{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "![img](ep8_2018-57x57.png)\n", "\n", "# To bar, or not to bar - what is the intention?\n", "\n", "**post @** [endlesspint.com](http://endlesspint.com/2018-11-07-to-bar-or-not-to-bar/)" ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "import numpy as np\n", "import pandas as pd\n", "from scipy import stats\n", "\n", "import matplotlib.pyplot as plt\n", "%matplotlib inline" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "# comfort level\n", "L = 60.\n", "\n", "# wishing to attend\n", "N = 100.\n", "\n", "# previous 'm' Friday nights\n", "m = 2." ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "array([0.87344206, 0.96854381, 0.86920762, 0.53090261, 0.23280506,\n", " 0.01149766, 0.43052577, 0.40241112, 0.5227224 , 0.47844396])" ] }, "execution_count": 3, "metadata": {}, "output_type": "execute_result" } ], "source": [ "N = 100\n", "\n", "def preferences(N):\n", " return np.random.uniform(.0001,1,N)\n", "\n", "np.random.seed(8)\n", "patron_p = preferences(N)\n", "patron_p[:10]" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# https://stackoverflow.com/questions/22744577/plotting-basic-uniform-distribution-on-python\n", "count, bins, ignored = plt.hist(patron_p, 25, facecolor='green')" ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "the count that went out & stayed in, respectively: 57 43\n", "the tally for the 'correct' choice: 57\n" ] } ], "source": [ "def any_given_friday_night(patron_p, N=100, L=60, ind_friday=True):\n", " \n", " tally = np.zeros(N)\n", " # either one rand number or one for each, we'll see...\n", " if ind_friday == True:\n", " friday = np.random.uniform(.0001,1, N)\n", " else:\n", " friday = np.random.uniform(.0001, 1)\n", " \n", " going_out = friday <= patron_p\n", " staying_in = np.invert(going_out)\n", " \n", " yes_cnt = np.sum(going_out)\n", " nos_cnt = N - yes_cnt\n", " \n", " if yes_cnt == L: # bar at optimal capacity everyone choose right/wins\n", " tally += going_out\n", " tally += staying_in\n", "# print(tally)\n", " elif yes_cnt < L: # extra space at bar, should've gone out\n", " tally += going_out\n", " tally -= staying_in\n", "# print(tally)\n", " else: # overcrowded bar, should've stayed in\n", " tally -= going_out\n", " tally += staying_in\n", "# print(tally)\n", " \n", " return yes_cnt, nos_cnt, tally, going_out\n", "\n", "\n", "np.random.seed(9)\n", "bar_y, bar_n, tally_run, _ = any_given_friday_night(patron_p, ind_friday=True)\n", "print(\"the count that went out & stayed in, respectively: \", bar_y, bar_n)\n", "print(\"the tally for the 'correct' choice: \", np.sum(tally_run==1))" ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[30, 30, 29, 24, 28, 87, 30, 9, 95, 85]\n", "[70, 70, 71, 76, 72, 13, 70, 91, 5, 15]\n", "15\n", "[ 1. -1. -1. -1. -1. -1. 1. -1. 1. -1.]\n", "[True, True, True, True, True, False, True, True, False, False]\n" ] }, { "data": { "text/plain": [ "array([ 1., 1., 1., 1., 1., -1., 1., 1., -1., -1.])" ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "yes = []\n", "nos = []\n", "tally_run = np.zeros(N)\n", "# m = [] # historical winning strategy\n", "\n", "# np.random.seed(10)\n", "for i in range(100):\n", " bar_y, bar_n, tally_run, going_out = any_given_friday_night(patron_p, ind_friday=False)\n", " yes.append(bar_y)\n", " nos.append(bar_n)\n", " \n", " \n", "print(yes[-10:])\n", "print(nos[-10:])\n", "print(np.sum(tally_run==1))\n", "\n", "m = [True if y<=60 else False for y in yes]\n", "print(tally_run[-10:])\n", "print(m[-10:])\n", "np.zeros(10) + m[-10:] - np.invert(m[-10:])" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "" ] }, "execution_count": 7, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# https://matplotlib.org/gallery/lines_bars_and_markers/bar_stacked.html\n", "plt.figure(figsize=(16,8))\n", "ind = np.arange(100)\n", "p1 = plt.bar(ind, yes)\n", "p2 = plt.bar(ind, nos, bottom=yes)\n", "plt.axhline(y=60, linewidth=5, color='red')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "** individual cut offs for going out/staying in result in too wild of swings, above just one example; later on, even with updating preferences, it remains eratic **\n", "\n", "## iterate through tally, compare against 'd' (wins - loses threshold)\n", "## set up 'patron_p' change for \"losers\"\n", "## use 'm' to determine whether to up or lower the ps" ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [], "source": [ "# check tally, is it -3?\n", "\n", "# review recent history, m[-2:], to determine go out more/less\n", "# if 50/50 choose randomly\n", "\n", "# determine by how much to update (+/-) ps\n", "\n", "# plot ps hist over time" ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "0.5227224038854343" ] }, "execution_count": 9, "metadata": {}, "output_type": "execute_result" } ], "source": [ "np.random.seed(8)\n", "patron_p = preferences(N)\n", "patron_p[8]" ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Sanity check, preferences for ten persons (pre-run):\n", "\n", "[0.17883914 0.61976344 0.56232716 0.76342827 0.80666567 0.29869534\n", " 0.13847087 0.63543437 0.00840754 0.78860441]\n", "\n", "Sanity check, preferences for ten persons (post-run):\n", "[0.45363968 0.69200839 0.56136785 0.76342827 0.80666567 0.49585112\n", " 0.51139845 0.63543437 0.5059887 0.78860441]\n", "\n", "Last ten 'yes's':\n", "[61, 57, 63, 61, 62, 54, 64, 60, 68, 55]\n", "\n", "Last ten 'no's':\n", "[39, 43, 37, 39, 38, 46, 36, 40, 32, 45]\n", "\n", "Public memory ('shoulda gone out'):\n", "[False, True, False, False, False, True, False, True, False, True]\n", "\n", "Ending score for ten persons:\n", "[42. 38. 30. 58. 56. 10. 12. 12. 4. 24.]\n" ] } ], "source": [ "temp_pref = patron_p.copy()\n", "\n", "yes = []\n", "nos = []\n", "tally_run = np.zeros(N)\n", "m = [] # historical winning strategy\n", "pref_08 = []\n", "tally08_run = []\n", "tally08 = []\n", "bar_08 = []\n", "m_rate = []\n", "\n", "print(\"Sanity check, preferences for ten persons (pre-run):\\n\")\n", "print(patron_p[-10:])\n", "\n", "bar_runs = 1000\n", "d = -3 # losing threshold\n", "last_m_nights = 4\n", "change_rate = .1\n", "\n", "for i in range(bar_runs):\n", " bar_y, bar_n, tally, going_out = any_given_friday_night(temp_pref)\n", " yes.append(bar_y)\n", " nos.append(bar_n)\n", " tally_run += tally\n", " tally08_run.append(tally[8])\n", " \n", " # historical winning strategy\n", " if bar_y <= 60:\n", " m.append(True)\n", " else:\n", " m.append(False)\n", " \n", " pref_08.append(patron_p[8])\n", " tally08.append(tally_run[8])\n", " \n", " # check tally, is it -3?\n", " # update directionally to recent past\n", " m_recent = np.sum(m[-last_m_nights:])/np.min([last_m_nights, len(m)])\n", " m_rate.append(m_recent)\n", "# print(m_recent)\n", " temp_pref[tally_run <= d] += (temp_pref[tally_run <= d] - m_recent) * -change_rate\n", " tally_run[tally_run <= d] = tally[tally_run <= d] * 0.\n", " \n", " \n", " bar_08.append(going_out[8])\n", " \n", "\n", "print(\"\\nSanity check, preferences for ten persons (post-run):\")\n", "print(temp_pref[-10:])\n", "print(\"\\nLast ten 'yes's':\")\n", "print(yes[-10:])\n", "print(\"\\nLast ten 'no's':\")\n", "print(nos[-10:])\n", "print(\"\\nPublic memory ('shoulda gone out'):\")\n", "print(m[-10:])\n", "print(\"\\nEnding score for ten persons:\")\n", "print(tally_run[-10:])\n", "\n", "# stats.describe(tally)" ] }, { "cell_type": "code", "execution_count": 11, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
barhomego_outbar08ps08+/-ty08m_rate
9906139FalseFalse0.5227221.01.00.25
9915743TrueFalse0.522722-1.00.00.50
9926337FalseTrue0.522722-1.0-1.00.50
9936139FalseFalse0.5227221.00.00.25
9946238FalseFalse0.5227221.01.00.25
9955446TrueFalse0.522722-1.00.00.25
9966436FalseTrue0.522722-1.0-1.00.25
9976040TrueTrue0.5227221.00.00.50
9986832FalseTrue0.522722-1.0-1.00.50
9995545TrueFalse0.522722-1.0-2.00.50
\n", "
" ], "text/plain": [ " bar home go_out bar08 ps08 +/- ty08 m_rate\n", "990 61 39 False False 0.522722 1.0 1.0 0.25\n", "991 57 43 True False 0.522722 -1.0 0.0 0.50\n", "992 63 37 False True 0.522722 -1.0 -1.0 0.50\n", "993 61 39 False False 0.522722 1.0 0.0 0.25\n", "994 62 38 False False 0.522722 1.0 1.0 0.25\n", "995 54 46 True False 0.522722 -1.0 0.0 0.25\n", "996 64 36 False True 0.522722 -1.0 -1.0 0.25\n", "997 60 40 True True 0.522722 1.0 0.0 0.50\n", "998 68 32 False True 0.522722 -1.0 -1.0 0.50\n", "999 55 45 True False 0.522722 -1.0 -2.0 0.50" ] }, "execution_count": 11, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# a look at one 'customer' over time\n", "# sanity check\n", "\n", "pd.DataFrame(\n", " {'bar': yes,\n", " 'home': nos,\n", " 'go_out': m, \n", " 'ps08': pref_08,\n", " '+/-': tally08_run,\n", " 'ty08': tally08,\n", " 'bar08': bar_08,\n", " 'm_rate': m_rate\n", " }, columns=['bar', 'home', 'go_out', 'bar08', 'ps08', '+/-', 'ty08', 'm_rate']).tail(10)" ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "" ] }, "execution_count": 12, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# https://stackoverflow.com/questions/22744577/plotting-basic-uniform-distribution-on-python\n", "count, bins, ignored = plt.hist(temp_pref, 25, facecolor='green')\n", "\n", "# https://matplotlib.org/gallery/lines_bars_and_markers/bar_stacked.html\n", "plt.figure(figsize=(16,8))\n", "ind = np.arange(bar_runs)\n", "p1 = plt.bar(ind, yes)\n", "p2 = plt.bar(ind, nos, bottom=yes)\n", "plt.axhline(y=60, linewidth=5, color='red')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "* timeseries: capture runnnig tally for all N (100)" ] }, { "cell_type": "code", "execution_count": 13, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "pre preferences: [0.87344206 0.96854381 0.86920762 0.53090261 0.23280506 0.01149766\n", " 0.43052577 0.40241112 0.5227224 0.47844396]\n", "\n", "post preferences: [ 8.73442060e-01 9.68543809e-01 8.69207621e-01 -4.23812220e+12\n", " -4.23913459e+13 -5.46196826e+13 5.38725875e-01 -1.15732262e+13\n", " 5.22722404e-01 -6.89893697e+12]\n", "\n", "interquartile (cumulative): -938.5 140.0 472.0\n", "\n", "interquartile (final tally): -0.0 146.5 472.0\n", "\n", "final tally:\n", " [748. 918. 736. -1. -2. -2. 80. -1. 72. -1. 298. 195. 512. 470.\n", " 246. -0. -0. 944. -0. -1. 198. 968. -0. 143. -2. 75. 42. 790.\n", " -2. 974. 81. -0. 138. 508. 860. -2. 90. -2. 654. 296. -2. -0.\n", " 628. 245. -2. 554. 970. 144. -0. 47. 158. 216. 221. 282. 106. -0.\n", " 478. 898. 876. -2. 202. 290. -2. -1. 251. -2. 74. -2. 77. 181.\n", " 220. 385. -1. 149. 638. 99. 412. 123. 690. 448. 690. -0. -2. -1.\n", " 48. 210. 604. 70. 884. 450. 122. 177. 152. 494. 610. 103. -0. 246.\n", " 49. 588.]\n" ] } ], "source": [ "# what if they weren't describing the threshold, e.g., -3, as an absolute tally \n", "# but rather a short term one, say -3 in last dozen, ten, or five nights out?\n", "# or even anytime it tallys -3 regardless window (too extreme?)\n", "\n", "\n", "def going_out_and_evaluate(patron_p, N=len(patron_p), L=int(N*.6), d=-3, m_nights=2, change_rate=.1, bar_runs=1000, \n", " ind_friday=True,\n", " re_eval=True, re_eval_binary=False, re_eval_w_contrarian=False):\n", " '''\n", " Parameters:\n", " -----------\n", " \n", " patron_p : ndarray of patron preferences\n", " N : number of patrons\n", " L : max comfort level of venue\n", " d : losing threshold\n", " m_nights : number of most recent nights' best policy\n", " bar_runs : \n", " change_rate : \n", " re_eval : whether to reset patron preferences\n", " '''\n", " \n", " temp_pref = patron_p\n", " yes = []\n", " nos = []\n", " \n", " # historical winning strategy\n", " m = []\n", " \n", " # historical v contrarian preference\n", " if re_eval_w_contrarian==True:\n", " hist_contra = np.random.randint(2, size=N)\n", " hist_contra += (hist_contra-1)\n", " else:\n", " hist_contra = np.ones(N)\n", " \n", " tally_run = np.zeros(N)\n", " M = bar_runs+1\n", " tally_matrix = np.zeros(N*M).reshape(N, M)\n", "\n", " for i in range(bar_runs):\n", " bar_y, bar_n, tally, going_out = any_given_friday_night(temp_pref, ind_friday=ind_friday)\n", " yes.append(bar_y)\n", " nos.append(bar_n)\n", " tally_run += tally\n", " tally_matrix[:, i+1] = tally_matrix[:, i] + tally\n", "\n", " # historical winning strategy\n", " if bar_y <= L:\n", " m.append(True)\n", " else:\n", " m.append(False)\n", "\n", "\n", " if re_eval==True:\n", " \n", " # check tally, update directionally to recent past\n", " m_recent = np.sum(m[-m_nights:])/np.min([m_nights, len(m)])\n", " \n", " if re_eval_binary==True:\n", " # only care if should've gone out more/less & update\n", " discrete_target = (m_recent >= 0.5) * 1.0\n", " temp_pref[tally_run <= d] += (temp_pref[tally_run <= d] - discrete_target) * (-change_rate * hist_contra[tally_run <= d])\n", " tally_run[tally_run <= d] = tally[tally_run <= d] * 0.\n", " else:\n", " # by what ratio should've gone out more/less & update\n", " temp_pref[tally_run <= d] += (temp_pref[tally_run <= d] - m_recent) * (-change_rate * hist_contra[tally_run <= d])\n", " tally_run[tally_run <= d] = tally[tally_run <= d] * 0.\n", "\n", " return yes, nos, bar_runs, hist_contra, tally_run, tally_matrix, temp_pref\n", "\n", "\n", "## run function\n", "\n", "np.random.seed(8)\n", "patron_p = preferences(N)\n", "print(\"pre preferences: \", patron_p[:10])\n", "yes, nos, bar_runs, _, final_tally, bar_matrix, new_pref = going_out_and_evaluate(patron_p, ind_friday=True,\n", " re_eval=True, re_eval_binary=True, re_eval_w_contrarian=True)\n", "print(\"\\npost preferences: \", new_pref[:10])\n", "print(\"\\ninterquartile (cumulative): \", np.percentile(bar_matrix[:,-1], 25), np.percentile(bar_matrix[:,-1], 50), np.percentile(bar_matrix[:,-1], 75))\n", "print(\"\\ninterquartile (final tally): \", np.percentile(final_tally, 25), np.percentile(final_tally, 50), np.percentile(final_tally, 75))\n", "print(\"\\nfinal tally:\\n\", final_tally)" ] }, { "cell_type": "code", "execution_count": 14, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "## plots\n", "\n", "f, ((ax1, ax2), (ax3, ax4)) = plt.subplots(2, 2, figsize=(18, 8))\n", "\n", "ax1.hist(patron_p, 25, facecolor='green')\n", "ax1.set_title(\"Final 'Going Out' Preferences\")\n", "\n", "\n", "# https://matplotlib.org/gallery/lines_bars_and_markers/bar_stacked.html\n", "ind = np.arange(bar_runs)\n", "p1 = ax2.bar(ind, yes)\n", "p2 = ax2.bar(ind, nos, bottom=yes)\n", "ax2.axhline(y=60, linewidth=5, color='red')\n", "ax2.set_title(\"Simulated Nights Out\")\n", "\n", "for row in range(bar_matrix.shape[0]):\n", " ax3.plot(bar_matrix[row])\n", "ax3.set_title(\"Cumulative Benefit\")\n", "\n", "ax4.scatter(new_pref, bar_matrix[:,-1])\n", "ax4.set_xlim(0.,1.)\n", "ax4.axhline(y=np.percentile(bar_matrix[:,-1], 25), linewidth=2, color='green', ls='--')\n", "ax4.axhline(y=np.percentile(bar_matrix[:,-1], 75), linewidth=2, color='green', ls='--')\n", "ax4.set_title(\"Final Pref x Cumulative Scores\")\n", " \n", "plt.tight_layout()" ] }, { "cell_type": "code", "execution_count": 15, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "C:\\Users\\rstancut\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:63: RuntimeWarning: overflow encountered in add\n", "C:\\Users\\rstancut\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:63: RuntimeWarning: invalid value encountered in add\n", "C:\\Users\\rstancut\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:10: RuntimeWarning: invalid value encountered in less_equal\n", " # Remove the CWD from sys.path while we load stuff.\n", "C:\\Users\\rstancut\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:67: RuntimeWarning: overflow encountered in add\n", "C:\\Users\\rstancut\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:67: RuntimeWarning: invalid value encountered in add\n" ] }, { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
d_thresholdm_nightschange_ratebinary_evalincl_contrasetting_runmedian_tallymean_tallymean_cuml
0-3100.9TrueTrue080.098.2279.20
1-3100.9TrueTrue1285.0257.27245.36
2-3100.9TrueFalse0274.0316.18316.12
3-3100.9TrueFalse1360.0303.69302.10
4-3100.9FalseTrue0300.0311.06310.82
\n", "
" ], "text/plain": [ " d_threshold m_nights change_rate binary_eval incl_contra setting_run \\\n", "0 -3 10 0.9 True True 0 \n", "1 -3 10 0.9 True True 1 \n", "2 -3 10 0.9 True False 0 \n", "3 -3 10 0.9 True False 1 \n", "4 -3 10 0.9 False True 0 \n", "\n", " median_tally mean_tally mean_cuml \n", "0 80.0 98.22 79.20 \n", "1 285.0 257.27 245.36 \n", "2 274.0 316.18 316.12 \n", "3 360.0 303.69 302.10 \n", "4 300.0 311.06 310.82 " ] }, "execution_count": 15, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# ds = [-10, -5, -3, -2, -1]\n", "# ms = [10, 5, 3, 2, 1]\n", "# cr = [.9, .75, .5, .2, .1]\n", "\n", "ds = [-3, -2, -1]\n", "ms = [10, 5, 3]\n", "cr = [.9, .5, .2]\n", "bn = [True, False]\n", "ct = [True, False]\n", "\n", "param_stats = []\n", "\n", "\n", "for i in range(len(ds)):\n", " for j in range(len(ms)):\n", " for k in range(len(cr)):\n", " for l in range(len(bn)):\n", " for m in range(len(ct)):\n", " for n in range(2):\n", "\n", " temp_dict = {}\n", " temp_dict['d_threshold'] = ds[i]\n", " temp_dict['m_nights'] = ms[j]\n", " temp_dict['change_rate'] = cr[k]\n", " temp_dict['binary_eval'] = bn[l]\n", " temp_dict['incl_contra'] = ct[m]\n", " temp_dict['setting_run'] = n\n", "\n", " yes, nos, bar_runs, _, final_tally, bar_matrix, new_pref = going_out_and_evaluate(patron_p, d=ds[i], m_nights=ms[j], \n", " change_rate=cr[k], re_eval_binary=bn[l], \n", " re_eval_w_contrarian=ct[m])\n", "\n", " # mean, minmax = stats.bayes_mvs(bar_matrix[:,-1], .95)[0]\n", " # temp_dict['min_ci_95'] = minmax[0]\n", " # temp_dict['max_ci_95'] = minmax[1]\n", "\n", " temp_dict['median_tally'] = np.median(final_tally)\n", " temp_dict['mean_tally'] = np.mean(final_tally)\n", " temp_dict['mean_cuml'] = np.mean(bar_matrix[:,-1])\n", "\n", " param_stats.append(temp_dict)\n", "\n", "\n", "df_param_stats = pd.DataFrame(param_stats)\n", "df_param_stats = df_param_stats[['d_threshold', 'm_nights', 'change_rate', \n", " 'binary_eval', 'incl_contra', 'setting_run', \n", " 'median_tally', 'mean_tally', 'mean_cuml']]\n", "df_param_stats.head()" ] }, { "cell_type": "code", "execution_count": 16, "metadata": {}, "outputs": [], "source": [ "# df_param_stats.to_excel(\"param_stats.xlsx\")" ] }, { "cell_type": "code", "execution_count": 19, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "C:\\Users\\rstancut\\AppData\\Local\\Continuum\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:10: RuntimeWarning: invalid value encountered in less_equal\n", " # Remove the CWD from sys.path while we load stuff.\n" ] }, { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
d_thresholdm_nightschange_ratebinary_evalincl_contrasetting_runmean_cumlhist_cntmean_cuml_histcontr_cntmean_cuml_contr
0-3100.9TrueTrue0103.1446-101.69565254277.629630
1-3100.9TrueTrue1102.665375.39622647133.404255
2-3100.9TrueTrue2103.6648107.7500005299.884615
3-3100.9TrueTrue3103.345224.26923148189.000000
4-3100.9TrueTrue4104.405573.38181845142.311111
\n", "
" ], "text/plain": [ " d_threshold m_nights change_rate binary_eval incl_contra setting_run \\\n", "0 -3 10 0.9 True True 0 \n", "1 -3 10 0.9 True True 1 \n", "2 -3 10 0.9 True True 2 \n", "3 -3 10 0.9 True True 3 \n", "4 -3 10 0.9 True True 4 \n", "\n", " mean_cuml hist_cnt mean_cuml_hist contr_cnt mean_cuml_contr \n", "0 103.14 46 -101.695652 54 277.629630 \n", "1 102.66 53 75.396226 47 133.404255 \n", "2 103.66 48 107.750000 52 99.884615 \n", "3 103.34 52 24.269231 48 189.000000 \n", "4 104.40 55 73.381818 45 142.311111 " ] }, "execution_count": 19, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# just the high performing settings\n", "\n", "ds = [-3, -2]\n", "ms = [10, 5, 3]\n", "cr = [.9, .5, .2]\n", "bn = [True, False]\n", "ct = [True]\n", "\n", "param_stats = []\n", "\n", "\n", "for i in range(len(ds)):\n", " for j in range(len(ms)):\n", " for k in range(len(cr)):\n", " for l in range(len(bn)):\n", " for m in range(len(ct)):\n", " for n in range(5): # set to 5 for example purpose only, actual run at 100\n", "\n", " temp_dict = {}\n", " temp_dict['d_threshold'] = ds[i]\n", " temp_dict['m_nights'] = ms[j]\n", " temp_dict['change_rate'] = cr[k]\n", " temp_dict['binary_eval'] = bn[l]\n", " temp_dict['incl_contra'] = ct[m]\n", " temp_dict['setting_run'] = n\n", "\n", " yes, nos, bar_runs, hist_contra, final_tally, bar_matrix, new_pref = going_out_and_evaluate(patron_p, d=ds[i], m_nights=ms[j], \n", " change_rate=cr[k], re_eval_binary=bn[l], \n", " re_eval_w_contrarian=ct[m])\n", "\n", " # mean, minmax = stats.bayes_mvs(bar_matrix[:,-1], .95)[0]\n", " # temp_dict['min_ci_95'] = minmax[0]\n", " # temp_dict['max_ci_95'] = minmax[1]\n", "\n", "# temp_dict['median_tally'] = np.median(final_tally)\n", "# temp_dict['mean_tally'] = np.mean(final_tally)\n", " temp_dict['mean_cuml'] = np.mean(bar_matrix[:,-1])\n", " temp_dict['hist_cnt'] = np.sum([hist_contra==1])\n", " temp_dict['mean_cuml_hist'] = np.mean(bar_matrix[:,-1][hist_contra==1])\n", " temp_dict['contr_cnt'] = np.sum([hist_contra==-1])\n", " temp_dict['mean_cuml_contr'] = np.mean(bar_matrix[:,-1][hist_contra==-1])\n", "\n", " param_stats.append(temp_dict)\n", "\n", "\n", "df_param_stats = pd.DataFrame(param_stats)\n", "df_param_stats = df_param_stats[['d_threshold', 'm_nights', 'change_rate', \n", " 'binary_eval', 'incl_contra', 'setting_run', \n", " 'mean_cuml', 'hist_cnt', 'mean_cuml_hist', 'contr_cnt', 'mean_cuml_contr']]\n", "df_param_stats.head()" ] }, { "cell_type": "code", "execution_count": 20, "metadata": {}, "outputs": [], "source": [ "# df_param_stats.to_excel(\"param_stats_high_perf.xlsx\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## random notes below" ] }, { "cell_type": "code", "execution_count": 21, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "Text(0,0.5,'settings')" ] }, "execution_count": 21, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# y = np.arange(df_param_stats.shape[0])\n", "x = df_param_stats.mean_cuml\n", "y = np.arange(df_param_stats.shape[0])\n", "plt.plot(x, y, 'o')\n", "# plt.plot(*z, '-')\n", "plt.xlabel(\"simulation mean\")\n", "plt.ylabel(\"settings\")" ] }, { "cell_type": "code", "execution_count": 22, "metadata": {}, "outputs": [], "source": [ "# use decision tree to identify parameters/thresholds to get high score (25+)\n", " # i.e., change_rate == 0.2, m_nights == 10, etc.\n", "\n", "# dot plot with ci lines for settings\n", " # or: https://seaborn.pydata.org/generated/seaborn.violinplot.html\n", "\n", "# also, CLT: run above N times and grab mean" ] }, { "cell_type": "code", "execution_count": 23, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "" ] }, "execution_count": 23, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAD8CAYAAAB+UHOxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAGitJREFUeJzt3X2QXNV95vHvw1gmUipmhF4SGGkiiGUWG1mWMgatWbQO3iCsYqNBaydQ9qJgB1VS2ix4y9qgmFrHWSiBlUrKLle8q1rYgrzIBiNPlApGVuwNrjiRnJFlIclEQbYx0ggb2dIoRWmMR8Nv/+jbojW6/TL39sx0930+VVPTfe7bOX1n7q/vuedFEYGZmRXPRdOdATMzmx4OAGZmBeUAYGZWUA4AZmYF5QBgZlZQDgBmZgXlAGBmVlAOAGZmBeUAYGZWUG+Y7gzUMnfu3Fi0aNF0Z8PMrK3s3bv3RxExr956LR0AFi1axODg4HRnw8ysrUj6fiPruQrIzKygHADMzArKAcDMrKAcAMzMCsoBwMysoOq2ApL0CHAL8HJEXJOkXQp8HlgEvAD8ekSckiTgU8Bq4AzwmxHxzWSbdcB9yW7vj4hHm1sUK4KBfUNs2XmY48MjXN49k42rrqJ/WU/HHK/RPAATylezyjGwb4g/2HGI4ZFRAGbPmsHH/+PbGtrXwL4hfn/7s5wZfQ0ACd515aW88OORqvmqdbxqZbpv4ADb9hxlLIIuiSvnzeK7J86ce3/7dQu5v39JzXx+4q8PcerM6Ll8RkBPxs+tnM+h4RG6JMYiMu+r2VRvRjBJK4FXgMcqAsAngZMR8aCke4HZEfF7klYDv0spAFwHfCoirksCxiDQBwSwF/jliDhV69h9fX3hZqBWNrBviE3bDzAyOnYubeaMLjavXTIp/0hTfbxG8zDjIoFgdOz1/91a+WpWOQb2DbHxif2Mvnb+NWNGl9jyvqV1A9B/e/xbvFZnAsLKfNU63m+8cyFP7h26oEzLey/h6985WbcsH1zRmxoEBvYNsfEL+8/7bKvlrxFpn33WfU2EpL0R0VdvvbpVQBHxNWD8J7oGKH+DfxTor0h/LEp2A92SLgNWAbsi4mRy0d8F3NxYUcxKtuw8fME/0sjoGFt2Hu6I4zWah9HX4oILVK18NascW3YevuBiDKVAVG9fW3YernvxH5+vWsfbtudoapkaufgDbNtztGo+q138x+evEWmffdZ9TYaszwB+PiJeAkh+z0/Se4DKT/ZYklYt/QKS1ksalDR44sSJjNmzTnR8eGRC6e12vLzHmmh+J1qOWuvX21eWctTaZiznXObVtm8kn804J1n2NRma/RBYKWlRI/3CxIitEdEXEX3z5tXtyWwFcnn3zAmlt9vx8h5rovmdaDlqrV9vX1nKUWubLqVdUhpXbftG8tmMc5JlX5MhawD4YVK1Q/L75ST9GLCwYr0FwPEa6WYN27jqKmbO6DovbeaMrnMPRdv9eI3mYcZFYkbX+RewWvlqVjk2rrqq9PxhnBldqruvjauuImXTC1Tmq9bxbr9uYWqZrv+lS+sfBLj9uoWp6RtXXXXBZ1stf41I++yz7msyZA0AO4B1yet1wF9VpN+hkhXA6aSKaCdwk6TZkmYDNyVpZg3rX9bD5rVL6OmeiSi1ypjMB7JTfbxG87Dl/UvZ8r6lDeerWeXoX9bDlvcvpXvmjHNps2fNqPsAuLztH//6O5g14/VLjgTX/9KlVfNV63j39y9JLdNf3PVv+eCK3nPf8LskFs//2fPeV3sAfO6Y71vK7FmvH7N8s5Dlc6v87MvHz7qvydBIK6BtwLuBucAPgY8DA8DjQC/wIvD+iDiZNAP9DKUHvGeAOyNiMNnPh4DfT3b7QET833qZcysgM7OJa7QVUN0AMJ0cAMzMJq5pzUDNzKwztfR8ADY1WqG3q5lNPQeAghvfU3FoeIRN2w8AOAiYdThXARVcK/R2NbPp4QBQcK3Q29XMpocDQMG1Qm9XM5seDgAF1wq9Xc1sevghcMGVH/S6FZBZ8fgOwBj8/kl+cPonBPCD0z9h8PuNDalrZu3NdwAFd9/AAf5894vn3o9FnHtfa9YkM2t/vgMouGoTY1RLN7PO4QBQcNUmxsg74YaZtT4HgIKrNjFG3gk3zKz1OQAUXLWJMaqlm1nn8EPggis/6N225yhjEXSpNNuSHwCbdT4HgIJJG/nz/v4lvuCbFZADQIF45E8zq+RnAAXikT/NrJIDQIF45E8zq+QAUCAe+dPMKjkAFIhH/jSzSn4IXCAe+dPMKjkAFEz/sh5f8M0McBWQmVlh5QoAku6WdFDSIUn3JGnvkLRb0rckDUq6NkmXpE9LOiLpWUnLm1EAMzPLJnMVkKRrgLuAa4GfAk9L+hvgk8AnIuJLklYn798NvBdYnPxcB3w2+W1TIK0HsKuCzIotzzOAq4HdEXEGQNIzwK1AAG9K1rkEOJ68XgM8FhEB7JbULemyiHgpRx6sAe4BbGZp8lQBHQRWSpojaRawGlgI3ANskXQU+CNgU7J+D1A5y8ixJM0mmXsAm1mazAEgIp4DHgJ2AU8D+4GzwO8AH4mIhcBHgIeTTdIGmL9g1hFJ65NnB4MnTpzImj2r4B7AZpYm10PgiHg4IpZHxErgJPA8sA7YnqzyBKVnBFD6xl85yPwCXq8eqtzn1ojoi4i+efPm5cmeJdwD2MzS5G0FND/53QusBbZRuqj/+2SVGykFBYAdwB1Ja6AVwGnX/08N9wA2szR5O4I9KWkOMApsiIhTku4CPiXpDcBPgPXJuk9Rek5wBDgD3Jnz2NaAcuufkdExuiTGIuhxKyAzI2cAiIgbUtL+HvjllPQANuQ5nk3M+NY/YxHnvvn74m9m7gncwdz6x8xqcQDoYG79Y2a1OAB0MLf+MbNaHAA6mFv/mFktHg66g3n8fzOrxQGgA9Qa6M3j/5tZNQ4Abc4DvZlZVn4G0Obc1NPMsnIAaHNu6mlmWTkAtDk39TSzrBwA2pybeppZVn4I3Obc1NPMsvIdgJlZQfkOoM25GaiZZeU7gDbnZqBmlpUDQBsb2DfEkJuBmllGDgBtqlz1U42bgZpZPQ4AbSqt6qfMzUDNrBEOAG2qVhXP5rVL/ADYzOpyAGhT1ap4uqQpzomZtSsHgDaV1gMYShO/b9p+gIF9Q9OQKzNrJw4Abap/WQ+b1y5J/cbvZqBm1ggHgDbWv6yHsYjUZW4Gamb1OAC0sYF9Q1Sr8XczUDOrxwGgjW3ZeZi07/8CNwM1s7pyBQBJd0s6KOmQpHsq0n9X0uEk/ZMV6ZskHUmWrcpzbKtezRN4HCAzqy/zYHCSrgHuAq4Ffgo8LelvgAXAGuDtEfGqpPnJ+m8FbgPeBlwO/K2kt0REem8mq+vy7pmpQ0H0uPrHzBqQ5w7gamB3RJyJiLPAM8CtwO8AD0bEqwAR8XKy/hrgcxHxakR8DzhCKXhYRp4MxszyyBMADgIrJc2RNAtYDSwE3gLcIGmPpGckvTNZvwc4WrH9sSTtPJLWSxqUNHjixIkc2et85aagPd0zEaVv/u4FbGaNylwFFBHPSXoI2AW8AuwHzib7nA2sAN4JPC7pSkhtsHLBM8yI2ApsBejr60tv41gQA/uG6s701b+sxxd8M8sk10PgiHg4IpZHxErgJPA8pW/226PkG8BrwNwkfWHF5guA43mO38nKo30ODY8QvD7Ri3v4mlmz5G0FVH7A2wusBbYBA8CNSfpbgDcCPwJ2ALdJuljSFcBi4Bt5jt/JPNGLmU22vFNCPilpDjAKbIiIU5IeAR6RdJBS66B1ERHAIUmPA9+mVFW0wS2AqqvWxLMyvZEqIjOzanIFgIi4ISXtp8AHq6z/APBAnmMWRbUmnuUevp4L2Mzyck/gFlWviaeriMwsr7xVQDZJyt/it+w8zNDwCF0SI6NjfOKvD/EHOw4xPDKaup0HgTOzRjkAtLByEKis6jl1Jv3CX+ZB4MysUa4CanG15v4dz72AzWwifAfQ4hqt0ulxKyAzmyAHgBZXrTVQpZ7umXz93hunKEdm1ilcBdTiqs39W+ZqHzPLyncALa6yNdDx4RG6Z80gAk6PjLrzl5nl4gDQBsYP+FbZA7jc7t9BwMwmygGgzbgHsJk1i58BtBn3ADazZnEAaDONDBJnZtYIB4A2U62nr3sAm9lEOQC0Gc8DbGbN4ofAbWZ8s1A3BTWzrBwAWkzaJC/gfgBm1nwOAC0krYnnxif2g2B0LIDzRwN1E1Azy8PPAFpIWhPP0dfi3MU/jZuAmllWDgAtJGtTTjcBNbMsHABaSNamnG4CamZZOAC0kHojf6ZxE1Azy8oPgVvI+Cae1Wv+QeBWQGaWiwNAi6kc+fP6B7+aOhmMJ4Axs2ZwFVALc69fM5tMuQKApLslHZR0SNI945Z9VFJImpu8l6RPSzoi6VlJy/Mcuwj6l/Wwee0SerpnIkrf/DevXeIqHzNriswBQNI1wF3AtcBS4BZJi5NlC4FfBV6s2OS9wOLkZz3w2azHLopyr+Ch4REukhhKJoAZ2Dc03Vkzsw6Q5w7gamB3RJyJiLPAM8CtybI/Af47nPcccw3wWJTsBrolXZbj+B2t3Cu4/AxgLEofZbn3r4OAmeWVJwAcBFZKmiNpFrAaWCjp14ChiNg/bv0e4GjF+2NJmqVI6xVc5t6/ZtYMmVsBRcRzkh4CdgGvAPuBs8DHgJtSNlHabi5YSVpPqYqI3t7erNlre/V697r3r5nlleshcEQ8HBHLI2IlcBJ4AbgC2C/pBWAB8E1Jv0DpG//Cis0XAMdT9rk1Ivoiom/evHl5stfW6vXude9fM8srbyug+cnvXmAtpTr++RGxKCIWUbroL4+IHwA7gDuS1kArgNMR8VK+7HeuWr2C3RTUzJohb0ewJyXNAUaBDRFxqsa6T1F6TnAEOAPcmfPYLS9tbP9Gm3BW9goeGh6hS2Isgh73/jWzJlFErQEHpldfX18MDg5OdzYyGT+2P5S+ubsdv5lNNkl7I6Kv3nruCTxJ0lrxuPWOmbUSB4BJUq2VjlvvmFmrcACYJNVa6bj1jpm1CgeASeKB3Mys1Xk46Ekyfmx/j91vZq3GAWASVY7tb2bWalwFZGZWUA4AZmYF5QBgZlZQDgBmZgXlAGBmVlAOAGZmBeUAYGZWUA4AZmYF5QBgZlZQDgBmZgXlAGBmVlAOAGZmBeUAYGZWUB4NtEXkmUDezCwLB4AWMH4C+aHhETZtPwDgIGBmk8ZVQC3AE8ib2XRwAGgB1SaKHxoeYWDf0BTnxsyKwgGgBdSaKH7T9gMOAmY2KRwAWkDaBPJlrgoys8mSKwBIulvSQUmHJN2TpG2R9M+SnpX0RUndFetvknRE0mFJq/JmvlP0L+th89olVZcPDY9w/YNf9Z2AmTVV5gAg6RrgLuBaYClwi6TFwC7gmoh4O/AvwKZk/bcCtwFvA24G/lRS+tfeAupf1kNPjaqgcssgBwEza5Y8dwBXA7sj4kxEnAWeAW6NiC8n7wF2AwuS12uAz0XEqxHxPeAIpeBhiVpVQeDqIDNrrjwB4CCwUtIcSbOA1cDCcet8CPhS8roHOFqx7FiSdh5J6yUNSho8ceJEjuy1n3JVUK07gWothszMJipzAIiI54CHKFX5PA3sB8rf/JH0seT9X5ST0naTst+tEdEXEX3z5s3Lmr221b+sh6/fe2PVIFCrxZCZ2UTkeggcEQ9HxPKIWAmcBJ4HkLQOuAX4QESUL/LHOP8OYQFwPM/xO1laddDMGV1sXHXVNOXIzDpN3lZA85PfvcBaYJukm4HfA34tIs5UrL4DuE3SxZKuABYD38hz/E5WWR0koKd7JpvXLvHQEGbWNHnHAnpS0hxgFNgQEackfQa4GNglCUoPin87Ig5Jehz4NqWqoQ0RMVZ1z0b/sh5f8M1s0uQKABFxQ0ram2us/wDwQJ5jmplZc7gnsJlZQTkAmJkVlAOAmVlBOQCYmRWUA4CZWUF5SsgpdN/AAbbtOcpYBF0St1+3kPv7q48CamY2mRwApsh9Awf4890vnns/FnHuvYOAmU0HVwFNkW17jk4o3cxssvkOIKOBfUNs2XmY48MjXN49k42rrqrZa3csLhj3rma6mdlkcwDIYGDfEJu2H2BktDSSRXmyFqBqEOiSUi/2XUobJNXMbPK5CiiDLTsPn7v4l9WbrOX268ZPlVA73cxssvkOIINqk7LUmqyl/KDXrYDMrFU4AGRwefdMhlIu9vUma7m/f4kv+GbWMlwFlIEnazGzTuA7gAzKD3on0grIzKzVOABk5MlazKzduQrIzKygHADMzArKAcDMrKAcAMzMCsoBwMysoBwAzMwKygHAzKygHADMzAoqVwCQdLekg5IOSbonSbtU0i5Jzye/ZyfpkvRpSUckPStpeTMKYGZm2WQOAJKuAe4CrgWWArdIWgzcC3wlIhYDX0neA7wXWJz8rAc+myPfZmaWU547gKuB3RFxJiLOAs8AtwJrgEeTdR4F+pPXa4DHomQ30C3pshzHNzOzHPIEgIPASklzJM0CVgMLgZ+PiJcAkt/zk/V7gMoJcI8laWZmNg0yDwYXEc9JegjYBbwC7AfO1tgkbe7DC+ZIlLSeUhURvb29WbM3bSY6V7CZ2XTJ9RA4Ih6OiOURsRI4CTwP/LBctZP8fjlZ/RilO4SyBcDxlH1ujYi+iOibN29enuxNufJcwUPDIwSvzxU8sG9ourNmZnaBvK2A5ie/e4G1wDZgB7AuWWUd8FfJ6x3AHUlroBXA6XJVUafIMlewmdl0yTsfwJOS5gCjwIaIOCXpQeBxSR8GXgTen6z7FKXnBEeAM8CdOY/dcrLMFWxmNl1yBYCIuCEl7cfAe1LSA9iQ53itLutcwWZm08E9gZvIcwWbWTvxlJBN5LmCzaydOAA0mecKNrN24SogM7OCcgAwMyuoQlcBZem1656+ZtYpChsAyr12yx23yr12gaoX9CzbmJm1qsJWAWXpteuevmbWSQobALL02nVPXzPrJIUNANV659bqtZtlGzOzVlXYAJCl1657+ppZJynsQ+AsvXbd09fMOolKY7S1pr6+vhgcHJzubJiZtRVJeyOir956ha0CMjMrOgcAM7OCcgAwMysoBwAzs4JyADAzKygHADOzgnIAMDMrKAcAM7OCcgAwMyuowg4FMZ4nejGzonEAwBO9mFkxuQoIT/RiZsWUKwBI+oikQ5IOStom6WckvUfSNyV9S9LfS3pzsu7Fkj4v6YikPZIWNaMAzeCJXsysiDIHAEk9wH8F+iLiGqALuA34LPCBiHgH8JfAfckmHwZORcSbgT8BHsqT8WbyRC9mVkR5q4DeAMyU9AZgFnAcCOBNyfJLkjSANcCjyesvAO+RpJzHbwpP9GJmRZT5IXBEDEn6I+BFYAT4ckR8WdJvAU9JGgH+FViRbNIDHE22PSvpNDAH+FHlfiWtB9YD9Pb2Zs3ehHiiFzMroswBQNJsSt/qrwCGgSckfRBYC6yOiD2SNgJ/DPwWkPZt/4LZaCJiK7AVShPCZM3fRPUv6/EF38wKJU8V0H8AvhcRJyJiFNgOXA8sjYg9yTqfB96VvD4GLARIqowuAU7mOL6ZmeWQJwC8CKyQNCupy38P8G3gEklvSdb5VeC55PUOYF3y+n3AV6OV56M0M+tweZ4B7JH0BeCbwFlgH6Wqm2PAk5JeA04BH0o2eRj4M0lHKH3zvy1Pxs3MLB9PCm9m1mE8KbyZmdXU0ncAkk4A35/ufNQwl3HNWDuMy9feXL72lqd8vxgR8+qt1NIBoNVJGmzkNqtduXztzeVrb1NRPlcBmZkVlAOAmVlBOQDks3W6MzDJXL725vK1t0kvn58BmJkVlO8AzMwKygGgCkk3SzqcTGBzb8ryX5T0FUnPSvo7SQsqlo0lE+J8S9KOqc15fZIekfSypINVlkvSp5OyPytpecWydZKeT37WpW0/3XKWr6XPHTRUvn8j6R8lvSrpo+OW1fy7bgU5y/eCpAPJ+WvJXqQNlO8Dyd/ls5L+QdLSimXNPX8R4Z9xP5Qmt/kOcCXwRmA/8NZx6zwBrEte3wj8WcWyV6a7DHXKtxJYDhyssnw18CVKI7iuAPYk6ZcC301+z05ez57u8jSrfO1w7hos33zgncADwEcr0uv+XbfCT9byJcteAOZOdxlylu9d5f8r4L0V/39NP3++A0h3LXAkIr4bET8FPkdp6OtKbwW+krz+fynLW1ZEfI3aI7GuAR6Lkt1At6TLgFXArog4GRGngF3AzZOf44nJUb62UK98EfFyRPwTMDpuUSN/19MuR/naQgPl+4fk/wtgN1CuXWj6+XMASHdu8prEsSSt0n7gPyWvbwV+TtKc5P3PSBqUtFtS/+RmdVJUK38jn0s7qFWOdj93tXTK+aslgC9L2ptMLtXuPkzpbhUm4fxlHg20wzUyec1Hgc9I+k3ga8AQpVFRAXoj4rikK4GvSjoQEd+ZtNw2X7XyNzSpTxuoVY52P3e1dMr5q+X65PzNB3ZJ+ufkG3fbkfQrlALAvysnpayW6/z5DiDduclrEgt4fW5jACLieESsjYhlwMeStNPlZcnv7wJ/Byybgjw3U7Xy1/1c2kTVcnTAuaulU85fVRXn72Xgi5SqTdqOpLcD/wdYExE/TpKbfv4cANL9E7BY0hWS3khp7oLzWoRImiup/PltAh5J0mdLuri8DqVZ0r49ZTlvjh3AHUlrmRXA6Yh4CdgJ3JSUcTZwU5LWblLL1yHnrpa6f9ftTNLPSvq58mtKf5+pLW1amaReSjMs/ueI+JeKRU0/f64CShGlSev/C6WLWxfwSEQckvSHwGBE7ADeDWyWFJSqgDYkm18N/G+VJsS5CHgwIlrqIiJpG6X8z5V0DPg4MAMgIv4X8BSlljJHgDPAncmyk5L+J6U/RIA/jIiWm9Yza/log3MH9csn6ReAQeBNwGuS7qHUWuRf0/6up6MMtWQtH6XRM78oCUrXtr+MiKenvgS1NfD3+T+AOcCfJmU5GxF91a5LufKSNC8yM7OCcRWQmVlBOQCYmRWUA4CZWUE5AJiZFZQDgJlZQTkAmJkVlAOAmVlBOQCYmRXU/wfZf2UT++OOxAAAAABJRU5ErkJggg==\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "#, new_pref\n", "plt.scatter(new_pref, bar_matrix[:,-1])" ] }, { "cell_type": "code", "execution_count": 24, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "(array([44., 0., 0., 0., 0., 0., 0., 0., 0., 56.]),\n", " array([-1000., -800., -600., -400., -200., 0., 200., 400.,\n", " 600., 800., 1000.]),\n", " )" ] }, "execution_count": 24, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAD8CAYAAABn919SAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAD0JJREFUeJzt3X+sZGddx/H3x12Kyg/b0ttm01K3NSuh/9CWm6amQrSFUn7IVgVSYmSjTTYaSCBoZLGJwcQ/Wg0/YjSS1TYsBmjLj6YNoLCurcQECrtlKa1L2W1dcO26u0ArJZrqwtc/5rlmur13Z+bemXu3D+9XMplznnnOnu8+587nnjlnzrmpKiRJz3w/sdYFSJKmw0CXpE4Y6JLUCQNdkjphoEtSJwx0SeqEgS5JnTDQJakTBrokdWL9aq7srLPOqo0bN67mKiXpGW/Pnj3fqaq5Uf1WNdA3btzI7t27V3OVkvSMl+Rb4/TzkIskdcJAl6ROGOiS1AkDXZI6YaBLUicMdEnqhIEuSZ0w0CWpEwa6JHViVa8UlaS1tHHbZ9ZkvQdvfO2qrMc9dEnqhIEuSZ0w0CWpEwa6JHXCQJekThjoktQJA12SOmGgS1InDHRJ6oSBLkmdMNAlqRMGuiR1wkCXpE4Y6JLUCQNdkjphoEtSJwx0SerEWH+xKMlB4Angh8DxqppPciZwG7AROAi8qaoem02ZkqRRJtlD/+Wquriq5tv8NmBXVW0CdrV5SdIaWckhl83Ajja9A7h25eVIkpZr3EAv4PNJ9iTZ2trOqarDAO357FkUKEkaz1jH0IErqurRJGcDO5N8Y9wVtF8AWwHOP//8ZZQoSRrHWHvoVfVoez4K3AFcBhxJsgGgPR9dYtntVTVfVfNzc3PTqVqS9DQjAz3Jc5I8b2EauBp4ALgL2NK6bQHunFWRkqTRxjnkcg5wR5KF/h+tqr9P8hXg9iTXA98G3ji7MiVJo4wM9Kp6BHjJIu3fBa6aRVGSpMl5pagkdcJAl6ROGOiS1AkDXZI6YaBLUicMdEnqhIEuSZ0w0CWpEwa6JHXCQJekThjoktQJA12SOmGgS1InDHRJ6oSBLkmdMNAlqRMGuiR1Ypw/QXdK2LjtM2uy3oM3vnZN1itJk3IPXZI6YaBLUicMdEnqhIEuSZ0w0CWpEwa6JHXCQJekThjoktQJA12SOmGgS1Inxg70JOuSfDXJp9v8BUnuTbI/yW1JTptdmZKkUSbZQ387sG9o/ibg/VW1CXgMuH6ahUmSJjNWoCc5D3gt8DdtPsCVwCdalx3AtbMoUJI0nnH30D8A/AHwozb/AuDxqjre5g8B5065NknSBEYGepLXAUeras9w8yJda4nltybZnWT3sWPHllmmJGmUcfbQrwBen+QgcCuDQy0fAE5PsnA/9fOARxdbuKq2V9V8Vc3Pzc1NoWRJ0mJGBnpVvbuqzquqjcB1wD9W1W8AdwNvaN22AHfOrEpJ0kgr+R76u4B3JjnA4Jj6zdMpSZK0HBP9Cbqquge4p00/Alw2/ZIkScvhlaKS1AkDXZI6YaBLUicMdEnqhIEuSZ0w0CWpEwa6JHXCQJekThjoktQJA12SOmGgS1InDHRJ6oSBLkmdMNAlqRMGuiR1wkCXpE4Y6JLUCQNdkjphoEtSJwx0SeqEgS5JnTDQJakTBrokdcJAl6ROGOiS1AkDXZI6YaBLUicMdEnqxMhAT/KTSb6c5GtJHkzyx639giT3Jtmf5LYkp82+XEnSUsbZQ38SuLKqXgJcDFyT5HLgJuD9VbUJeAy4fnZlSpJGGRnoNfCDNvus9ijgSuATrX0HcO1MKpQkjWWsY+hJ1iXZCxwFdgIPA49X1fHW5RBw7mxKlCSNY6xAr6ofVtXFwHnAZcCLF+u22LJJtibZnWT3sWPHll+pJOmkJvqWS1U9DtwDXA6cnmR9e+k84NElltleVfNVNT83N7eSWiVJJzHOt1zmkpzepn8KeAWwD7gbeEPrtgW4c1ZFSpJGWz+6CxuAHUnWMfgFcHtVfTrJvwC3JvkT4KvAzTOsU5I0wshAr6r7gUsWaX+EwfF0SdIpwCtFJakTBrokdcJAl6ROGOiS1AkDXZI6YaBLUicMdEnqhIEuSZ0w0CWpEwa6JHXCQJekThjoktQJA12SOmGgS1InDHRJ6oSBLkmdMNAlqRMGuiR1wkCXpE4Y6JLUCQNdkjphoEtSJwx0SeqEgS5JnTDQJakTBrokdcJAl6ROGOiS1ImRgZ7khUnuTrIvyYNJ3t7az0yyM8n+9nzG7MuVJC1lnD3048DvVdWLgcuBtya5CNgG7KqqTcCuNi9JWiMjA72qDlfVfW36CWAfcC6wGdjRuu0Arp1VkZKk0SY6hp5kI3AJcC9wTlUdhkHoA2dPuzhJ0vjGDvQkzwU+Cbyjqr4/wXJbk+xOsvvYsWPLqVGSNIaxAj3JsxiE+Ueq6lOt+UiSDe31DcDRxZatqu1VNV9V83Nzc9OoWZK0iHG+5RLgZmBfVb1v6KW7gC1tegtw5/TLkySNa/0Yfa4AfhP4epK9re0PgRuB25NcD3wbeONsSpQkjWNkoFfVPwNZ4uWrpluOJGm5vFJUkjphoEtSJwx0SeqEgS5JnTDQJakTBrokdcJAl6ROGOiS1AkDXZI6YaBLUicMdEnqhIEuSZ0w0CWpEwa6JHXCQJekThjoktQJA12SOmGgS1InDHRJ6oSBLkmdMNAlqRMGuiR1wkCXpE4Y6JLUCQNdkjphoEtSJwx0SeqEgS5JnRgZ6EluSXI0yQNDbWcm2Zlkf3s+Y7ZlSpJGGWcP/UPANSe0bQN2VdUmYFeblyStoZGBXlVfAL53QvNmYEeb3gFcO+W6JEkTWu4x9HOq6jBAez57eiVJkpZj5idFk2xNsjvJ7mPHjs16dZL0Y2u5gX4kyQaA9nx0qY5Vtb2q5qtqfm5ubpmrkySNstxAvwvY0qa3AHdOpxxJ0nKN87XFjwFfBF6U5FCS64EbgVcm2Q+8ss1LktbQ+lEdqurNS7x01ZRrkSStgFeKSlInDHRJ6oSBLkmdMNAlqRMGuiR1wkCXpE4Y6JLUCQNdkjphoEtSJwx0SeqEgS5JnTDQJakTBrokdcJAl6ROGOiS1AkDXZI6YaBLUicMdEnqhIEuSZ0w0CWpEwa6JHXCQJekThjoktQJA12SOmGgS1InDHRJ6oSBLkmdMNAlqRMrCvQk1yR5KMmBJNumVZQkaXLLDvQk64C/BF4NXAS8OclF0ypMkjSZleyhXwYcqKpHqup/gFuBzdMpS5I0qZUE+rnAvw3NH2ptkqQ1sH4Fy2aRtnpap2QrsLXN/iDJQ8tc31nAd5a57LLlppFd1qSuMVjXZKxrMtY1gdy04rp+dpxOKwn0Q8ALh+bPAx49sVNVbQe2r2A9ACTZXVXzK/13ps26JmNdk7Guyfy417WSQy5fATYluSDJacB1wF3TKUuSNKll76FX1fEkbwM+B6wDbqmqB6dWmSRpIis55EJVfRb47JRqGWXFh21mxLomY12Tsa7J/FjXlaqnnceUJD0Deem/JHXilAj0JG9M8mCSHyWZP+G1d7dbCzyU5FVD7YvedqCdpL03yf4kt7UTttOq87Yke9vjYJK9rX1jkv8eeu2DQ8u8NMnXW51/nmSxr3uutK73JPn3ofW/Zui1icZvynX9WZJvJLk/yR1JTm/tazpei9S5ZrewSPLCJHcn2dfeA29v7RNv0xnUdrBti71Jdre2M5PsbO+vnUnOaO1p2+tA296XzqimFw2Nyd4k30/yjrUYryS3JDma5IGhtonHJ8mW1n9/ki0rKqqq1vwBvBh4EXAPMD/UfhHwNeDZwAXAwwxOwK5r0xcCp7U+F7Vlbgeua9MfBH53RjW/F/ijNr0ReGCJfl8GfoHB9/b/Dnj1DGp5D/D7i7RPPH5TrutqYH2bvgm46VQYrxPWtypjcZL1bwAubdPPA77ZtttE23RGtR0Ezjqh7U+BbW1629A2fU3bXgEuB+5dhbFbB/wHg+9or/p4AS8HLh3+WZ50fIAzgUfa8xlt+ozl1nRK7KFX1b6qWuyCo83ArVX1ZFX9K3CAwS0HFr3tQNubuxL4RFt+B3DttOtt63kT8LER/TYAz6+qL9Zg6314FvWcxETjN+2VV9Xnq+p4m/0Sg2sVlrRG47Wmt7CoqsNVdV+bfgLYx8mvuF5qm66WzQzeV/DU99dm4MM18CXg9LY9Z+kq4OGq+tZJ+sxsvKrqC8D3FlnfJOPzKmBnVX2vqh4DdgLXLLemUyLQT2Kp2wss1f4C4PGhEJnV7QheBhypqv1DbRck+WqSf0rysqH6Dy1S5yy8rX2Uu2XhYx6Tj98s/TaDPZQFaz1eC06ZW1gk2QhcAtzbmibZprNQwOeT7Mngim+Ac6rqMAx+GQFnr0FdC67jqTtVaz1eMPn4TLW+VQv0JP+Q5IFFHifbG1rq9gKTtk+7zjfz1B+kw8D5VXUJ8E7go0meP416xqzrr4CfAy5utbx3YbEl1r9adS30uQE4DnykNc18vCb5L6zBOp9eRPJc4JPAO6rq+0y+TWfhiqq6lMEdVd+a5OUn6buq45jBubHXAx9vTafCeJ3MzN+LsMLvoU+iql6xjMVOdnuBxdq/w+CjzPq2l77o7QhWUmeS9cCvAS8dWuZJ4Mk2vSfJw8DPt/qHDzNMXM+4dQ3V99fAp9vspOM39braSZ7XAVe1wyirMl4TGOsWFrOU5FkMwvwjVfUpgKo6MvT6uNt0qqrq0fZ8NMkdDA5VHEmyoaoOt0MGR1e7rubVwH0L43QqjFcz6fgcAn7phPZ7lrvyU/2Qy13AdUmeneQCYBODk2aL3nagBcbdwBva8luAO6dc0yuAb1TV/x8aSDKXwf3hSXJhq/OR9pHriSSXt+Pub5lBPQvHnhf8KrBw1n2i8ZtBXdcA7wJeX1X/NdS+puN1gjW9hUX7f94M7Kuq9w21T7pNp13Xc5I8b2GawQnuB9r6F76JMfz+ugt4S/s2x+XAfy4cepiRp3xKXuvxGjLp+HwOuDrJGe0w0dWtbXmmcbZ3pQ8GG+AQg722I8Dnhl67gcGZ6YcY+sYDg7PG32yv3TDUfiGDDXaAwcexZ0+51g8Bv3NC268DDzI4m34f8CtDr80z+OF6GPgL2sVcU67pb4GvA/e3H5wNyx2/Kdd1gMHxwb3t8cFTYbwWqXPmY3GSdf8ig4/Y9w+N02uWs02nXNeFbft8rW2rG1r7C4BdwP72fGZrD4M/ePNwq3t+FnW1df008F3gZ1byHphCHR9jcHjnfxnk1/XLGR8G55cOtMdvraQmrxSVpE6c6odcJEljMtAlqRMGuiR1wkCXpE4Y6JLUCQNdkjphoEtSJwx0SerE/wFctYWxulWKVwAAAABJRU5ErkJggg==\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plt.hist(bar_matrix[:,-1])" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.6.4" } }, "nbformat": 4, "nbformat_minor": 2 }