pyplot
  4 | import pandas as pd
  5 | from datetime import datetime
  6 | from sklearn.preprocessing import MinMaxScaler
  7 | from sklearn.preprocessing import LabelEncoder
  8 | from sklearn.metrics import mean_squared_error
  9 | from keras.models import Sequential
 10 | from keras.layers import Dense
 11 | from keras.layers import LSTM
 12 | import numpy as np
 13 | import datetime
 14 | from pytz import timezone
 15 | est = timezone('US/Eastern')
 16 | 
 17 | 
 18 | from twilio.rest import Client
 19 | 
 20 | # Your Account SID from twilio.com/console
 21 | account_sid = "*******"
 22 | # Your Auth Token from twilio.com/console
 23 | auth_token  = "*******"
 24 | 
 25 | client = Client(account_sid, auth_token)
 26 | 
 27 | 
 28 | 
 29 | data = pd.read_csv("merged_data.csv")
 30 | 
 31 | datag = data[['Price','Sentiment']].groupby(data['Time']).mean()
 32 | 
 33 | from sklearn.preprocessing import MinMaxScaler
 34 | values = datag['Price'].values.reshape(-1,1)
 35 | sentiment = datag['Sentiment'].values.reshape(-1,1)
 36 | values = values.astype('float32')
 37 | sentiment = sentiment.astype('float32')
 38 | scaler = MinMaxScaler(feature_range=(0, 1))
 39 | scaled = scaler.fit_transform(values)
 40 | 
 41 | train_size = int(len(scaled) * 0.7)
 42 | test_size = len(scaled) - train_size
 43 | train, test = scaled[0:train_size,:], scaled[train_size:len(scaled),:]
 44 | print(len(train), len(test))
 45 | split = train_size
 46 | 
 47 | def create_dataset(dataset, look_back, sentiment, sent=False):
 48 |     dataX, dataY = [], []
 49 |     for i in range(len(dataset) - look_back):
 50 |         if i >= look_back:
 51 |             a = dataset[i-look_back:i+1, 0]
 52 |             a = a.tolist()
 53 |             if(sent==True):
 54 |                 a.append(sentiment[i].tolist()[0])
 55 |             dataX.append(a)
 56 |             dataY.append(dataset[i + look_back, 0])
 57 |     #print(len(dataY))
 58 |     return np.array(dataX), np.array(dataY)
 59 | 
 60 | look_back = 2
 61 | trainX, trainY = create_dataset(train, look_back, sentiment[0:train_size],sent=True)
 62 | testX, testY = create_dataset(test, look_back, sentiment[train_size:len(scaled)], sent=True)
 63 | 
 64 | trainX = np.reshape(trainX, (trainX.shape[0], 1, trainX.shape[1]))
 65 | testX = np.reshape(testX, (testX.shape[0], 1, testX.shape[1]))
 66 | 
 67 | model = Sequential()
 68 | model.add(LSTM(100, input_shape=(trainX.shape[1], trainX.shape[2]), return_sequences=True))
 69 | model.add(LSTM(100))
 70 | model.add(Dense(1))
 71 | model.compile(loss='mae', optimizer='adam')
 72 | history = model.fit(trainX, trainY, epochs=300, batch_size=100, validation_data=(testX, testY), verbose=0, shuffle=False)
 73 | 
 74 | yhat = model.predict(testX)
 75 | 
 76 | yhat_inverse_sent = scaler.inverse_transform(yhat.reshape(-1, 1))
 77 | testY_inverse_sent = scaler.inverse_transform(testY.reshape(-1, 1))
 78 | 
 79 | rmse_sent = sqrt(mean_squared_error(testY_inverse_sent, yhat_inverse_sent))
 80 | print('Test RMSE: %.3f' % rmse_sent)
 81 | 
 82 | import MySQLdb
 83 | #Enter the values for you database connection
 84 | dsn_database = "bitcoin"         # e.g. "MySQLdbtest"
 85 | dsn_hostname = "173.194.231.244"      # e.g.: "mydbinstance.xyz.us-east-1.rds.amazonaws.com"
 86 | dsn_port = 3306                  # e.g. 3306 
 87 | dsn_uid = "demo"             # e.g. "user1"
 88 | dsn_pwd = "qwerty@123"              # e.g. "Password123"
 89 | 
 90 | conn = MySQLdb.connect(host=dsn_hostname, port=dsn_port, user=dsn_uid, passwd=dsn_pwd, db=dsn_database)
 91 | 
 92 | cursor=conn.cursor()
 93 | 
 94 | 
 95 | import queue 
 96 | import time
 97 | 
 98 | import queue
 99 | import matplotlib.pyplot as plt
100 | true_q = queue.Queue()
101 | pred_q = queue.Queue()
102 | '''
103 | fig = plt.figure()
104 | ax = fig.add_subplot(111)
105 | fig.show()
106 | fig.canvas.draw()
107 | plt.ion()
108 | '''
109 | 
110 | def process_data(in_data):
111 |     out_data = []
112 |     for line in in_data:
113 |         out_data.append(float(line.split(',')[0]))
114 |     return np.array(out_data).reshape(-1,1)
115 | prev = 15000
116 | threshold = 0.05
117 | while True:
118 |     btc = open('live_bitcoin.csv','r')
119 |     sent = open('live_tweet.csv','r')
120 |     bit_data = btc.readlines()
121 |     sent_data = sent.readlines()
122 |     bit_data = process_data(bit_data[len(bit_data)-5:])
123 |     sent_data = process_data(sent_data[len(sent_data)-5:])
124 |     live = scaler.transform(bit_data)
125 |     testX, testY = create_dataset(live, 2, sent_data, sent=True)
126 |     testX = np.reshape(testX, (testX.shape[0], 1, testX.shape[1]))
127 |     yhat = model.predict(testX)
128 |     yhat_inverse = scaler.inverse_transform(yhat.reshape(-1, 1))
129 |     true_q.put(bit_data[4])
130 |     pred_q.put(yhat_inverse[0])
131 |     val = 100*((yhat_inverse[0][0] - prev)/prev)
132 |     if val > threshold:
133 |         decision = 'Buy!!!'
134 |         message = client.messages.create(to="+15184234418‬", from_="+15188883052", body=decision+' - Price of Bitcoin is expected to rise.')
135 |     elif val <-threshold:
136 |         decision = 'Sell!!!'
137 |         message = client.messages.create(to="+15184234418", from_="+15188883052", body=decision+' - Price of Bitcoin is expected to drop.')
138 |     else:
139 |         decision = ''
140 |     prev = yhat_inverse[0][0]
141 |     input_string = "INSERT INTO live_data values ({},{},{},'{}','{}');".format(yhat_inverse[0][0],bit_data[0][0],sent_data[4][0],datetime.datetime.now(tz=est).strftime('%Y-%m-%d %H:%M:%S'),decision)
142 |     cursor.execute(input_string)
143 |     conn.commit()
144 |     time.sleep(60)
145 |     
146 |     
147 | 
148 | 
149 | 
150 | 
151 | 


--------------------------------------------------------------------------------
/future_virtual_trading.py:
--------------------------------------------------------------------------------
  1 | from math import sqrt
  2 | from numpy import concatenate
  3 | from matplotlib import pyplot
  4 | import pandas as pd
  5 | from datetime import datetime
  6 | from sklearn.preprocessing import MinMaxScaler
  7 | from sklearn.preprocessing import LabelEncoder
  8 | from sklearn.metrics import mean_squared_error
  9 | from keras.models import Sequential
 10 | from keras.layers import Dense
 11 | from keras.layers import LSTM
 12 | import numpy as np
 13 | import datetime
 14 | from pytz import timezone
 15 | est = timezone('US/Eastern')
 16 | 
 17 | 
 18 | from twilio.rest import Client
 19 | 
 20 | # Your Account SID from twilio.com/console
 21 | account_sid = "AC3899557667afcb2bbc20e9ebecf8befb"
 22 | # Your Auth Token from twilio.com/console
 23 | auth_token  = "18c9c35bafb8abb1c009e9aad25a0aec"
 24 | 
 25 | client = Client(account_sid, auth_token)
 26 | 
 27 | 
 28 | 
 29 | data = pd.read_csv("merged_data.csv")
 30 | 
 31 | datag = data[['Price','Sentiment']].groupby(data['Time']).mean()
 32 | 
 33 | from sklearn.preprocessing import MinMaxScaler
 34 | values = datag['Price'].values.reshape(-1,1)
 35 | sentiment = datag['Sentiment'].values.reshape(-1,1)
 36 | values = values.astype('float32')
 37 | sentiment = sentiment.astype('float32')
 38 | scaler = MinMaxScaler(feature_range=(0, 1))
 39 | scaled = scaler.fit_transform(values)
 40 | 
 41 | train_size = int(len(scaled) * 0.7)
 42 | test_size = len(scaled) - train_size
 43 | train, test = scaled[0:train_size,:], scaled[train_size:len(scaled),:]
 44 | print(len(train), len(test))
 45 | split = train_size
 46 | 
 47 | def create_dataset(dataset, look_back, sentiment, sent=False):
 48 |     dataX, dataY = [], []
 49 |     for i in range(len(dataset) - look_back):
 50 |         if i >= look_back:
 51 |             a = dataset[i-look_back:i+1, 0]
 52 |             a = a.tolist()
 53 |             if(sent==True):
 54 |                 a.append(sentiment[i].tolist()[0])
 55 |             dataX.append(a)
 56 |             dataY.append(dataset[i + look_back, 0])
 57 |     #print(len(dataY))
 58 |     return np.array(dataX), np.array(dataY)
 59 | 
 60 | look_back = 2
 61 | trainX, trainY = create_dataset(train, look_back, sentiment[0:train_size],sent=True)
 62 | testX, testY = create_dataset(test, look_back, sentiment[train_size:len(scaled)], sent=True)
 63 | 
 64 | trainX = np.reshape(trainX, (trainX.shape[0], 1, trainX.shape[1]))
 65 | testX = np.reshape(testX, (testX.shape[0], 1, testX.shape[1]))
 66 | 
 67 | model = Sequential()
 68 | model.add(LSTM(100, input_shape=(trainX.shape[1], trainX.shape[2]), return_sequences=True))
 69 | model.add(LSTM(100))
 70 | model.add(Dense(1))
 71 | model.compile(loss='mae', optimizer='adam')
 72 | history = model.fit(trainX, trainY, epochs=300, batch_size=100, validation_data=(testX, testY), verbose=0, shuffle=False)
 73 | 
 74 | yhat = model.predict(testX)
 75 | 
 76 | yhat_inverse_sent = scaler.inverse_transform(yhat.reshape(-1, 1))
 77 | testY_inverse_sent = scaler.inverse_transform(testY.reshape(-1, 1))
 78 | 
 79 | rmse_sent = sqrt(mean_squared_error(testY_inverse_sent, yhat_inverse_sent))
 80 | print('Test RMSE: %.3f' % rmse_sent)
 81 | 
 82 | import MySQLdb
 83 | #Enter the values for you database connection
 84 | dsn_database = "bitcoin"         # e.g. "MySQLdbtest"
 85 | dsn_hostname = "173.194.231.244"      # e.g.: "mydbinstance.xyz.us-east-1.rds.amazonaws.com"
 86 | dsn_port = 3306                  # e.g. 3306 
 87 | dsn_uid = "demo"             # e.g. "user1"
 88 | dsn_pwd = "qwerty@123"              # e.g. "Password123"
 89 | 
 90 | conn = MySQLdb.connect(host=dsn_hostname, port=dsn_port, user=dsn_uid, passwd=dsn_pwd, db=dsn_database)
 91 | 
 92 | cursor=conn.cursor()
 93 | 
 94 | 
 95 | import queue 
 96 | import time
 97 | 
 98 | import queue
 99 | import matplotlib.pyplot as plt
100 | true_q = queue.Queue()
101 | pred_q = queue.Queue()
102 | '''
103 | fig = plt.figure()
104 | ax = fig.add_subplot(111)
105 | fig.show()
106 | fig.canvas.draw()
107 | plt.ion()
108 | '''
109 | 
110 | def process_data(in_data):
111 |     out_data = []
112 |     for line in in_data:
113 |         out_data.append(float(line.split(',')[0]))
114 |     return np.array(out_data).reshape(-1,1)
115 | prev = 15000
116 | threshold = 0.05
117 | bank = 15000
118 | bitcoin = 0
119 | while True:
120 |     btc = open('live_bitcoin.csv','r')
121 |     sent = open('live_tweet.csv','r')
122 |     bit_data = btc.readlines()
123 |     sent_data = sent.readlines()
124 |     bit_data = process_data(bit_data[len(bit_data)-5:])
125 |     sent_data = process_data(sent_data[len(sent_data)-5:])
126 |     live = scaler.transform(bit_data)
127 |     testX, testY = create_dataset(live, 2, sent_data, sent=True)
128 |     testX = np.reshape(testX, (testX.shape[0], 1, testX.shape[1]))
129 |     yhat = model.predict(testX)
130 |     yhat_inverse = scaler.inverse_transform(yhat.reshape(-1, 1))
131 |     true_q.put(bit_data[4])
132 |     pred_q.put(yhat_inverse[0])
133 |     val = 100*((yhat_inverse[0][0] - prev)/prev)
134 |     if val > threshold:
135 |         decision = 'Buy!!!'
136 |         #message = client.messages.create(to="+15184234418‬", from_="+15188883052", body=decision+' - Price of Bitcoin is expected to rise.')
137 |     elif val <-threshold:
138 |         decision = 'Sell!!!'
139 |         #message = client.messages.create(to="+15184234418", from_="+15188883052", body=decision+' - Price of Bitcoin is expected to drop.')
140 |     else:
141 |         decision = ''
142 |     prev = yhat_inverse[0][0]
143 |     #input_string = "INSERT INTO live_data values ({},{},{},'{}','{}');".format(yhat_inverse[0][0],bit_data[0][0],sent_data[4][0],datetime.datetime.now(tz=est).strftime('%Y-%m-%d %H:%M:%S'),decision)
144 |     #cursor.execute(input_string)
145 |     #conn.commit()
146 |     time.sleep(60)
147 |     
148 |     
149 | 
150 | 
151 | 
152 | 
153 | 


--------------------------------------------------------------------------------
/get_bitcoin_new.py:
--------------------------------------------------------------------------------
 1 | import requests, json
 2 | from time import sleep
 3 | from datetime import datetime
 4 | import sys 
 5 | import traceback
 6 | import argparse
 7 | 
 8 | parser = argparse.ArgumentParser()
 9 | 
10 | parser.add_argument("outputfile", nargs='?', default="bitcoin_price.json")
11 | parser.add_argument("errorfile", nargs='?', default="bitcoin_price_error.txt")
12 | args = parser.parse_args()
13 | 
14 | def getBitcoinPrice():
15 |     URL = 'https://www.bitstamp.net/api/ticker/'
16 |     try:
17 |         r = requests.get(URL)
18 |         bitcoindata = json.loads(r.text)
19 |         bitcoindata['datetime'] = datetime.utcfromtimestamp(int(bitcoindata['timestamp'])).strftime('%Y-%m-%d-%H-%M-%S')        
20 | 
21 |         with open(args.outputfile, mode='a') as file:
22 |             file.write('{},\n'.format(json.dumps(bitcoindata)))     
23 | 
24 |     except:        
25 |         exc_type, exc_value, exc_traceback = sys.exc_info()
26 |         with open(args.errorfile, mode='a') as file:           
27 |            traceback.print_exc(file=file)
28 |            file.write(('-'*100)+'\n\n')
29 | 
30 | 
31 | while True:	
32 |     getBitcoinPrice()
33 |     sleep(10)


--------------------------------------------------------------------------------
/get_bitcoin_price.py:
--------------------------------------------------------------------------------
 1 | __author__ = 'Chandra S Narain Kappera'
 2 | 
 3 | import requests
 4 | 
 5 | def getPrices():
 6 |     date_prices = {}
 7 |     for page in range(1,1000):
 8 |         page_string = str(page)
 9 |         page_link = 'https://api.coinbase.com/v1/prices/historical?page=' + page_string
10 |         r = requests.get(page_link)
11 |         r = r.text
12 |         year_string = r[0:4]
13 |         month_string = r[5:7]
14 |         day_string = r[8:10]
15 |         year_int = int(year_string)
16 |         month_int = int(month_string)
17 |         day_int = int(day_string)
18 |         date_tuple = (year_int, month_int, day_int)
19 |         if date_tuple not in date_prices:
20 |             date_string = year_string + "-" + month_string + "-" + day_string
21 |             start_point = r.index(date_string) + 26
22 |             end_point = start_point
23 |             while r[end_point] != '.':
24 |                 end_point = end_point + 1
25 |             end_point = end_point + 2
26 |             value_string = r[start_point:end_point]
27 |             value_float = float(value_string)
28 |             date_prices[date_tuple] = value_float
29 |         day_on_page = True
30 |         while day_on_page:
31 |             if month_int == 3 and day_int == 1:
32 |                 month_int = 2
33 |                 day_int = 28
34 |             elif (month_int == 5 or month_int == 7 or month_int == 10 or month_int == 12) and day_int == 1:
35 |                 month_int = month_int - 1
36 |                 day_int = 30
37 |             elif (month_int == 2 or month_int == 4 or month_int == 6 or month_int == 8 or month_int == 9 or month_int == 11) and day_int == 1:
38 |                 month_int = month_int - 1
39 |                 day_int = 31
40 |             elif month_int == 1 and day_int == 1:
41 |                 month_int = 12
42 |                 day_int = 31
43 |                 year_int = year_int - 1
44 |             else:
45 |                 day_int = day_int - 1
46 |             month_string = str(month_int)
47 |             if len(month_string) < 2:
48 |                 month_string = '0' + month_string
49 |             day_string = str(day_int)
50 |             if len(day_string) < 2:
51 |                 day_string = '0' + day_string
52 |             date_string = str(year_int) + '-' + month_string + '-' + day_string
53 |             if r.find(date_string) == -1:
54 |                 day_on_page = False
55 |                 break
56 |             start_point = r.index(date_string) + 26
57 |             end_point = start_point
58 |             while r[end_point] != '.':
59 |                 end_point = end_point + 1
60 |             end_point = end_point + 2
61 |             value_string = r[start_point:end_point]
62 |             value_float = float(value_string)
63 |             date_tuple = (year_int, month_int, day_int)
64 |             date_prices[date_tuple] = value_float
65 |     return date_prices
66 | 
67 | 
68 |         #figure out first day in list
69 |         #if we don't have data for that day
70 |             #get data
71 |         #otherwise
72 |             #find next day's data
73 |             #add day to days list
74 |             #if we can't find next day's data
75 |                 #next page
76 | 
77 | 
78 | target = open('bitcoinprices.txt', 'w')
79 | 
80 | price_dict = getPrices()
81 | # not very elegant - need to modify this, but commiting the current logic - Chandra
82 | for key in price_dict.keys():
83 |     year = key[0]
84 |     month = key[1]
85 |     date = key[2]
86 |     year = year*10000
87 |     month = month*100
88 |     stamp = year+month+date
89 |     value = price_dict[key]
90 |     target.write(str(stamp)+','+str(value)+'\n')
91 | 


--------------------------------------------------------------------------------
/get_twitter.py:
--------------------------------------------------------------------------------
 1 | __author__ = 'Chandra S Narain Kappera'
 2 | 
 3 | 
 4 | from pprint import pprint
 5 | import requests, json
 6 | import selenium
 7 | from selenium import webdriver
 8 | from selenium.webdriver.common.by import By
 9 | from selenium.webdriver.support.ui import WebDriverWait
10 | from selenium.webdriver.support import expected_conditions as EC
11 | import time
12 | import datetime
13 | 
14 | 
15 | #Format
16 | #{"importance": 0-10, "sentiment": positive/negative/neutral, "score":0-1}
17 | #
18 | 
19 | def get_sentiment(year, month, day):
20 |     month_string = str(month)
21 |     if len(month_string) < 2:
22 |         month_string = "0" + month_string
23 |     day_string = str(day)
24 |     if len(day_string) < 2:
25 |         day_string = "0" + day_string
26 |     date = str(year) + month_string + day_string
27 | 
28 |     r1 = requests.get("http://archive.org/wayback/available?url=reddit.com/r/bitcoin&timestamp=" + date)
29 | 
30 |     if(r1.status_code == 200):
31 |         data1 = json.loads(r1.text)
32 |         archive_url = data1['archived_snapshots']['closest']['url']
33 |     else:
34 |         archive_url = None
35 |         print("Error return code = "+str(r1.status_code))
36 | 
37 |     r2 = requests.get("https://api.idolondemand.com/1/api/sync/analyzesentiment/v1?apikey=fe6dea49-084f-4cd8-be86-0976baf9a714&url=" + archive_url)
38 | 
39 |     if(r2.status_code == 200):
40 |         data2 = json.loads(r2.text)
41 |         return data2['aggregate']['score']
42 |     else:
43 |         print("Error return code = "+str(r2.status_code))
44 | 
45 | 
46 | def make_dict():
47 |     scores = {}
48 |     date = datetime.date(2017, 11, 14)
49 |     target = open('sentiment6.txt', 'w')
50 |     for i in range(643):
51 |         #print(date.year)
52 |         #print(date.month)
53 |         #print(date.day)
54 |         stamp = date.year*10000+date.month*100+date.day
55 |         value = get_sentiment(date.year, date.month, date.day)
56 |         scores[(date.year, date.month, date.day)] = value
57 |         date -= datetime.timedelta(days=1)
58 |         target.write(str(stamp)+','+str(value))
59 |         target.write('\n')
60 |     target.close()
61 | make_dict()
62 | 


--------------------------------------------------------------------------------
/get_twitter_new.py:
--------------------------------------------------------------------------------
  1 | #Import the necessary methods from tweepy library
  2 | from tweepy.streaming import StreamListener
  3 | from tweepy import OAuthHandler
  4 | from tweepy import Stream
  5 | from textblob import TextBlob
  6 | import re 
  7 | import json
  8 | import datetime
  9 | import traceback
 10 | import sys 
 11 | 
 12 | import argparse
 13 | 
 14 | parser = argparse.ArgumentParser()
 15 | 
 16 | parser.add_argument("outputfile", nargs='?', default="tweets_bitcoin.json")
 17 | parser.add_argument("errorfile", nargs='?', default="tweets_bitcoin_error.txt")
 18 | 
 19 | 
 20 | args = parser.parse_args()
 21 | 
 22 | #Variables that contains the user credentials to access Twitter API 
 23 | access_token = "2569501620-dOSDk5kftHZiOTlWDzu66k9S7xKEh2GrpjNn1SA"
 24 | access_token_secret = "makKH9EdIErBW8JhQdHdc594qWyH09VoQkZm4FFCAHCQk"
 25 | consumer_key = "VZ4evnFfxQ0f1vyRpyoJrf5VT"
 26 | consumer_secret = "dcSDKuSz0ODD8kqPgBdBX6u6l0JoGq8NDbGQ2Vvg5bukDo3weh"
 27 | 
 28 | 
 29 | class Tweet:
 30 |     def __init__(self, text, sentiment_text, polarity, created_at):
 31 |         self.text = text
 32 |         self.sentiment_text = sentiment_text
 33 |         self.polarity = polarity
 34 |         self.created_at = created_at
 35 | 
 36 |     def __str__(self):
 37 |         return 'Sentiment: {} {} \nText: {}\nCreated_At: {}\n'.format(self.sentiment_text, self.polarity, self.text, self.created_at)
 38 | 
 39 | #This is a basic listener that just prints received tweets to stdout.
 40 | class StdOutListener(StreamListener):
 41 |     def __init__(self):
 42 |         self.mylist = []
 43 |     def on_data(self, data):
 44 |                 
 45 |         tweet = json.loads(data)
 46 |         
 47 |         try:
 48 |             if 'retweeted_status' in tweet:
 49 |             # Retweets 
 50 |                 if 'extended_tweet' in tweet['retweeted_status']:
 51 |                 #When extended beyond 140 Characters limit
 52 |                     tweet_text = tweet['retweeted_status']['extended_tweet']['full_text']
 53 |                 else:
 54 |                     tweet_text = tweet['retweeted_status']['text']
 55 |             else:
 56 |             #Normal Tweets
 57 |                 if 'extended_tweet' in tweet:
 58 |                 #When extended beyond 140 Characters limit            
 59 |                     tweet_text = tweet['extended_tweet']['full_text']
 60 |                 else:
 61 |                     tweet_text = tweet['text']
 62 |             
 63 |             tweet_sentiment, polarity = self.get_tweet_sentiment(tweet_text)        
 64 |             
 65 |             newTweetObj = Tweet(text=tweet_text, 
 66 |                                 sentiment_text=tweet_sentiment, 
 67 |                                 polarity=polarity, 
 68 |                                 created_at= datetime.datetime.strptime(tweet['created_at'],'%a %b %d %H:%M:%S %z %Y').strftime('%Y-%m-%d-%H-%M-%S'))
 69 |             
 70 |             with open(args.outputfile, mode='a') as file:
 71 |                 file.write('{},\n'.format(json.dumps(newTweetObj.__dict__)))     
 72 | 
 73 |         except:
 74 |             exc_type, exc_value, exc_traceback = sys.exc_info()            
 75 |             with open(args.errorfile, mode='a') as file:
 76 |                 file.write('Error: {}\n\nTweet Info: {}\n--------------------------\n'.format(repr(traceback.format_tb(exc_traceback)), tweet))            
 77 | 
 78 |         return True
 79 | 
 80 |     def on_error(self, status):
 81 |         print(status)
 82 |         print('-----------------')
 83 | 
 84 |     def clean_tweet(self, tweet):
 85 |         '''
 86 |         Utility function to clean tweet text by removing links, special characters
 87 |         using simple regex statements.
 88 |         '''
 89 |         return ' '.join(re.sub("(@[A-Za-z0-9]+)|([^0-9A-Za-z \t])|(\w+:\/\/\S+)", " ", tweet).split())
 90 | 
 91 |     def get_tweet_sentiment(self, tweet):
 92 |         '''
 93 |         Utility function to classify sentiment of passed tweet
 94 |         using textblob's sentiment method
 95 |         '''
 96 |         # create TextBlob object of passed tweet text
 97 |         analysis = TextBlob(self.clean_tweet(tweet))
 98 | 
 99 |         # set sentiment
100 |         sentiment = None 
101 |         if analysis.sentiment.polarity > 0:
102 |             sentiment = 'positive'
103 |         elif analysis.sentiment.polarity == 0:
104 |             sentiment = 'neutral'
105 |         else:
106 |             sentiment = 'negative'
107 | 
108 |         return sentiment, analysis.sentiment.polarity
109 | 
110 |     def convert_sentiment_to_emoticon(self, sentiment_text):
111 |         if sentiment_text == 'positive':
112 |             return '✅'
113 |         elif sentiment_text == 'negative':
114 |             return '❌'
115 |         else:
116 |             return '🤷'
117 | 
118 | if __name__ == '__main__':
119 | 
120 |     #This handles Twitter authetification and the connection to Twitter Streaming API
121 |     l = StdOutListener()
122 |     auth = OAuthHandler(consumer_key, consumer_secret)
123 |     auth.set_access_token(access_token, access_token_secret)
124 |     stream = Stream(auth, l, tweet_mode='extended')
125 | 
126 |     #This line filter Twitter Streams to capture data by the keywords: 'python', 'javascript', 'ruby'
127 |     stream.filter(track=['bitcoin', 'BTC'])


--------------------------------------------------------------------------------
/sentiment6.txt:
--------------------------------------------------------------------------------
  1 | 20171114,0.38021943904690936
  2 | 20171113,0.38021943904690936
  3 | 20171112,0.38021943904690936
  4 | 20171111,0.38021943904690936
  5 | 20171110,0.38021943904690936
  6 | 20171109,0.38021943904690936
  7 | 20171108,0.38021943904690936
  8 | 20171107,0.38021943904690936
  9 | 20171106,0.38021943904690936
 10 | 20171105,0.38021943904690936
 11 | 20171104,0.38021943904690936
 12 | 20171103,0.38021943904690936
 13 | 20171102,0.38021943904690936
 14 | 20171101,0.38021943904690936
 15 | 20171031,0.38021943904690936
 16 | 20171030,0.38021943904690936
 17 | 20171029,0.38021943904690936
 18 | 20171028,0.38021943904690936
 19 | 20171027,0.38021943904690936
 20 | 20171026,0.38021943904690936
 21 | 20171025,0.38021943904690936
 22 | 20171024,0.38021943904690936
 23 | 20171023,0.38021943904690936
 24 | 20171022,0.38021943904690936
 25 | 20171021,0.38021943904690936
 26 | 20171020,0.38021943904690936
 27 | 20171019,0.38021943904690936
 28 | 20171018,0.38021943904690936
 29 | 20171017,0.38021943904690936
 30 | 20171016,0.38021943904690936
 31 | 20171015,0.38021943904690936
 32 | 20171014,0.38021943904690936
 33 | 20171013,0.38021943904690936
 34 | 20171012,0.38021943904690936
 35 | 20171011,0.38021943904690936
 36 | 20171010,0.38021943904690936
 37 | 20171009,0.38021943904690936
 38 | 20171008,0.38021943904690936
 39 | 20171007,0.38021943904690936
 40 | 20171006,0.38021943904690936
 41 | 20171005,0.38021943904690936
 42 | 20171004,0.38021943904690936
 43 | 20171003,0.38021943904690936
 44 | 20171002,0.2505911694572912
 45 | 20171001,0.2883043981056517
 46 | 20170930,0.22871109056917982
 47 | 20170929,0.3806278258579074
 48 | 20170928,0.24809526009512026
 49 | 20170927,0.2625429079708684
 50 | 20170926,0.2625429079708684
 51 | 20170925,0.347151782235405
 52 | 20170924,0.3069042819449387
 53 | 20170923,0.22740568414417128
 54 | 20170922,0.07673560909897316
 55 | 20170921,0.318641544217865
 56 | 20170920,0.318641544217865
 57 | 20170919,0.1980971309285184
 58 | 20170918,0.3049536211823958
 59 | 20170917,0.292749113295036
 60 | 20170916,0.2596776886772127
 61 | 20170915,0.25284511111632846
 62 | 20170914,0.09739457147044052
 63 | 20170913,0.16079828706521007
 64 | 20170912,0.25591845918058764
 65 | 20170911,0.3467723630228228
 66 | 20170910,0.2917620893733621
 67 | 20170909,0.3370602849342434
 68 | 20170908,0.11018949237374867
 69 | 20170907,0.24013439101853573
 70 | 20170906,0.17587381606832045
 71 | 20170905,0.2585821529543866
 72 | 20170904,0.3363159213279587
 73 | 20170903,0.3452091436650593
 74 | 20170902,0.410246838560282
 75 | 20170901,0.34961641313783115
 76 | 20170831,0.371981770591173
 77 | 20170830,0.25394288550817595
 78 | 20170829,0.21874676944921964
 79 | 20170828,0.21874676944921964
 80 | 20170827,0.3598057072808753
 81 | 20170826,0.26887507826993684
 82 | 20170825,0.22403883946656988
 83 | 20170824,0.16035248567793672
 84 | 20170823,0.18468143537746343
 85 | 20170822,0.3523245671300772
 86 | 20170821,0.3151493556819943
 87 | 20170820,0.10876035749010954
 88 | 20170819,0.27383642364593547
 89 | 20170818,0.2287504066057202
 90 | 20170817,0.29796878488148787
 91 | 20170816,0.152765307186486
 92 | 20170815,0.15091089305019034
 93 | 20170814,0.30374067910015107
 94 | 20170813,0.37456593042579167
 95 | 20170812,0.34140746243192505
 96 | 20170811,0.3758329745600802
 97 | 20170810,0.243127877976264
 98 | 20170809,0.16266069617076145
 99 | 20170808,0.3751261826692668
100 | 20170807,0.3620419192525838
101 | 20170806,0.4232222785640444
102 | 20170805,0.03370941542788476
103 | 20170804,0.2959419357690376
104 | 20170803,0.43510436996532165
105 | 20170802,0.5045077403204518
106 | 20170801,0.38579380415996084
107 | 20170731,0.20538887673196699
108 | 20170730,0.20538887673196699
109 | 20170729,0.21995154183116494
110 | 20170728,0.2540792928216312
111 | 20170727,0.2800075998644351
112 | 20170726,0.29979533674737996
113 | 20170725,0.1149573421446629
114 | 20170724,0.5002746636988847
115 | 20170723,0.43108252354295723
116 | 20170722,0.2777090506131157
117 | 20170721,0.4854968579086719
118 | 20170720,0.3376769740993703
119 | 20170719,0.25877486715415804
120 | 20170718,0.39237710237276197
121 | 20170717,0.4685222099196068
122 | 20170716,0.3360522452634446
123 | 20170715,0.12597394868986372
124 | 20170714,0.4542367943161525
125 | 20170713,0.15406235227181583
126 | 20170712,0.29094835438614375
127 | 20170711,0.38862384789905235
128 | 20170710,0.510614340603299
129 | 20170709,0.510614340603299
130 | 20170708,0.21502846353920602
131 | 20170707,0.41530443718718674
132 | 20170706,0.3480927641863872
133 | 20170705,0.18694484779031673
134 | 20170704,0.09310500971278221
135 | 20170703,0.0669665313504003
136 | 20170702,0.18882317328969558
137 | 20170701,0.29059122537644677
138 | 20170630,0.2683340719507728
139 | 20170629,0.2703343362665517
140 | 20170628,0.11723817518397327
141 | 20170627,0.17665150748971625
142 | 20170626,0.27715151044653025
143 | 20170625,0.21685749726903347
144 | 20170624,0.31850331612785815
145 | 20170623,0.30021087154094567
146 | 20170622,0.3063278445266835
147 | 20170621,0.18945601707829474
148 | 20170620,0.15540824256029975
149 | 20170619,0.35493137474853276
150 | 20170618,0.3774529849124963
151 | 20170617,0.1677529919427456
152 | 20170616,0.1677529919427456
153 | 20170615,0.24729676184854765
154 | 20170614,0.1485524891552037
155 | 20170613,-0.040371922823996126
156 | 20170612,0.2635892723228725
157 | 20170611,0.3833741890425256
158 | 20170610,0.383941440827027
159 | 20170609,0.1026646062635423
160 | 20170608,0.1778802234480251
161 | 20170607,0.37726742269057745
162 | 20170606,0.17640603920465855
163 | 20170605,0.308088085615714
164 | 20170604,0.22873950096075696
165 | 20170603,0.31959951414354215
166 | 20170602,0.24101319963498843
167 | 20170601,0.17012411429790877
168 | 20170531,0.07014599143180046
169 | 20170530,0.2120864947064709
170 | 20170529,0.2120864947064709
171 | 20170528,0.24970388274243369
172 | 20170527,0.21238023410492496
173 | 20170526,0.4586725585380068
174 | 20170525,0.3496578489114199
175 | 20170524,0.2657564279634523
176 | 20170523,0.4040768133626713
177 | 20170522,0.17056525325442085
178 | 20170521,0.3477333770707802
179 | 20170520,0.3871956140364719
180 | 20170519,0.27137216552019106
181 | 20170518,0.16802333077423814
182 | 20170517,0.3408091753367907
183 | 20170516,0.09615415156119812
184 | 20170515,0.3166774104506486
185 | 20170514,0.3166774104506486
186 | 20170513,0.19639805699271165
187 | 20170512,0.44894053321408467
188 | 20170511,0.3571365450558292
189 | 20170510,0.09300275107338975
190 | 20170509,0.4622597164277392
191 | 20170508,0.48518585161221944
192 | 20170507,0.26710718852102594
193 | 20170506,0.3301874089233446
194 | 20170505,0.4239203443238371
195 | 20170504,0.19717928195127438
196 | 20170503,0.3043593035203883
197 | 20170502,0.258361727843722
198 | 20170501,0.1548289066240211
199 | 20170430,0.12990744579958433
200 | 20170429,0.22800852108652264
201 | 20170428,0.2488288819297634
202 | 20170427,0.11637409389031098
203 | 20170426,0.21643355160391808
204 | 20170425,0.23901394017001873
205 | 20170424,0.3199119640299874
206 | 20170423,0.3832959983080909
207 | 20170422,0.4756622159477592
208 | 20170421,0.3326870202884664
209 | 20170420,0.113556894177197
210 | 20170419,0.08850622377644807
211 | 20170418,0.31039919383459413
212 | 20170417,0.32015354816417074
213 | 20170416,0.2668320514746584
214 | 20170415,0.3146846905800985
215 | 20170414,0.18776066278887835
216 | 20170413,0.3861244497759137
217 | 20170412,0.3245588418385859
218 | 20170411,0.2706404838308102
219 | 20170410,0.05208776327594738
220 | 20170409,0.14081391368435534
221 | 20170408,0.018005442837751316
222 | 20170407,0.10197464411414206
223 | 20170406,0.21070069051355553
224 | 20170405,0.31689766946821524
225 | 20170404,0.3130415263684466
226 | 20170403,0.2937196950891288
227 | 20170402,0.2020168499160209
228 | 20170401,0.347791460207737
229 | 20170331,0.1933171620404274
230 | 20170330,0.1280023859962342
231 | 20170329,0.20834956976157623
232 | 20170328,0.08202415035239306
233 | 20170327,0.19121891274857855
234 | 20170326,0.33775771723697706
235 | 20170325,0.16654225656105356
236 | 20170324,0.3380867839375403
237 | 20170323,0.02211859305684538
238 | 20170322,0.2976134405851208
239 | 20170321,0.14764289040277032
240 | 20170320,0.25813452009376425
241 | 20170319,0.33426197666399865
242 | 20170318,0.17964069971106156
243 | 20170317,0.38572162832908496
244 | 20170316,0.16859617964154316
245 | 20170315,0.34317467076191693
246 | 20170314,0.20229882288975878
247 | 20170313,0.09448256062329684
248 | 20170312,0.18984718710618875
249 | 20170311,0.3022397051171501
250 | 20170310,0.2936294101724601
251 | 20170309,0.3733741706581603
252 | 20170308,0.25024096164196835
253 | 20170307,0.3941018610733407
254 | 20170306,0.3542635100549983
255 | 20170305,0.2789727409736776
256 | 20170304,0.3314061861830554
257 | 20170303,0.3164806095850169
258 | 20170302,0.28878392460718255
259 | 20170301,0.40588685047251355
260 | 20170228,0.28341045287698624
261 | 20170227,0.3367314636743158
262 | 20170226,0.22406379250486005
263 | 20170225,0.30071073432284867
264 | 20170224,0.3099757475965949
265 | 20170223,0.32129743094466207
266 | 20170222,0.34275875923163224
267 | 20170221,0.30195474875529793
268 | 20170220,0.180401339140928
269 | 20170219,0.31327113422582126
270 | 20170218,0.2723475484402832
271 | 20170217,0.3801157651513554
272 | 20170216,0.24271048363100944
273 | 20170215,0.2244451215373433
274 | 20170214,0.3924237636474582
275 | 20170213,0.4404530729518508
276 | 20170212,0.2884606042943219
277 | 20170211,0.36334298682595995
278 | 20170210,0.31825489902703885
279 | 20170209,0.30814201443544653
280 | 20170208,0.3363796866351565
281 | 20170207,0.246464174660013
282 | 20170206,0.28548048700214934
283 | 20170205,0.37650587135258096
284 | 20170204,0.3380776867872938
285 | 20170203,0
286 | 20170202,0
287 | 20170201,0.39827594800090765
288 | 20170131,0.07155334251923583
289 | 20170130,0.22133041873843623
290 | 20170129,0.28767802714622037
291 | 20170128,0.38002948093755384
292 | 20170127,0.288970847040402
293 | 20170126,0.39624662982350156
294 | 20170125,0.24827846182889104
295 | 20170124,0.18583339211594546
296 | 20170123,0.34650153809464296
297 | 20170122,0.4650471322880682
298 | 20170121,0.4549192867376071
299 | 20170120,0.2981663638620774
300 | 20170119,0.22691515539686366
301 | 20170118,0.44305754264078745
302 | 20170117,0.30475499376618664
303 | 20170116,0.34536157379325333
304 | 20170115,0.29702551851818665
305 | 20170114,0.30380038488592304
306 | 20170113,0.25818400423533855
307 | 20170112,0.3930190926850269
308 | 20170111,0.27800254733237345
309 | 20170110,0.39870320162901374
310 | 20170109,0.4312088404650657
311 | 20170108,0.32130783266007645
312 | 20170107,0.3048873987910217
313 | 20170106,0.4691977309069293
314 | 20170105,0.4691977309069293
315 | 20170104,0.4691977309069293
316 | 20170103,0.37969648057585376
317 | 20170102,0.37969648057585376
318 | 20170101,0.37969648057585376
319 | 20161231,0.3575934832883298
320 | 20161230,0.3575934832883298
321 | 20161229,0.3575934832883298
322 | 20161228,0.3575934832883298
323 | 20161227,0.39987223183786164
324 | 20161226,0.39987223183786164
325 | 20161225,0.39987223183786164
326 | 20161224,0.4277726025271001
327 | 20161223,0.4277726025271001
328 | 20161222,0.4277726025271001
329 | 20161221,0.2816483590386045
330 | 20161220,0.2816483590386045
331 | 20161219,0.2816483590386045
332 | 20161218,0.2816483590386045
333 | 20161217,0.2816483590386045
334 | 20161216,0.3796936736131927
335 | 20161215,0.3796936736131927
336 | 20161214,0.3796936736131927
337 | 20161213,0.3796936736131927
338 | 20161212,0.3796936736131927
339 | 20161211,0.3796936736131927
340 | 20161210,0.3796936736131927
341 | 20161209,0.43782319004576975
342 | 20161208,0.43782319004576975
343 | 20161207,0.43782319004576975
344 | 20161206,0.43782319004576975
345 | 20161205,0.43782319004576975
346 | 20161204,0.43782319004576975
347 | 20161203,0.43782319004576975
348 | 20161202,0.29379326196655037
349 | 20161201,0.29379326196655037
350 | 20161130,0.29379326196655037
351 | 20161129,0.29379326196655037
352 | 20161128,0.29379326196655037
353 | 20161127,0.29379326196655037
354 | 20161126,0.35271100358911633
355 | 20161125,0.35271100358911633
356 | 20161124,0.35271100358911633
357 | 20161123,0.35271100358911633
358 | 20161122,0.35271100358911633
359 | 20161121,0.35271100358911633
360 | 20161120,0.3579517383343933
361 | 20161119,0.3579517383343933
362 | 20161118,0.3579517383343933
363 | 20161117,0.3579517383343933
364 | 20161116,0.3579517383343933
365 | 20161115,0.3579517383343933
366 | 20161114,0.3815485566088281
367 | 20161113,0.3815485566088281
368 | 20161112,0.3815485566088281
369 | 20161111,0.3815485566088281
370 | 20161110,0.4243532163835117
371 | 20161109,0.4243532163835117
372 | 20161108,0.4243532163835117
373 | 20161107,0.4243532163835117
374 | 20161106,0.4243532163835117
375 | 20161105,0.4243532163835117
376 | 20161104,0.37459906884480104
377 | 20161103,0.37459906884480104
378 | 20161102,0.37459906884480104
379 | 20161101,0.37459906884480104
380 | 20161031,0.42351258674997966
381 | 20161030,0.42351258674997966
382 | 20161029,0.42351258674997966
383 | 20161028,0.26480713905437514
384 | 20161027,0.26480713905437514
385 | 20161026,0.42814055674111173
386 | 20161025,0.3444665708146327
387 | 20161024,0.3444665708146327
388 | 20161023,0.3444665708146327
389 | 20161022,0.3444665708146327
390 | 20161021,0.3444665708146327
391 | 20161020,0.3632954194498043
392 | 20161019,0.3632954194498043
393 | 20161018,0.3632954194498043
394 | 20161017,0.45439791338417085
395 | 20161016,0.45439791338417085
396 | 20161015,0.2645992256385194
397 | 20161014,0.2645992256385194
398 | 20161013,0.2645992256385194
399 | 20161012,0.2645992256385194
400 | 20161011,0.231493274256136
401 | 20161010,0.231493274256136
402 | 20161009,0.231493274256136
403 | 20161008,0.231493274256136
404 | 20161007,0.4181447658397086
405 | 20161006,0.4181447658397086
406 | 20161005,0.4181447658397086
407 | 20161004,0.4181447658397086
408 | 20161003,0.39236930003947107
409 | 20161002,0.39236930003947107
410 | 20161001,0.39236930003947107
411 | 20160930,0.39236930003947107
412 | 20160929,0.39236930003947107
413 | 20160928,0.39236930003947107
414 | 20160927,0.3667874881555046
415 | 20160926,0.3667874881555046
416 | 20160925,0.3667874881555046
417 | 20160924,0.3667874881555046
418 | 20160923,0.3667874881555046
419 | 20160922,0.3667874881555046
420 | 20160921,0.3667874881555046
421 | 20160920,0.3667874881555046
422 | 20160919,0.41950303467340844
423 | 20160918,0.41950303467340844
424 | 20160917,0.41950303467340844
425 | 20160916,0.41950303467340844
426 | 20160915,0.41950303467340844
427 | 20160914,0.30583876254339193
428 | 20160913,0.30583876254339193
429 | 20160912,0.30583876254339193
430 | 20160911,0.30583876254339193
431 | 20160910,0.30583876254339193
432 | 20160909,0.30583876254339193
433 | 20160908,0.30583876254339193
434 | 20160907,0.30583876254339193
435 | 20160906,0.30583876254339193
436 | 20160905,0.30583876254339193
437 | 20160904,0.27703349209994277
438 | 20160903,0.27703349209994277
439 | 20160902,0.27703349209994277
440 | 20160901,0.27703349209994277
441 | 20160831,0.27703349209994277
442 | 20160830,0.27703349209994277
443 | 20160829,0.27703349209994277
444 | 20160828,0.27703349209994277
445 | 20160827,0.27703349209994277
446 | 20160826,0.27703349209994277
447 | 20160825,0.27703349209994277
448 | 20160824,0.27703349209994277
449 | 20160823,0.27703349209994277
450 | 20160822,0.27703349209994277
451 | 20160821,0.27703349209994277
452 | 20160820,0.27703349209994277
453 | 20160819,0.12397104529288874
454 | 20160818,0.12397104529288874
455 | 20160817,0.12397104529288874
456 | 20160816,0.12397104529288874
457 | 20160815,0.12397104529288874
458 | 20160814,0.12397104529288874
459 | 20160813,0.12397104529288874
460 | 20160812,0.12397104529288874
461 | 20160811,0.12397104529288874
462 | 20160810,0.12397104529288874
463 | 20160809,0.12397104529288874
464 | 20160808,0.12397104529288874
465 | 20160807,0.181541810372209
466 | 20160806,0.181541810372209
467 | 20160805,0.181541810372209
468 | 20160804,0.181541810372209
469 | 20160803,0.183626473495043
470 | 20160802,0.3933434479951281
471 | 20160801,0.19121252355762677
472 | 20160731,0.19121252355762677
473 | 20160730,0.2836800268328441
474 | 20160729,0.2836800268328441
475 | 20160728,0.2814179008072436
476 | 20160727,0.2814179008072436
477 | 20160726,0.3431396276824513
478 | 20160725,0.3431396276824513
479 | 20160724,0.2711219255654779
480 | 20160723,0.2711219255654779
481 | 20160722,0.2711219255654779
482 | 20160721,0.2711219255654779
483 | 20160720,0.2711219255654779
484 | 20160719,0.23900804057476338
485 | 20160718,0.23900804057476338
486 | 20160717,0.23900804057476338
487 | 20160716,0.2786617379853121
488 | 20160715,0.2786617379853121
489 | 20160714,0.2786617379853121
490 | 20160713,0.24593233476051457
491 | 20160712,0.24593233476051457
492 | 20160711,0.24593233476051457
493 | 20160710,0.24593233476051457
494 | 20160709,0.2546622350050327
495 | 20160708,0.2546622350050327
496 | 20160707,0.2546622350050327
497 | 20160706,0.2377765381091954
498 | 20160705,0.1550795454052526
499 | 20160704,0.2543687845687291
500 | 20160703,0.2238612967239434
501 | 20160702,0.2238612967239434
502 | 20160701,0.26863153815701046
503 | 20160630,0.26863153815701046
504 | 20160629,0.26863153815701046
505 | 20160628,0.3602535487769374
506 | 20160627,0.27628626041629173
507 | 20160626,0.27628626041629173
508 | 20160625,0.36405573272179753
509 | 20160624,0.36405573272179753
510 | 20160623,0.3428234880089395
511 | 20160622,0.29220762558609514
512 | 20160621,0.1892551974030994
513 | 20160620,0.1892551974030994
514 | 20160619,0.3540412868518666
515 | 20160618,None
516 | 20160617,0.36513453672807467
517 | 20160616,0.1911184778363609
518 | 20160615,0.1911184778363609
519 | 20160614,0.1911184778363609
520 | 20160613,0.44591831699141976
521 | 20160612,0.44591831699141976
522 | 20160611,0.44591831699141976
523 | 20160610,0.44591831699141976
524 | 20160609,0.44591831699141976
525 | 20160608,0.2633669782783534
526 | 20160607,0.2633669782783534
527 | 20160606,0.2633669782783534
528 | 20160605,0.2633669782783534
529 | 20160604,0.1723816707783527
530 | 20160603,0.1723816707783527
531 | 20160602,0.1723816707783527
532 | 20160601,0.1723816707783527
533 | 20160531,0.3738575063088357
534 | 20160530,0.3738575063088357
535 | 20160529,0.3738575063088357
536 | 20160528,0.3738575063088357
537 | 20160527,0.24549939463680057
538 | 20160526,0.30560050785306075
539 | 20160525,0.30560050785306075
540 | 20160524,0.30560050785306075
541 | 20160523,0.1959593726279536
542 | 20160522,0.1959593726279536
543 | 20160521,0.1959593726279536
544 | 20160520,0.22818581619549058
545 | 20160519,0.22818581619549058
546 | 20160518,0.22818581619549058
547 | 20160517,0.3356428160035545
548 | 20160516,0.3356428160035545
549 | 20160515,0.3356428160035545
550 | 20160514,0.3356428160035545
551 | 20160513,0.45336031366018936
552 | 20160512,0.35878698100300555
553 | 20160511,0.35878698100300555
554 | 20160510,0.35878698100300555
555 | 20160509,0.35878698100300555
556 | 20160508,0.35878698100300555
557 | 20160507,0.29144092102925595
558 | 20160506,0.29144092102925595
559 | 20160505,0.29144092102925595
560 | 20160504,0.29144092102925595
561 | 20160503,0.2561029997281787
562 | 20160502,0.30442258076342743
563 | 20160501,0.30442258076342743
564 | 20160430,0.30442258076342743
565 | 20160429,0.27596816240948013
566 | 20160428,0.27596816240948013
567 | 20160427,0.27596816240948013
568 | 20160426,0.27596816240948013
569 | 20160425,0.27596816240948013
570 | 20160424,0.27596816240948013
571 | 20160423,0.33823057784887334
572 | 20160422,0.33823057784887334
573 | 20160421,0.33823057784887334
574 | 20160420,0.39943268249482516
575 | 20160419,0.4019893026176416
576 | 20160418,0.3760449666982592
577 | 20160417,0.3760449666982592
578 | 20160416,0.3760449666982592
579 | 20160415,0.28413737896522867
580 | 20160414,0.28413737896522867
581 | 20160413,0.28413737896522867
582 | 20160412,0.28413737896522867
583 | 20160411,0.28413737896522867
584 | 20160410,0.5185095039224366
585 | 20160409,0.5185095039224366
586 | 20160408,0.5185095039224366
587 | 20160407,0.24866301703706298
588 | 20160406,0.24866301703706298
589 | 20160405,0.24866301703706298
590 | 20160404,0.24866301703706298
591 | 20160403,0.24866301703706298
592 | 20160402,0.3776666102163249
593 | 20160401,0.3776666102163249
594 | 20160331,0.44691124539982524
595 | 20160330,0.35660893969401114
596 | 20160329,0.36029458182896135
597 | 20160328,0.36029458182896135
598 | 20160327,0.2509085301823485
599 | 20160326,0.2509085301823485
600 | 20160325,0.2509085301823485
601 | 20160324,0.2509085301823485
602 | 20160323,0.29806920588018937
603 | 20160322,0.29806920588018937
604 | 20160321,0.29806920588018937
605 | 20160320,0.2632188453422218
606 | 20160319,0.2632188453422218
607 | 20160318,0.2632188453422218
608 | 20160317,0.2940941216704473
609 | 20160316,0.2940941216704473
610 | 20160315,0.2940941216704473
611 | 20160314,0.2940941216704473
612 | 20160313,0.1675868356079048
613 | 20160312,0.1675868356079048
614 | 20160311,0.1675868356079048
615 | 20160310,0.470822254763685
616 | 20160309,0.470822254763685
617 | 20160308,0.470822254763685
618 | 20160307,0.27763577626939734
619 | 20160306,0.27763577626939734
620 | 20160305,0.30604978887883344
621 | 20160304,0.30604978887883344
622 | 20160303,0.30604978887883344
623 | 20160302,0.34289601025196725
624 | 20160301,0.34289601025196725
625 | 20160229,0.34289601025196725
626 | 20160228,0.456150843448238
627 | 20160227,0.456150843448238
628 | 20160226,0.456150843448238
629 | 20160225,0.456150843448238
630 | 20160224,0.2835236239049316
631 | 20160223,0.2835236239049316
632 | 20160222,0.2910002071019829
633 | 20160221,0.2910002071019829
634 | 20160220,0.2910002071019829
635 | 20160219,0.21771782491281808
636 | 20160218,0.21771782491281808
637 | 20160217,0.32126829533814855
638 | 20160216,0.32126829533814855
639 | 20160215,0.32126829533814855
640 | 20160214,0.30863521159297325
641 | 20160213,0.30490166804054747
642 | 20160212,0.26014177070001776
643 | 20160211,0.3064875485393007
644 | 


--------------------------------------------------------------------------------