plot_figures.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "74260\n"
     ]
    }
   ],
   "source": [
    "filename = '~/data/patient_report/all_patient_data.csv'\n",
    "df = pd.read_csv(filename)\n",
    "df['Primary / secondary category'].fillna(' ',inplace=True)\n",
    "df['Full action description'].fillna(' ',inplace=True)\n",
    "\n",
    "df2 = df[df['Primary / secondary category'] != ' ']\n",
    "\n",
    "df2 = df2[df2['Full action description'] != ' ']\n",
    "\n",
    "print(len(df2))\n",
    "\n",
    "subject = df2['Action subject'].fillna(' ').to_list()\n",
    "full_text = df2['Full action description'].fillna(' ').to_list()\n",
    "label = df2['Primary / secondary category'].fillna(' ').to_list()\n",
    "solution = df2['Resolution response'].fillna(' ').to_list()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "primary_list = []\n",
    "second_list = []\n",
    "for l in label:\n",
    "    split = l.split('-')\n",
    "    pri = split[0]\n",
    "    sec = split[1]\n",
    "    primary_list.append(pri)\n",
    "    second_list.append(sec)\n",
    "df2['Primary'] = primary_list\n",
    "df2['Secondary'] = second_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "74260"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(primary_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "26\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1280x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Import Data\n",
    "#df = pd.read_csv(\"https://github.com/selva86/datasets/raw/master/mpg_ggplot2.csv\")\n",
    "df = df2\n",
    "# Prepare data\n",
    "x_var = 'Primary'\n",
    "groupby_var = 'Secondary'\n",
    "df_agg = df.loc[:, [x_var, groupby_var]].groupby(groupby_var)\n",
    "vals = [df[x_var].values.tolist() for i, df in df_agg]\n",
    "\n",
    "# Draw\n",
    "plt.figure(figsize=(16,9), dpi= 80)\n",
    "len_val = 13\n",
    "color_per_primary = []\n",
    "for i in [13,3,6,4]:\n",
    "    for j in range(i):\n",
    "        color_per_primary.append(plt.cm.Spectral(j/float(i-1)))\n",
    "print(len(color_per_primary))\n",
    "n, bins, patches = plt.hist(vals, df[x_var].unique().__len__(), stacked=True, density=False, color=color_per_primary,rwidth=0.5)\n",
    "\n",
    "# Decoration\n",
    "#plt.legend({group:col for group, col in zip(np.unique(df[groupby_var]).tolist(), colors[:len(vals)])})\n",
    "plt.title(\"Stacked Histogram of {} colored by {}\".format(x_var,groupby_var), fontsize=22)\n",
    "plt.xlabel(x_var)\n",
    "plt.ylabel(\"Frequency\")\n",
    "plt.ylim(0, 70000)\n",
    "plt.xticks(ticks=bins+0.25, labels=['Delivery of care ', 'Access','Finance ', 'Environment'], rotation=90, horizontalalignment='left')\n",
    "#plt.xticks(ticks=bins, labels=np.unique(df[x_var]).tolist(), rotation=90, horizontalalignment='left')\n",
    "plt.show()\n",
    "#sort the secondary by porpotion\n",
    "#label with abbr, arrows for the narrow "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Access ', 'Delivery of care ', 'Environment ', 'Finance ']"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.unique(df[x_var]).tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[' Accommodation',\n",
       " ' Adaptive aids / equipment',\n",
       " ' Availability / arranging services',\n",
       " ' Billing',\n",
       " ' Care plan',\n",
       " ' Cleanliness',\n",
       " ' Communication style',\n",
       " ' Diagnosis / assessment',\n",
       " ' Discharge',\n",
       " ' EMS Billing ',\n",
       " ' Emotional support',\n",
       " ' Food',\n",
       " ' Funding',\n",
       " ' Governance / policies & procedures',\n",
       " ' Health information / confidentiality',\n",
       " ' Information sharing / participation',\n",
       " ' Maintenance / upkeep',\n",
       " ' Parking',\n",
       " ' Personal property',\n",
       " ' Physical comfort',\n",
       " ' Postponement',\n",
       " ' Practice standards',\n",
       " ' Preferred accommodation',\n",
       " ' Request for Compensation / Reimbursement',\n",
       " ' Services coordination',\n",
       " ' Wait times']"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.unique(df[groupby_var]).tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "\n",
    "def removeOutliers(x, outlierConstant):\n",
    "    a = np.array(x)\n",
    "    upper_quartile = np.percentile(a, 75)\n",
    "    print('upper_quartile: ' + str(upper_quartile))\n",
    "    lower_quartile = np.percentile(a, 25)\n",
    "    print('lower_quartile: ' + str(lower_quartile))\n",
    "    IQR = (upper_quartile - lower_quartile) * outlierConstant\n",
    "    quartileSet = (lower_quartile - IQR, upper_quartile + IQR)\n",
    "    resultList = []\n",
    "    for y in a.tolist():\n",
    "        if y >= quartileSet[0] and y <= quartileSet[1]:\n",
    "            resultList.append(y)\n",
    "    return resultList"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "129.54729329383247\n",
      "24787.64049159645\n"
     ]
    }
   ],
   "source": [
    "print(np.mean(len_statis))\n",
    "print(np.var(len_statis))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "upper_quartile: 173.0\n",
      "lower_quartile: 30.0\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import math\n",
    "from matplotlib import pyplot as plt\n",
    "total_list_clean = removeOutliers(len_statis, 2)\n",
    "data2 = total_list_clean\n",
    "bins = np.linspace(math.ceil(min(data2)), \n",
    "                   math.floor(max(data2)),\n",
    "                   40) # fixed number of bins\n",
    "\n",
    "plt.xlim([min(data2)-5, max(data2)+5])\n",
    "plt.hist(data2, bins=bins, alpha=0.5)\n",
    "plt.show()\n",
    "# LDA topics of docs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "upper_quartile: 173.0\n",
      "lower_quartile: 30.0\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import math\n",
    "from matplotlib import pyplot as plt\n",
    "total_list_clean = removeOutliers(len_statis, 2)\n",
    "data2 = len_statis\n",
    "bins = np.linspace(math.ceil(min(data2)), \n",
    "                   math.floor(max(data2)),\n",
    "                   40) # fixed number of bins\n",
    "\n",
    "plt.xlim([min(data2)-5, max(data2)+5])\n",
    "plt.hist(data2, bins=bins, alpha=0.5)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "import numpy as np\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn import metrics\n",
    "from scipy.sparse import vstack\n",
    "import spacy\n",
    "import tools\n",
    "import imp\n",
    "import scipy\n",
    "import time\n",
    "from sklearn.decomposition import online_lda\n",
    "from sklearn.linear_model import LogisticRegression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "sw = tools.stop_word()\n",
    "FIXED_SEED = 0\n",
    "lemma_text = tools.read_lemma_data()\n",
    "imp.reload(tools)\n",
    "Cvectorizer = CountVectorizer(max_df = 0.5,min_df=10,stop_words=sw,token_pattern=r\"(?u)\\b\\d?[a-z]{2,}\\d?\\b\",ngram_range=(1,1))\n",
    "Cvectorizer.fit_transform(lemma_text)\n",
    "total_vec,myCViterator = tools.total_vect_cv_prepare(vectorizer=Cvectorizer,dummy_document = 'dummy_doc_1015_tfidf_exclusive')\n",
    "Count_vec = total_vec[:-4]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Topic 26: 0.1058 \n",
      "Topic 22: 0.0882 \n",
      "Topic 33: 0.0880 \n",
      "Topic 32: 0.0752 \n",
      "Topic 21: 0.0720 \n",
      "Topic 9: 0.0492 \n",
      "Topic 15: 0.0485 \n",
      "Topic 1: 0.0427 \n",
      "Topic 34: 0.0408 \n",
      "Topic 27: 0.0388 \n",
      "Topic 11: 0.0358 \n",
      "Topic 16: 0.0352 \n",
      "Topic 17: 0.0338 \n",
      "Topic 18: 0.0314 \n",
      "Topic 29: 0.0311 \n",
      "Topic 19: 0.0307 \n",
      "Topic 31: 0.0274 \n",
      "Topic 25: 0.0211 \n",
      "Topic 30: 0.0156 \n",
      "Topic 24: 0.0105 \n",
      "Topic 28: 0.0096 \n",
      "Topic 36: 0.0076 \n",
      "Topic 2: 0.0059 \n",
      "Topic 39: 0.0054 \n",
      "Topic 8: 0.0045 \n",
      "Topic 35: 0.0044 \n",
      "Topic 7: 0.0042 \n",
      "Topic 3: 0.0041 \n",
      "Topic 20: 0.0037 \n",
      "Topic 37: 0.0036 \n",
      "Topic 6: 0.0034 \n",
      "Topic 10: 0.0033 \n",
      "Topic 13: 0.0030 \n",
      "Topic 12: 0.0029 \n",
      "Topic 14: 0.0026 \n",
      "Topic 0: 0.0023 \n",
      "Topic 23: 0.0022 \n",
      "Topic 38: 0.0020 \n",
      "Topic 5: 0.0018 \n",
      "Topic 4: 0.0017 \n"
     ]
    }
   ],
   "source": [
    "from joblib import load\n",
    "lda = load('final_lda')\n",
    "topics = lda.transform(Count_vec)\n",
    "topic_count = np.sum(topics,axis=0)\n",
    "topic_count = topic_count / np.sum(topic_count)\n",
    "log = 'Topic {}: {:0.4f} '\n",
    "for i in np.argsort(-topic_count):\n",
    "    print(log.format(i,topic_count[i]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 3600x2400 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "topic_count.sort()\n",
    "y_pos = np.arange(len(topic_count))\n",
    "plt.figure(dpi= 600)\n",
    "\n",
    "plt.barh(y_pos, topic_count)#, color=['black', 'red', 'green', 'blue', 'cyan'])\n",
    "#plt.yticks(y_pos, y_pos)\n",
    "plt.xlabel('Proportion')\n",
    "plt.ylabel('Topics given by LDA model')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<BarContainer object of 40 artists>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "i=35\n",
    "curr_color = ['royalblue']*40\n",
    "curr_color[i] = 'salmon'\n",
    "plt.xlabel('Proportion of topics')\n",
    "plt.ylabel('Topic')\n",
    "\n",
    "plt.barh(y_pos, topic_count,color=curr_color)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Access : \n",
      "[22  9  1 29 26]\n",
      "[0.05182201 0.09139938 0.0979688  0.16870217 0.32806   ]\n",
      "\n",
      "Delivery of care : \n",
      "[16 27 33 22 21]\n",
      "[0.05650657 0.16575313 0.17731222 0.06356018 0.1395881 ]\n",
      "\n",
      "Environment : \n",
      "[10 33 17 29 25]\n",
      "[0.03261364 0.03312411 0.50245662 0.30039811 0.04388648]\n",
      "\n",
      "Finance : \n",
      "[35 16 34 31 19]\n",
      "[0.03719671 0.05190349 0.37555389 0.26498016 0.1076021 ]\n",
      "Entropy Loss: -8.506989214428922\n",
      "Corr Loss: 0.1529750112814523\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "-10.036739327243446"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "NB_vect = total_vec[-4:]\n",
    "#lda = grid.best_estimator_\n",
    "NB_topic_dist = lda.transform(NB_vect)\n",
    "#report_topic = lda2.transform(count_vec)\n",
    "print('Access : ')\n",
    "to_topic = np.argpartition(NB_topic_dist[0],-5)[-5:]\n",
    "print(to_topic)\n",
    "print(NB_topic_dist[0,to_topic])\n",
    "\n",
    "print('\\nDelivery of care : ')\n",
    "to_topic = np.argpartition(NB_topic_dist[1],-5)[-5:]\n",
    "print(to_topic)\n",
    "print(NB_topic_dist[1,to_topic])\n",
    "\n",
    "print('\\nEnvironment : ')\n",
    "to_topic = np.argpartition(NB_topic_dist[2],-5)[-5:]\n",
    "print(to_topic)\n",
    "print(NB_topic_dist[2,to_topic])\n",
    "\n",
    "print('\\nFinance : ')\n",
    "to_topic = np.argpartition(NB_topic_dist[3],-5)[-5:]\n",
    "print(to_topic)\n",
    "print(NB_topic_dist[3,to_topic])\n",
    "\n",
    "tools.NB_loss_function(0,NB_topic_dist)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "weight_list.reverse()\n",
    "weight_cumsum = np.array(weight_list).cumsum()\n",
    "plt.xticks(rotation=90)\n",
    "plt.figure(dpi= 600)\n",
    "\n",
    "plt.rc('xtick', labelsize=5) \n",
    "word_list.reverse()\n",
    "plt.plot(word_list, weight_cumsum)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "weight_cumsum = np.array(weight_list).cumsum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "with open('index_mapping_list','rb') as f:\n",
    "    index_mapping_list = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "filename = '~/data/patient_report/all_patient_data.csv'\n",
    "df = pd.read_csv(filename)\n",
    "df['Primary / secondary category'].fillna(' ',inplace=True)\n",
    "df['Full action description'].fillna(' ',inplace=True)\n",
    "\n",
    "df2 = df[df['Primary / secondary category'] != ' ']\n",
    "\n",
    "df2 = df2[df2['Full action description'] != ' ']\n",
    "\n",
    "print(len(df2))\n",
    "\n",
    "subject = df2['Action subject'].fillna(' ').to_list()\n",
    "full_text = df2['Full action description'].fillna(' ').to_list()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "lemma_filtered_text = tools.read_filtered_lemma_data()\n",
    "Count_vec2 = Cvectorizer.transform(lemma_filtered_text)\n",
    "imp.reload(tools)\n",
    "tools.show_lda_example(Count_vec2,lda,full_text,index_mapping_list,example_num=15,thres=0.8)\n",
    "#tools.show_lda_example(Count_vec2,lda,lemma_filtered_text,example_num=15,thres=0.8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "tools.show_lda_example(Count_vec2,lda,full_text,index_mapping_list,example_num=15,thres=0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "tools.show_lda_example(Count_vec2,lda,full_text,index_mapping_list,example_num=15,thres=0.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Topic 26 appointment, wait, time, clinic, speak, book, clerk, information, contact, explain, phone, office, number, staff, hour, receptionist, doctor, process, schedule, ask, person, leave, procedure, like, cancel, follow, question, appt, date, answer, pcc, advise, booking, happen, request, year, minute, provide, day,\n",
      "\n",
      "Topic 22 care, home, staff, facility, provide, manager, nursing, plan, term, change, case, meeting, dementia, homecare, transition, bath, need, palliative, senior, hc, management, lack, long, support, ltc, caregiver, agent, living, receive, transfer, client, assist, assistance, level, ensure, staffing, live, resident, lodge,\n",
      "\n",
      "Topic 33 nurse, come, triage, ask, leave, staff, room, sit, rude, iv, way, hand, check, comment, talk, unprofessional, shift, desk, nursing, clean, hour, night, treat, manner, look, walk, yell, minute, help, attitude, like, reply, start, wait, male, cry, water, pain, compassion,\n",
      "\n",
      "Topic 32 ed, physician, pain, ct, scan, rdrhc, symptom, emergency, diagnosis, department, chest, diagnose, abdominal, test, assessment, assess, present, treat, attend, hour, doctor, result, severe, discharge, kidney, review, heart, er, history, blood, medical, hospital, send, condition, return, admit, fever, treatment, home,\n",
      "\n",
      "Topic 21 health, centre, emergency, practice, information, standard, department, hospital, worker, social, communication, medical, tooth, style, foothills, south, link, record, dental, community, care, university, lake, dentist, deer, plan, campus, confidentiality, regional, attach, red, healthcare, breach, treatment, uah, mental, sharing, hills, staff,\n",
      "\n",
      "Topic 9 medication, psychiatrist, mental, health, mg, remand, psychiatric, request, prescribe, anxiety, prairie, suicide, psychologist, disorder, depression, ago, form, crisis, suicidal, chronic, grande, addiction, sleep, pain, erc, psychiatry, submit, help, year, prescription, counselor, incarcerate, correctional, centre, hsr, methadone, release, month, dose,\n",
      "\n",
      "Topic 15 antibiotic, injury, ray, infection, left, fracture, wound, leg, orthopedic, foot, right, discharge, arm, follow, fall, home, hip, st, hospital, bone, wrist, eye, sustain, day, ankle, require, assessment, physician, care, shoulder, referral, swollen, august, heal, rehabilitation, week, wecc, stitch, fractured,\n",
      "\n",
      "Topic 1 surgery, surgeon, baby, post, delivery, procedure, surgical, section, breast, day, op, deliver, labour, recovery, biopsy, february, nicu, perform, grey, nuns, feed, pre, cancer, tumor, time, operative, labor, anesthesiologist, week, operating, undergo, complication, obstetrician, operatively, epidural, birth, anesthetic, spinal, bear,\n",
      "\n",
      "Topic 34 discharge, miss, belonging, item, mch, glass, lose, personal, shoe, wallet, allergy, flu, bag, unit, friend, aid, locate, shot, find, allergic, transfer, black, fmc, pharmacy, ms, return, key, clothing, ahe, id, prescription, reaction, clothe, plain, pharmacist, purse, pair, rocky, rmcc,\n",
      "\n",
      "Topic 27 medication, relay, physician, resident, ask, oxygen, death, die, administer, rn, advise, review, morphine, management, drug, communication, deceased, administration, error, discuss, order, partner, manner, provide, discussion, dtr, follow, appropriate, chart, stop, prior, injection, like, jun, question, icu, inappropriate, attend, receive,\n",
      "\n",
      "Topic 11 report, rgh, dialysis, rah, denture, document, forward, outcome, peter, lougheed, unit, relations, royal, lacc, rockyview, general, alexandra, grh, site, pcc, phone, reviewer, feedback, manor, hospitalist, hip, sep, review, renal, questions, lacombe, like, action, secure, discuss, partial, delay, clinic, send,\n",
      "\n",
      "Topic 16 ems, ambulance, crew, transport, paramedic, attendant, female, arrive, claim, driver, metro, scene, drive, personnel, medic, event, house, seizure, elevator, dispatch, male, apparently, fire, emt, police, truck, hospital, car, hit, stretcher, siren, ride, niece, believe, neighbour, lane, think, rude, dispatcher,\n",
      "\n",
      "Topic 17 bed, food, unit, floor, room, meal, chair, bathroom, eat, stretcher, roommate, tray, place, staff, cleanliness, tv, safety, serve, lunch, capacity, breakfast, commode, ocp, space, uah, sleep, fall, bell, ducc, visitor, hospital, use, transfer, housekeeping, privacy, sink, restraint, noise, rest,\n",
      "\n",
      "Topic 18 blood, oncologist, test, cci, lab, fort, result, work, sample, chemotherapy, ultrasound, river, cancer, draw, saskatchewan, chemo, insulin, treatment, april, sugar, diabetic, urine, high, testing, pregnancy, opinion, order, infusion, second, laboratory, level, fscc, pregnant, mcmurray, diabetes, miscarriage, urologist, medical, bloodwork,\n",
      "\n",
      "Topic 29 program, parking, access, service, wheelchair, machine, park, group, building, therapy, zone, available, door, entrance, ticket, outside, spot, main, deny, volunteer, availability, live, use, aish, lot, physio, walk, handicapped, area, doc, unable, central, ot, location, parkade, exit, handicap, stall, limited,\n",
      "\n",
      "Topic 19 private, room, charge, semi, accommodation, sign, request, fee, stay, billing, ward, preferred, waive, isolation, bill, available, unit, form, admit, hospital, night, technician, receive, day, lois, theme, admission, pay, hole, agreement, reverse, rah, adjust, place, blue, discussion, rate, lab, la,\n",
      "\n",
      "Topic 31 bill, pay, invoice, tech, cost, receive, ahs, insurance, card, md, money, therapist, cover, financial, payment, collection, finance, company, credit, billing, dispute, afford, account, coverage, canada, income, receipt, outstanding, ahc, waive, ar, hardship, send, refund, service, visit, receivable, bc, accounts,\n",
      "\n",
      "Topic 25 child, security, parent, stollery, guard, children, smoke, pediatrician, public, school, smoking, aunt, pediatric, policy, cigarette, smoker, ear, hospital, nechc, police, protective, infant, minor, sw, age, enforce, area, stoll, autism, video, phn, vaccination, visitor, services, butt, birth, immunization, property, ban,\n",
      "\n",
      "Topic 30 knee, medicine, hat, physiotherapy, specialist, gi, erc, brace, tylenol, northern, lights, physiotherapist, pain, accord, replacement, regional, mhrh, exercise, session, inmate, express, enema, describe, movement, ip, carewest, vision, constipation, joint, belief, jaw, involved, lump, spine, day, wear, oxycontin, orthotic, splint,\n",
      "\n",
      "Topic 24 stroke, placement, hearing, guardian, ltc, legal, neurologist, aide, speech, grandson, um, abuse, await, westlock, robbins, sl4, neuro, pavilion, slp, egh, rdrh, angiogram, welfare, dgtr, pathologist, loss, slur, rosedale, investigation, battery, sl, waitlist, confused, acute, hanna, cva, team, tia, tc,\n",
      "\n",
      "Topic 28 urgent, file, chumir, sheldon, cardiologist, pcc, close, pursue, prc, wish, norwood, survey, reciev, covenant, records, operations, ucc, uc, paint, columbia, prioritize, british, downtown, reopen, unsatisfied, echocardiogram, director, smchc, richmond, th, corporate, trace, sliver, venipuncture, remedy, rd, john, operational, joseph,\n",
      "\n",
      "Topic 36 catheter, voice, mail, message, law, crutch, insert, brooks, urinary, insertion, foley, common, catheterization, penis, echo, void, catheterize, razor, intake, balloon, trustee, urologist, urethra, indwell, tract, tubing, foreskin, licensed, culture, thousand, suprapubic, vancomycin, urinate, cindy, dispose, audiology, pox, carpet, irrigation,\n",
      "\n",
      "Topic 2 cast, air, throat, ach, boot, sore, fit, strep, nephrologist, employer, splint, ch, swallow, ankle, laptop, break, virus, gnch, slab, pic, barium, donate, swab, aircast, ortho, microwave, nephrology, contrary, crcc, osteoporosis, size, mono, angle, conditioning, pric, tonsil, tight, gland, fitting,\n",
      "\n",
      "Topic 39 line, caller, bruise, nlrhc, form, disappointment, television, photo, ft, woods, rehc, caleo, making, express, kipnes, jb, mill, renfrew, alpha, sunridge, detox, foip, shape, pinch, observed, music, unsupportive, automate, bruising, hide, cls, du, laurier, cable, feature, trusteeship, sweet, electrode, kim,\n",
      "\n",
      "Topic 8 mri, chinook, crh, stocking, kin, reduction, compression, quarantine, regional, diagnostic, fh, server, dislocated, dash, oc, exasperate, lonely, dislocation, privately, reimbursement, altogether, digital, narrow, hopeless, halloween, measles, fitting, isolate, gauge, proceeding, unwanted, wait, complete, urgent, month, shoulder, functional, warmth, timely,\n",
      "\n",
      "Topic 35 grandmother, ring, np, scope, gold, practitioner, stool, granddaughter, grandfather, wedding, sound, grand, ultra, necklace, chain, diamond, sterile, laxative, packing, belt, pendant, grandma, bloody, egccc, kit, diarrhea, grandaughter, reddrc, fcc, sky, softener, copper, circumcision, herpe, standby, color, tail, garment, stomp,\n",
      "\n",
      "Topic 7 walker, officer, peace, camrose, apr, uncle, overflow, foam, bonnyville, peanut, garbage, grove, boyle, spruce, officers, whcc, handcuff, knife, liaison, wetaskawin, anonymous, butter, bashaw, resist, onoway, aged, monie, religion, disagreement, dirt, ground, eps, loan, chaplain, mccauley, agressive, teen, mary, degeneration,\n",
      "\n",
      "Topic 3 plc, swab, operator, contrast, mazankowski, switchboard, housing, lobby, suite, mask, pac, maz, licence, signage, marguerite, masterpiece, stepdaughter, institute, nations, approval, eip, working, attire, emg, terminal, eviction, tighten, racism, precious, rattle, bee, cashier, revoke, ulceration, holistic, guards, offense, african, villa,\n",
      "\n",
      "Topic 20 tube, vehicle, co, bariatric, cpap, nose, motor, km, ng, nasal, apnea, tpn, telehealth, accident, feeding, par, gastric, passenger, everybody, collision, victoria, mvc, landlord, insert, pessary, dislodge, nephrostomy, adhesion, rear, wood, prong, specialty, morbidity, bereavement, damaged, seatbelt, terminology, castor, rec,\n",
      "\n",
      "Topic 37 colonoscopy, website, drcc, valley, drayton, coordination, kaye, healthlink, computer, colon, inefficient, description, screening, advertise, estates, opthalmology, lewis, beleiv, ez, cochrane, summerwood, endocrinology, landing, retirement, mayfair, forcibly, streamline, dire, advertising, residency, fortunately, market, designated, neurologic, rapidly, map, recording, transcribe, center,\n",
      "\n",
      "Topic 6 bladder, appeal, neurosurgeon, radiology, dissatisfaction, permission, arthritis, college, endoscopy, surgeons, aboriginal, rheumatologist, cpsa, gall, physicians, rheumatology, vancouver, polyp, rheumatoid, endometriosis, shunt, camis, banff, overbook, york, racist, instrument, cath, profile, lime, misdiagnosis, ethic, nation, methotrexate, teleconference, scarring, biased, malignant, yellowknife,\n",
      "\n",
      "Topic 10 diet, special, dietician, pco, restriction, dietary, gas, milk, hill, product, als, weigh, celiac, vm, scale, animal, log, lactose, stock, contraindicate, vegetarian, phsycian, cipro, dairy, supplement, dietitian, neice, daugther, protein, cereal, ombudsman, kg, vegan, strict, intolerant, fo, investigator, listing, mayo,\n",
      "\n",
      "Topic 13 shc, free, girlfriend, scent, perfume, gluten, sacred, project, convenient, ipad, religious, ibs, solitary, mclennan, motility, bird, contractor, inefficiency, toll, spiritual, lawn, bead, humane, pray, rainbow, confinement, distribution, gut, tot, forest, neuromuscular, vast, bureaucracy, chapel, mclennon, menopause, society, agecare, juvenile,\n",
      "\n",
      "Topic 12 pcu, bowel, obstruction, fridge, mayerthorpe, southwood, bunk, educator, military, perforation, manning, wants, thicken, miscommunication, choke, prep, hepatology, hemiarthroplasty, sandal, virtual, linger, subcutaneous, pc, dicu, dull, markedly, shaking, spoon, hys, wvhc, supplemental, choking, perforate, fleet, coed, esophageal, bin, clostridium, rah,\n",
      "\n",
      "Topic 14 rash, cream, innisfail, incc, soap, hemorrhoid, itchy, container, skin, cortisone, scabie, dermatologist, ointment, scratch, spread, measle, sm, lotion, pit, eczema, paitent, sunset, thermometer, scalp, topical, diseases, cocnern, canteen, sized, wendy, psoriasis, contagious, genital, miseracordia, torso, patietn, itch, dot, unhooked,\n",
      "\n",
      "Topic 0 mahi, ii, pcm, refusal, fp, queen, broken, faint, complaiant, qe, owner, elizabeth, camp, eehc, unsympathetic, pedway, pact, basket, cutting, crooked, inattentive, enclose, borrow, flutter, transaction, strand, proceedure, suspension, mazenkowski, cabg, locked, kiosk, unassist, qeii, union, suspend, syncope, mailbox, mazankowski,\n",
      "\n",
      "Topic 23 dog, fentanyl, affected, bite, eeg, ob, tubal, ligation, midwife, patch, inspector, gyne, gyn, toxic, cycle, bug, campbell, pateint, rabie, immunocompromis, circulate, midwifery, intensivist, vbac, reassignment, trainer, notation, infestation, strictly, unauthorized, frontline, eve, string, sterilization, bark, environmental, underneath, pushing, bedr,\n",
      "\n",
      "Topic 38 blame, breech, jean, vascular, satisfaction, phlebotomy, abortion, mailing, phlebotomist, compalinant, nod, ass, complaintant, lifeline, supplier, ski, daniel, somebody, cca, samsung, antecubital, army, gruff, prostrate, misunderstand, radway, denie, untie, privacy, buck, brenda, analysis, goodbye, generic, hoodie, vaginally, coma, dnr, flomax,\n",
      "\n",
      "Topic 5 anesthetist, pfic, mirror, po, butterfly, intubation, reserved, rigid, overloaded, million, chipped, uninsured, dayward, oncall, convert, heparin, needle, dental, anesthetize, rhythm, reevaluate, dentist, emphatically, context, laparoscopic, underlying, hysterical, curly, anesthetic, late, pay, calm, relax, undermine, reparative, halfway, anesthesia, improperly, obesity,\n",
      "\n",
      "Topic 4 mate, lloydminster, pca, india, user, meth, oxytocin, lately, al, crystal, anethetist, veterans, fixate, degraded, cursory, invalid, mercy, urinated, navigate, affairs, invasive, production, english, septicemia, naturally, warn, irreversible, sour, emphatically, retaliation, precedence, undermine, stun, pipe, cell, stabbing, chuckle, beautiful, version,\n",
      "\n"
     ]
    }
   ],
   "source": [
    "import imp\n",
    "imp.reload(tools)\n",
    "Count_vec = total_vec[:-4]\n",
    "topics = lda.transform(Count_vec)\n",
    "topic_count = np.sum(topics,axis=0)\n",
    "topic_sort = np.argsort(topic_count)\n",
    "rev_sort = []\n",
    "for i in range(len(topic_sort)-1,-1,-1):\n",
    "    rev_sort.append(topic_sort[i])\n",
    "representative_words,representative_weight = tools.print_topic_word(lda,Cvectorizer,word_num=40,weight_list = rev_sort)"
   ]
  },
  {
   "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}