ON and OFF times (minutes) in the FCC Help Chatroom.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# User chatting time (ON-chatting) in the FCC Help Room (Jan-Jul 2015)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Definitions:\n",
    "* **OFF** from the chatroom if the user waits more than 5 minutes or more to send the next message\n",
    "* **ON** if the user sent messages during the period before being considered off \n",
    "* assume the time for any first message to add 1 minute "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import os, sys\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import pickle\n",
    "import nltk\n",
    "from datetime import datetime, date, timedelta\n",
    "import re, math\n",
    "from collections import Counter\n",
    "\n",
    "#http://stackoverflow.com/questions/15173225/how-to-calculate-cosine-similarity-given-2-sentence-strings-python\n",
    "WORD = re.compile(r'\\w+')\n",
    "\n",
    "def get_cosine(vec1, vec2): \n",
    "     intersection = set(vec1.keys()) & set(vec2.keys())\n",
    "     numerator = sum([vec1[x] * vec2[x] for x in intersection])\n",
    "\n",
    "     sum1 = sum([vec1[x]**2 for x in vec1.keys()])\n",
    "     sum2 = sum([vec2[x]**2 for x in vec2.keys()])\n",
    "     denominator = math.sqrt(sum1) * math.sqrt(sum2)\n",
    "\n",
    "     if not denominator:\n",
    "        return 0.0\n",
    "     else:\n",
    "        return float(numerator) / denominator\n",
    "\n",
    "def text_to_vector(text):\n",
    "     words = WORD.findall(text)\n",
    "     #if len(words) == 0:\n",
    "     #     words = ['EMPTY']\n",
    "     return Counter(words)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After getting the data from disk, the following code (datausers dict):\n",
    "* will visit all the messages per user\n",
    "* it will make a score to evaluate if the current message is similar to itself or any other message after that\n",
    "* if the following message for comparison is beyond the 5 minutes, it will stop the comparison and start evaluating the closest message\n",
    "\n",
    "A internal iteration (usertimes dict) will take the score and evaluate if:\n",
    "* it is the begining of an ON condition\n",
    "* it is the last message before becoming OFF\n",
    "* it is a solitary message"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "USER DATA STARTED...\n",
      "USER DATA COMPLETED...\n"
     ]
    }
   ],
   "source": [
    "directory = \"/your/path/to/data/\"\n",
    "raw = pickle.load(open(directory+\"help.pkl\", \"rb\"))\n",
    "users = set()\n",
    "for u in raw:\n",
    "    if u['fromUser']['username'] not in users:\n",
    "        users.add(u['fromUser']['username'])\n",
    "\n",
    "# for u in users:\n",
    "#     print(u)\n",
    "#     break\n",
    "\n",
    "# for elem in raw:\n",
    "#     if elem['fromUser']['username'] == list(users)[5]:\n",
    "#         username = elem['fromUser']['username']\n",
    "#         break\n",
    "\n",
    "print(\"USER DATA STARTED...\")  \n",
    "datausers = {}\n",
    "usertimes = {}\n",
    "#http://stackoverflow.com/questions/2788871/python-date-difference-in-minutes\n",
    "fmt = '%Y-%m-%dT%H:%M'\n",
    "for username in users:\n",
    "     #CAPTURING TEXTS AND SCORES PER USER\n",
    "     # initialization of username an data collector (messages list)\n",
    "     if username not in list(datausers.keys()):\n",
    "          datausers[username] = []\n",
    "          #print(username)\n",
    "     messages = []\n",
    "     for i, elem in enumerate(raw):\n",
    "         # loop over all messages and if it is from username, capture that text and time\n",
    "         if elem['fromUser']['username'] == username:\n",
    "             #print(\"OUTSIDE \",elem)\n",
    "             text1 = elem['text']\n",
    "             if WORD.search(text1) == None:\n",
    "                text1 = 'BLANKorCHARACTERS'\n",
    "             #texttime = datetime(int(elem['sent'][:4]),int(elem['sent'][5:7]),int(elem['sent'][8:10]),int(elem['sent'][11:13]),int(elem['sent'][14:16]))\n",
    "             texttime = datetime.strptime(elem['sent'][:16],fmt)\n",
    "             # prepare text for calculations and initialise comparisons\n",
    "             vector1 = text_to_vector(text1)\n",
    "             count = 0\n",
    "             cosine = 0\n",
    "             starttime = texttime\n",
    "             for el in raw[i:]:\n",
    "                 # have a look of this and following messages sent by username until there is a gap higher of 5 minutes between the current message and the last one\n",
    "                 # (look if the person has been \"on\" for 5 minutes since this message)\n",
    "                 if el['fromUser']['username'] == username:\n",
    "                     #followingtime = datetime(int(el['sent'][:4]),int(el['sent'][5:7]),int(el['sent'][8:10]),int(el['sent'][11:13]), int(el['sent'][14:16]))\n",
    "                     followingtime = datetime.strptime(el['sent'][:16],fmt)\n",
    "                     if followingtime - starttime > timedelta(minutes=5):\n",
    "                         #print(followingtime - starttime)\n",
    "                         break\n",
    "                     text2 = el['text']\n",
    "                     if WORD.search(text2) == None:\n",
    "                         text2 = 'BLANKorCHARACTERS'\n",
    "                     #similarity scoring: cosine\n",
    "                     vector2 = text_to_vector(text2)\n",
    "                     cosine += get_cosine(vector1, vector2)\n",
    "                     count += 1\n",
    "             #print(count)\n",
    "             # average all scores\n",
    "             if count > 0: cosine = cosine/count\n",
    "             text1_id = elem['id']\n",
    "             # to append to messages list: the current text and id, the last text in the 5 minutes window, the time at which this text was sent, the score\n",
    "             msg = (text1, text1_id, text2, elem['sent'][:16], cosine)\n",
    "             messages.append(msg)\n",
    "             \n",
    "             #CAPTURING ON-OFF CHAT PERIODS\n",
    "             if username not in list(usertimes.keys()):\n",
    "                 usertimes[username] = []\n",
    "             if len(usertimes[username]) == 0:\n",
    "                 #first message sent: \"on\" message\n",
    "                 usertimes[username].append([elem['sent'][:16]])\n",
    "                 find_end = cosine\n",
    "             # the person STILL in the chat room\n",
    "             elif find_end <= .99:\n",
    "                 if cosine > .99:\n",
    "                      #if person \"off\" of chat room, attach to \"on\" message and set long activity to stop (find_end = 1)\n",
    "                      usertimes[username][-1].append(elem['sent'][:16])\n",
    "                      find_end = 1\n",
    "             # it is a request to start chatting\n",
    "             else:\n",
    "                 # last first \"on\": attach to collection as first\n",
    "                 usertimes[username].append([elem['sent'][:16]])\n",
    "                 # if a long \"on\" activity, set to long activity: find_end = 0\n",
    "                 if cosine <= .99:\n",
    "                       find_end = 0\n",
    "               \n",
    "     \n",
    "     #once completed to visit all messages of THAT person, attach to datausers     \n",
    "     datausers[username] = messages\n",
    "     #break\n",
    "     \n",
    "print(\"USER DATA COMPLETED...\")    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following section is to take the usertimes dict and count/calculate differences between beginning and end of ON individual periods when pertinent, to prepare for some descriptives"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "onoff_chatdur = []\n",
    "onoff_chats = 0\n",
    "for username in users:\n",
    "        registers = usertimes[username]\n",
    "        onoff_chats = onoff_chats + len(registers)\n",
    "        for t in registers:\n",
    "                if t == 1:\n",
    "                        onoff_chatdur.append(1)\n",
    "                else:\n",
    "                        timediff = datetime.strptime(t[-1],fmt) - datetime.strptime(t[0],fmt)\n",
    "                        diff_minutes = timediff.seconds/60 + 1\n",
    "                        if diff_minutes >= 1000: print(username, diff_minutes)\n",
    "                        onoff_chatdur.append(diff_minutes)\n",
    "total_min = sum(onoff_chatdur)/len(onoff_chatdur)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Finally, a simple histogram of the data (for all on/off chatting duration, per user). On sessions vs Time (minutes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": [
       "iVBORw0KGgoAAAANSUhEUgAAAX8AAAEACAYAAABbMHZzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\n",
       "AAALEgAACxIB0t1+/AAAGSRJREFUeJzt3X+MVed95/H3J3Yw2GFNabbDr6mMo6FrUtLYJCZtfo27\n",
       "FOEoBatdGbIKQhX1P7Qbt3+kgWq15p+lzkpVg1TZ0rZxDN6GFXUbhBWWMKGZbFZVje1Agj0mQDej\n",
       "MtPM4NYp7ibyLsSf/eM+I46nA3MvXO4dOJ+XNOK5z/lxv+cOfO7Dc869R7aJiIh6eUe3C4iIiM5L\n",
       "+EdE1FDCPyKihhL+ERE1lPCPiKihhH9ERA1NG/6SHpV0QtLLkh4tffMlDUg6JemwpHmV9bdLOi3p\n",
       "pKQ1lf6VZT+nJe26PocTERHNuGL4S/p54DeBDwK/AHxS0nuAbcCA7WXAkfIYScuBDcByYC3whCSV\n",
       "3T0JbLHdB/RJWnsdjiciIpow3cj/3wDP237T9k+AbwK/DqwDdpd1dgMPlfZ6YK/tC7aHgTPAKkkL\n",
       "gbm2j5b19lS2iYiIDpsu/F8GPlqmeW4HPgEsAXpsj5d1xoGe0l4EjFS2HwEWT9E/WvojIqILbr3S\n",
       "QtsnJX0eOAz8CDgO/GTSOpaU74iIiLiBXDH8AWw/BTwFIOk/0xjBj0taYHusTOmcK6uPAr2VzZeU\n",
       "9UdLu9o/OtXz5Y0kIqJ1tjX9Wm/f4Io/wM+UP38WeBW4E/gvwOdK/zbg8dJeTuN/B7OApcDfAirL\n",
       "ngdWAQIOAmsv83yerqZO/wA7ul1Darp5apqpdaWmG7omt7rNtCN/4FlJPw1cALbaPi/pcWCfpC3A\n",
       "MPBwefYhSfuAIeBiWX9iJL8VeBqYAxy0faiJ546IiOugmWmfj03R9zqw+jLr7wR2TtH/ErDiKmqM\n",
       "iIg2yyd8mzPY7QKmMNjtAqYw2O0CpjDY7QIuY7DbBUxhsNsFTGGw2wVMYbDbBbSDLs3KzAyS7FZP\n",
       "XERE1NjV5GZG/hERNZTwj4iooYR/REQNNXOpZ1dIWkzjy+GqDts+2416IiJuJjM2/IF7YP4fw4Pl\n",
       "6yS+dgv8w68DCf+IiGs0k8Mf6Pu/8N/ubLQ/dh6+1d1yIiJuEpnzj4iooYR/REQNJfwjImoo4R8R\n",
       "UUMJ/4iIGkr4R0TUUMI/IqKGEv4RETU0bfhL2i7pFUknJH1Z0m2S5ksakHRK0mFJ8yatf1rSSUlr\n",
       "Kv0ryz5OS9p1vQ4oIiKmd8Xwl3QX8Ahwn+0VwC3ARhr37R2wvQw4Uh4jaTmwgca9fNcCT0ia+I7p\n",
       "J4EttvuAPkmTv7cnIiI6ZLqR/xs07t17u6RbgduBvwfWAbvLOruBh0p7PbDX9gXbw8AZYJWkhcBc\n",
       "20fLensq20RERIddMfzLvXr/EPg7GqH/T7YHgB7b42W1caCntBcBI5VdjACLp+gfLf0REdEFV/xi\n",
       "N0nvAX4HuAs4D/y5pE9X17FtSW29F6SkHcDdcHZ243aZ/e3cfUTEDU1SP9cYjNN9q+cHgL+2/Y/l\n",
       "Cf8S+EVgTNIC22NlSudcWX8U6K1sv4TGiH+0tKv9o5d7Uts7JK2G3nXQf1tLRxQRcZOzPUjlRvKS\n",
       "Hmt1H9PN+Z8EPiRpTjlxuxoYAp4DNpd1NgP7S/sAsFHSLElLgT7gqO0x4A1Jq8p+NlW2iYiIDrvi\n",
       "yN/2dyTtAV4E3gK+DfxXYC6wT9IWYBh4uKw/JGkfjTeIi8BW2xNTQluBp4E5wEHbh9p+NBER0RRd\n",
       "yuaZQVI5jaDVsOpZ+JvqzVz+ve2D3a0wImJmmcjNVrbJJ3wjImoo4R8RUUMJ/4iIGkr4R0TUUMI/\n",
       "IqKGEv4RETWU8I+IqKGEf0REDSX8IyJqKOEfEVFDCf+IiBpK+EdE1FDCPyKihhL+ERE1lPCPiKih\n",
       "hH9ERA1NG/6Sfk7SscrPeUmfkTRf0oCkU5IOS5pX2Wa7pNOSTkpaU+lfKelEWbbreh1URERc2bTh\n",
       "b/t7tu+1fS+wEvgx8BVgGzBgexlwpDxG0nJgA7AcWAs8Ue7bC/AksMV2H9AnaW27DygiIqbX6rTP\n",
       "auCM7bPAOmB36d8NPFTa64G9ti/YHgbOAKskLQTm2j5a1ttT2SYiIjqo1fDfCOwt7R7b46U9DvSU\n",
       "9iJgpLLNCLB4iv7R0h8RER12a7MrSpoF/CrwucnLyh3X23YneEk7gLvh7GwYBPrbteuIiBuepH6u\n",
       "MRibDn/gQeAl26+Vx+OSFtgeK1M650r/KNBb2W4JjRH/aGlX+0eneiLbOyStht510H9bCzVGRNz0\n",
       "bA/SGBkDIOmxVvfRyrTPp7g05QNwANhc2puB/ZX+jZJmSVoK9AFHbY8Bb0haVU4Ab6psExERHdTU\n",
       "yF/SHTRO9j5S6X4c2CdpCzAMPAxge0jSPmAIuAhstT0xJbQVeBqYAxy0fagNxxARES1qKvxt/wh4\n",
       "96S+12m8IUy1/k5g5xT9LwErWi8zIiLaKZ/wjYiooYR/REQNJfwjImoo4R8RUUMJ/4iIGkr4R0TU\n",
       "UMI/IqKGEv4RETWU8I+IqKGEf0REDSX8IyJqKOEfEVFDCf+IiBpK+EdE1FDCPyKihhL+ERE11FT4\n",
       "S5on6VlJr0oaKrdinC9pQNIpSYclzausv13SaUknJa2p9K+UdKIs23U9DigiIqbX7Mh/F43bLt4D\n",
       "vA84CWwDBmwvA46Ux0haDmwAlgNrgSfKPXsBngS22O4D+iStbduRRERE06YNf0l3Ah+1/RSA7Yu2\n",
       "zwPrgN1ltd3AQ6W9Hthr+4LtYeAMsErSQmCu7aNlvT2VbSIiooOaGfkvBV6T9CVJ35b0J+WG7j22\n",
       "x8s640BPaS8CRirbjwCLp+gfLf0REdFhzdzA/VbgPuC3bb8g6QuUKZ4Jti3J7SpK0g7gbjg7GwaB\n",
       "/nbtOiLihiepn2sMxmbCfwQYsf1CefwssB0Yk7TA9liZ0jlXlo8CvZXtl5R9jJZ2tX90qie0vUPS\n",
       "auhdB/23NX84ERE3P9uDNEbGAEh6rNV9TDvtY3sMOCtpWelaDbwCPAdsLn2bgf2lfQDYKGmWpKVA\n",
       "H3C07OeNcqWQgE2VbSIiooOaGfkD/AfgzyTNAv4W+A3gFmCfpC3AMPAwgO0hSfuAIeAisNX2xJTQ\n",
       "VuBpYA6Nq4cOtek4IiKiBU2Fv+3vAB+cYtHqy6y/E9g5Rf9LwIpWCoyIiPbLJ3wjImoo4R8RUUMJ\n",
       "/4iIGkr4R0TUUMI/IqKGEv4RETWU8I+IqKGEf0REDSX8IyJqKOEfEVFDCf+IiBpK+EdE1FDCPyKi\n",
       "hhL+ERE1lPCPiKihpsJf0rCk70o6Julo6ZsvaUDSKUmHJc2rrL9d0mlJJyWtqfSvlHSiLNvV/sOJ\n",
       "iIhmNDvyN9Bv+17b95e+bcCA7WXAkfIYScuBDcByYC3wRLltI8CTwBbbfUCfpLVtOo6IiGhBK9M+\n",
       "mvR4HbC7tHcDD5X2emCv7Qu2h4EzwKpyk/e5to+W9fZUtomIiA5qZeT/dUkvSnqk9PXYHi/tcaCn\n",
       "tBcBI5VtR4DFU/SPlv6IiOiwZm/g/mHbP5D0r4EBSSerC21bki+zbUREzDDN3sD9B+XP1yR9Bbgf\n",
       "GJe0wPZYmdI5V1YfBXormy+hMeIfLe1q/+hUzydpB3A3nJ0Ng0B/s8cTEXHTk9TPNQbjtNM+km6X\n",
       "NLe07wDWACeAA8DmstpmYH9pHwA2SpolaSnQBxy1PQa8IWlVOQG8qbLN29jeAeyB3jcT/BERb2d7\n",
       "0PaOiZ+r2UczI/8e4Cvlgp1bgT+zfVjSi8A+SVuAYeDhUtSQpH3AEHAR2Gp7YkpoK/A0MAc4aPvQ\n",
       "1RQdERHXZtrwt/194P1T9L8OrL7MNjuBnVP0vwSsaL3MiIhop3zCNyKihhL+ERE1lPCPiKihhH9E\n",
       "RA0l/CMiaijhHxFRQwn/iIgaSvhHRNRQwj8iooYS/hERNZTwj4iooYR/REQNJfwjImoo4R8RUUMJ\n",
       "/4iIGkr4R0TUUFPhL+kWScckPVcez5c0IOmUpMOS5lXW3S7ptKSTktZU+ldKOlGW7Wr/oURERLOa\n",
       "Hfk/SuO2jBO3Y9wGDNheBhwpj5G0HNgALAfWAk+U+/UCPAlssd0H9Ela255DiIiIVjVzA/clwCeA\n",
       "PwUmgnwdsLu0dwMPlfZ6YK/tC7aHgTPAKkkLgbm2j5b19lS2iYiIDmtm5P9HwGeBtyp9PbbHS3uc\n",
       "xk3eARYBI5X1RoDFU/SPlv6IiOiCK4a/pE8C52wf49Ko/21sm0vTQRERcQO4dZrlvwSsk/QJYDbw\n",
       "ryQ9A4xLWmB7rEzpnCvrjwK9le2X0Bjxj5Z2tX/0ck8qaQdwN5ydDYNAf9MHFBFxs5PUzzUG4xVH\n",
       "/rZ/33av7aXARuCvbG8CDgCby2qbgf2lfQDYKGmWpKVAH3DU9hjwhqRV5QTwpso2Uz3vDmAP9L6Z\n",
       "4I+IeDvbg7Z3TPxczT6mG/n/i+csfz4O7JO0BRgGHi4FDUnaR+PKoIvA1jItBLAVeBqYAxy0fehq\n",
       "Co6IiGvXdPjb/ibwzdJ+HVh9mfV2Ajun6H8JWHF1ZUZERDvlE74RETWU8I+IqKGEf0REDSX8IyJq\n",
       "KOEfEVFDCf+IiBpK+EdE1FDCPyKihhL+ERE1lPCPiKihhH9ERA0l/CMiaijhHxFRQwn/iIgaSvhH\n",
       "RNRQwj8iooamu4H7bEnPSzouaUjSH5T++ZIGJJ2SdFjSvMo22yWdlnRS0ppK/0pJJ8qyXdfvkCIi\n",
       "YjrT3cP3TeAB2+8H3gc8IOkjwDZgwPYy4Eh5jKTlwAZgObAWeKLcsxfgSWCL7T6gT9La63FAEREx\n",
       "vWmnfWz/uDRnAbcAPwTWAbtL/27godJeD+y1fcH2MHAGWCVpITDX9tGy3p7KNhER0WHThr+kd0g6\n",
       "DowD37D9CtBje7ysMg70lPYiYKSy+QiweIr+0dIfERFdMO0N3G2/Bbxf0p3A1yQ9MGm5JbmdRUna\n",
       "AdwNZ2fDINDfzt1HRNzQJPVzjcE4bfhPsH1e0leBlcC4pAW2x8qUzrmy2ijQW9lsCY0R/2hpV/tH\n",
       "r/BcOyStht510H9bszVGRNSB7UEaI2MAJD3W6j6mu9rn3RNX8kiaA/wKcAw4AGwuq20G9pf2AWCj\n",
       "pFmSlgJ9wFHbY8AbklaVE8CbKttERESHTTfyXwjslvQOGm8Uz9g+IukYsE/SFmAYeBjA9pCkfcAQ\n",
       "cBHYantiSmgr8DQwBzho+1C7DyYiIppzxfC3fQK4b4r+14HVl9lmJ7Bziv6XgBVXV2ZERLRTPuEb\n",
       "EVFDCf+IiBpK+EdE1FDCPyKihhL+ERE1lPCPiKihhH9ERA0l/CMiaijhHxFRQwn/iIgaSvhHRNRQ\n",
       "wj8iooYS/hERNZTwj4iooYR/REQNJfwjImpo2vCX1CvpG5JekfSypM+U/vmSBiSdknR44naPZdl2\n",
       "SaclnZS0ptK/UtKJsmzX9TmkiIiYTjMj/wvA79p+L/Ah4Lck3QNsAwZsLwOOlMdIWg5sAJYDa4En\n",
       "yn17AZ4EttjuA/okrW3r0URERFOmDX/bY7aPl/b/AV4FFgPrgN1ltd3AQ6W9Hthr+4LtYeAMsErS\n",
       "QmCu7aNlvT2VbSIiooNamvOXdBdwL/A80GN7vCwaB3pKexEwUtlshMabxeT+0dIfEREddsUbuFdJ\n",
       "ehfwF8Cjtv/50kwO2LYkt6soSTuAu+HsbBgE+tu164iIG56kfq4xGJsKf0nvpBH8z9jeX7rHJS2w\n",
       "PVamdM6V/lGgt7L5Ehoj/tHSrvaPTvV8tndIWg2966D/tuYPJyLi5md7kMbIGABJj7W6j2au9hHw\n",
       "RWDI9hcqiw4Am0t7M7C/0r9R0ixJS4E+4KjtMeANSavKPjdVtomIiA5qZuT/YeDTwHclHSt924HH\n",
       "gX2StgDDwMMAtock7QOGgIvAVtsTU0JbgaeBOcBB24fadBwREdGCacPf9v/i8v9DWH2ZbXYCO6fo\n",
       "fwlY0UqBERHRfvmEb0REDTV9tc8M8dVJVxnpCutGRMRl3IAjf5efiIi4Wjdg+EdExLVK+EdE1FDC\n",
       "PyKihhL+ERE1lPCPiKihhH9ERA0l/CMiaijhHxFRQwn/iIgaSvhHRNRQwj8iooYS/hERNXSjfavn\n",
       "20y+b3C+5TMiojnN3MbxKUnjkk5U+uZLGpB0StJhSfMqy7ZLOi3ppKQ1lf6Vkk6UZbvaU77Jt3xG\n",
       "RLSumWmfLwFrJ/VtAwZsLwOOlMdIWg5sAJaXbZ7QpS/gfxLYYrsP6JM0eZ8REdEh04a/7W8BP5zU\n",
       "vQ7YXdq7gYdKez2w1/YF28PAGWCVpIXAXNtHy3p7Ktu0jSRP/LR73xERN5OrPeHbY3u8tMeBntJe\n",
       "BIxU1hsBFk/RP1r62yxTQBERzbjmE7622z7SlrQDuBvOzoZBoP9q9pGTwRFxU5LUz9UEY8XVhv+4\n",
       "pAW2x8qUzrnSPwr0VtZbQmPEP1ra1f7Ry+3c9g5Jq6F3HfTfdnUlVrNfeTOIiJuG7UEaI2MAJD3W\n",
       "6j6udtrnALC5tDcD+yv9GyXNkrQU6AOO2h4D3pC0qpwA3lTZpkNyZVBExIRpR/6S9gIfB94t6Szw\n",
       "n4DHgX2StgDDwMMAtock7QOGgIvAVtsTabsVeBqYAxy0fai9hxIREc3SpWyeGSSV0whaDauehb+5\n",
       "s7HkY+fhW3deGrmLyVM7zS1rPM60T0TcLCZys5Vt8vUOERE1lPCPiKihhH9ERA0l/CMiauiG/lbP\n",
       "a1G97j8nfyOibmo88s81/xFRXzUO/4iI+qrttE9VvvohIuom4Q/8yw+ERUTc3BL+U8jJ4Ii42SX8\n",
       "p3TpayIyJRQRN6OE/7Qu/9XQeSOIiBtVrvZpWS4RjYgbX0b+1yBTQhFxo0r4X5NMCUXEjanj0z6S\n",
       "1ko6Kem0pM91+vmvr0tTQpJ8pZ/u1hkRddfR8Jd0C/DHwFpgOfApSfd0soarM3gV20y+bWR7byNZ\n",
       "buA8o6Sm5s3EulJTc2ZiTVej0yP/+4EztodtXwD+O7C+wzVchcG277GZ/xVM8z+G/rYXde36u13A\n",
       "FPq7XcBl9He7gCn0d7uAKfR3u4Ap9He7gHbodPgvBs5WHo+Uvhq6/P8K3h70l132WKaSIuJqdfqE\n",
       "b4sB9cps+OXzjfZ3bmt/OTNV9V7EV1rW3AnnTr4xSHqs1W2uVGtOnEdcHx29gbukDwE7bK8tj7cD\n",
       "b9n+fGWdjGAjIlrU6kCp0+F/K/A94N8Cfw8cBT5l+9WOFREREZ2d9rF9UdJvA18DbgG+mOCPiOi8\n",
       "jo78IyJiZpgx3+0zEz78JekpSeOSTlT65ksakHRK0mFJ8zpcU6+kb0h6RdLLkj7T7bokzZb0vKTj\n",
       "koYk/UG3a5pU3y2Sjkl6bibUJWlY0ndLTUdnSE3zJD0r6dXyO1zV5b9TP1den4mf85I+MwNep+3l\n",
       "394JSV+WdFu3ayp1PVpqelnSo6WvpbpmRPjPoA9/fanUULUNGLC9DDhSHnfSBeB3bb8X+BDwW+W1\n",
       "6Vpdtt8EHrD9fuB9wAOSPtLNmiZ5FBji0uVQ3a7LQL/te23fP0Nq2gUctH0Pjd/hyW7WZPt75fW5\n",
       "F1gJ/Bj4SjdrknQX8Ahwn+0VNKaqN3azplLXzwO/CXwQ+AXgk5Le03Jdtrv+A/wicKjyeBuwrUu1\n",
       "3AWcqDw+CfSU9gLgZJdfq/3A6plSF3A78ALw3plQE7AE+DrwAPDcTPgdAt8HfnpSX9dqAu4E/vcU\n",
       "/V3//ZXnXgN8q9s1AfNpXKDyUzTOjz4H/Eq3Xyfg3wF/Wnn8H4Hfa7WuGTHyZ2Z/+KvH9nhpjwM9\n",
       "3SqkjETuBZ6ny3VJeoek4+W5v2H7lW7XVPwR8FngrUpft+sy8HVJL0p6ZAbUtBR4TdKXJH1b0p9I\n",
       "uqPLNVVtBPaWdtdqsv068IfA39G4OvGfbA90s6biZeCjZZrnduATNAY9LdU1U8L/hjjr7MZbaldq\n",
       "lfQu4C+AR23/c7frsv2WG9M+S4CPSXqg2zVJ+iRwzvYxLnMz5i79Dj/sxnTGgzSm7T7a5ZpuBe4D\n",
       "nrB9H/AjJk0RdOvvuqRZwK8Cfz55WadrKlMpv0NjNmAR8C5Jn+5mTeU5TwKfBw4D/wM4Dvyk1bpm\n",
       "SviPAr2Vx700Rv8zwbikBQCSFgLnOl2ApHfSCP5nbO+fKXUB2D4PfJXGPG23a/olYJ2k79MYOf6y\n",
       "pGe6XZftH5Q/X6Mxj31/l2saAUZsv1AeP0vjzWBsBvydehB4qbxW0N3X6QPAX9v+R9sXgb+kMUXd\n",
       "9dfJ9lO2P2D748APgVO0+FrNlPB/EeiTdFd5598AHOhyTRMOAJtLezONOfeOkSTgi8CQ7S/MhLok\n",
       "vXviSgJJc2jMgx7rZk0Atn/fdq/tpTSmDv7K9qZu1iXpdklzS/sOGvPZJ7pZk+0x4KykZaVrNfAK\n",
       "jTntrv3+ik9xacoHuvt36iTwIUlzyr/D1TQuJOj66yTpZ8qfPwv8GvBlWn2tOnmiYpqTGA/SOLly\n",
       "BtjepRr20pjb+380zkH8Bo2TPl+n8c56GJjX4Zo+QmP++jiNgD1G44qkrtUFrAC+XWr6LvDZ0t/V\n",
       "12pSjR8HDnS7Lhrz68fLz8sTf7e7/VrRuErkBeA7NEa0d86Amu4A/gGYW+nrdk2/R+ON8QSwG3hn\n",
       "t2sqdf3PUtdxGlfetfxa5UNeERE1NFOmfSIiooMS/hERNZTwj4iooYR/REQNJfwjImoo4R8RUUMJ\n",
       "/4iIGkr4R0TU0P8HTSCJWiWkoNEAAAAASUVORK5CYII=\n"
      ],
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1fb1ba1400>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(onoff_chatdur, bins=90)\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.4.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}