equity (3).ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Import packages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:12:58.550070Z",
     "start_time": "2019-01-24T22:12:55.526382Z"
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd #for data management and wrangling\n",
    "import numpy as np #for calculations and data wrangling\n",
    "from numpy import random\n",
    "import matplotlib.pyplot as plt #for graphs\n",
    "import seaborn as sns #for graphs\n",
    "import sklearn #for the TF-IDF vectorizer, cosine similarity, etc.\n",
    "import nltk #for nlp pre-processing\n",
    "from nltk import word_tokenize\n",
    "# nltk.download()\n",
    "import spacy #for nlp pre-processing\n",
    "import en_core_web_sm  # or any other model you downloaded via spacy download or pip\n",
    "nlp = en_core_web_sm.load()\n",
    "from spacy.lang.en import STOP_WORDS\n",
    "import string\n",
    "from sklearn.cluster import KMeans\n",
    "from sklearn.decomposition import PCA\n",
    "\n",
    "\n",
    "# Enable inline plotting\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Read in xlsx file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:12:58.878738Z",
     "start_time": "2019-01-24T22:12:58.551786Z"
    }
   },
   "outputs": [],
   "source": [
    "#df = pd.read_csv('ALLDATA_MERGE.csv')\n",
    "df = pd.read_excel('ALLDATA_MERGE.xlsx')\n",
    "#df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Get list of unique values in the LinkName column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:12:58.891406Z",
     "start_time": "2019-01-24T22:12:58.886986Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Advancing Equity - November 20, 2018 - DHC Employees',\n",
       "       'Advancing Equity - December 18, 2018 - DHC Employees',\n",
       "       'Advancing Equity - November 9 2018 - DHC Employees',\n",
       "       'Aspiring Manager Program, Session 4: Advancing Equity, Diversity & Inclusion - November 9 2018',\n",
       "       'MDP Track 1 , Session 3: Advancing Equity, Diversity & Inclusion - November 9, 2018',\n",
       "       'Journey Program: Advancing Equity, Diversity & Inclusion - November 9, 2018',\n",
       "       'LeadDOE Session1: Advancing Equity, Diversity & Inclusion - November 15, 2018',\n",
       "       'Advancing Equity – November 15, 2018 – DHC Employees',\n",
       "       'Journey Program: Advancing Equity, Diversity & Inclusion - November 15,2018',\n",
       "       'Advancing Equity – November 30, 2018 – DHC Employees',\n",
       "       'Advancing Equity – December 5, 2018 – DHC Employees',\n",
       "       'Advancing Equity – December 12, 2018 – DHC Employees',\n",
       "       'MDP Track 2, Session 3: Advancing Equity, Diversity & Inclusion - December 12, 2018'],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.LinkName.unique()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### create new column (emp_group) with distinct employee groups based on LinkName values - for easier grouping"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:12:58.924367Z",
     "start_time": "2019-01-24T22:12:58.893505Z"
    }
   },
   "outputs": [],
   "source": [
    "conditions = [\n",
    "    (df['LinkName'] == 'Advancing Equity - November 20, 2018 - DHC Employees') | (df['LinkName'] == 'Advancing Equity - December 18, 2018 - DHC Employees') | (df['LinkName'] == 'Advancing Equity - November 9 2018 - DHC Employees') | (df['LinkName'] == 'Advancing Equity – November 30, 2018 – DHC Employees') | (df['LinkName'] == 'Advancing Equity – December 5, 2018 – DHC Employees') | (df['LinkName'] == 'Advancing Equity – December 12, 2018 – DHC Employees') | (df['LinkName'] == 'Advancing Equity – November 15, 2018 – DHC Employees'),\n",
    "    (df['LinkName'] == 'Aspiring Manager Program, Session 4: Advancing Equity, Diversity & Inclusion - November 9 2018'),\n",
    "    (df['LinkName'] == 'MDP Track 1 , Session 3: Advancing Equity, Diversity & Inclusion - November 9, 2018'),\n",
    "    (df['LinkName'] == 'Journey Program: Advancing Equity, Diversity & Inclusion - November 9, 2018') | (df['LinkName'] == 'Journey Program: Advancing Equity, Diversity & Inclusion - November 15,2018'),\n",
    "    (df['LinkName'] == 'LeadDOE Session1: Advancing Equity, Diversity & Inclusion - November 15, 2018'),\n",
    "    (df['LinkName'] == 'MDP Track 2, Session 3: Advancing Equity, Diversity & Inclusion - December 12, 2018')]\n",
    "choices = ['DHC Employees', 'Aspiring Manager Program', 'MDP Track 1', 'Journey Program', 'LeadDOE', 'MDP Track 2']\n",
    "df['emp_group'] = np.select(conditions, choices)\n",
    "#df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['DHC Employees', 'Aspiring Manager Program', 'MDP Track 1',\n",
       "       'Journey Program', 'LeadDOE', 'MDP Track 2'], dtype=object)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.emp_group.unique()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Get word counts for text cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:12:58.953945Z",
     "start_time": "2019-01-24T22:12:58.926334Z"
    }
   },
   "outputs": [],
   "source": [
    "df['totalwords_takeaway'] = df['Whatwasthemostusefulpieceofinformationorskillthatyoutookawayfrom'].str.split().str.len()\n",
    "df['totalwords_improvement'] = df['Whatwouldhavemadethesessionmorevaluable\\xa0'].str.split().str.len()\n",
    "#df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:03:39.951816Z",
     "start_time": "2019-01-24T22:03:39.949145Z"
    }
   },
   "source": [
    "#### Analyse helpers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:12:58.960168Z",
     "start_time": "2019-01-24T22:12:58.956076Z"
    }
   },
   "outputs": [],
   "source": [
    "def top_tfidf_feats(row, features, top_n=25):\n",
    "    ''' Get top n tfidf values in row and return them with their corresponding feature names.'''\n",
    "    topn_ids = np.argsort(row)[::-1][:top_n]\n",
    "    top_feats = [(features[i], row[i]) for i in topn_ids]\n",
    "    df = pd.DataFrame(top_feats)\n",
    "    df.columns = ['feature', 'tfidf']\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:12:58.965808Z",
     "start_time": "2019-01-24T22:12:58.962566Z"
    }
   },
   "outputs": [],
   "source": [
    "def top_feats_in_doc(Xtr, features, row_id, top_n=25):\n",
    "    ''' Top tfidf features in specific document (matrix row) '''\n",
    "    row = np.squeeze(Xtr[row_id].toarray())\n",
    "    return top_tfidf_feats(row, features, top_n)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:12:58.972148Z",
     "start_time": "2019-01-24T22:12:58.968071Z"
    }
   },
   "outputs": [],
   "source": [
    "def top_mean_feats(Xtr, features, grp_ids=None, min_tfidf=0.1, top_n=25):\n",
    "    ''' Return the top n features that on average are most important amongst documents in rows\n",
    "        indentified by indices in grp_ids. '''\n",
    "    if grp_ids:\n",
    "        D = Xtr[grp_ids].toarray()\n",
    "    else:\n",
    "        D = Xtr.toarray()\n",
    "\n",
    "    D[D < min_tfidf] = 0\n",
    "    tfidf_means = np.mean(D, axis=0)\n",
    "    return top_tfidf_feats(tfidf_means, features, top_n)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:12:58.977898Z",
     "start_time": "2019-01-24T22:12:58.974243Z"
    }
   },
   "outputs": [],
   "source": [
    "def top_feats_by_class(Xtr, y, features, min_tfidf=0.1, top_n=25):\n",
    "    ''' Return a list of dfs, where each df holds top_n features and their mean tfidf value\n",
    "        calculated across documents with the same class label. '''\n",
    "    dfs = []\n",
    "    labels = np.unique(y)\n",
    "    for label in labels:\n",
    "        ids = np.where(y==label)\n",
    "        feats_df = top_mean_feats(Xtr, features, ids, min_tfidf=min_tfidf, top_n=top_n)\n",
    "        feats_df.label = label\n",
    "        dfs.append(feats_df)\n",
    "    return dfs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:12:58.982648Z",
     "start_time": "2019-01-24T22:12:58.980093Z"
    }
   },
   "outputs": [],
   "source": [
    "def wrap_title(title):\n",
    "    tokens = title.split()\n",
    "    if len(tokens) > 2:\n",
    "        tokens.insert(2, \"\\n\")\n",
    "    return \" \".join(tokens)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:12:58.989569Z",
     "start_time": "2019-01-24T22:12:58.984594Z"
    }
   },
   "outputs": [],
   "source": [
    "def plot_tfidf_classfeats_h(dfs):\n",
    "    ''' Plot the data frames returned by the function plot_tfidf_classfeats(). '''\n",
    "    fig = plt.figure(figsize=(17, 10), facecolor=\"w\")\n",
    "    x = np.arange(len(dfs[0]))\n",
    "    for i, df in enumerate(dfs):\n",
    "        ax = fig.add_subplot(2, len(dfs), i+1)\n",
    "        ax.spines[\"top\"].set_visible(False)\n",
    "        ax.spines[\"right\"].set_visible(False)\n",
    "        ax.set_frame_on(False)\n",
    "        ax.get_xaxis().tick_bottom()\n",
    "        ax.get_yaxis().tick_left()\n",
    "        if i == int(len(dfs) /2):\n",
    "            ax.set_xlabel(\"Mean TF-IDF Score\", labelpad=16, fontsize=14)\n",
    "        ax.set_title(wrap_title(str(df.label)))\n",
    "        ax.ticklabel_format(axis='x', style='sci', scilimits=(-2,2))\n",
    "        ax.barh(x, df.tfidf, align='center', color='#3F5D7D')\n",
    "        ax.set_yticks(x)\n",
    "        ax.set_ylim([-1, x[-1]+1])\n",
    "        yticks = ax.set_yticklabels(df.feature)\n",
    "        plt.subplots_adjust(bottom=0.09, right=0.97, left=0.15, top=0.95, wspace=0.52)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:03:39.951816Z",
     "start_time": "2019-01-24T22:03:39.949145Z"
    }
   },
   "source": [
    "#### Analyse takeaway"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Get tokens, lemmas, and pos tags for the key takeaways text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:00.639319Z",
     "start_time": "2019-01-24T22:12:58.991434Z"
    },
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "tokens = []\n",
    "lemma = []\n",
    "pos = []\n",
    "\n",
    "for doc in nlp.pipe(df['Whatwasthemostusefulpieceofinformationorskillthatyoutookawayfrom'].astype('unicode').values, batch_size=30,\n",
    "                        n_threads=3):\n",
    "    if doc.is_parsed:\n",
    "        tokens.append([n.text for n in doc if not n.is_punct and not n.is_stop and not n.is_space])\n",
    "        lemma.append([n.lemma_ for n in doc if not n.is_punct and not n.is_stop and not n.is_space])\n",
    "        pos.append([n.pos_ for n in doc if not n.is_punct and not n.is_stop and not n.is_space])\n",
    "    else:\n",
    "        tokens.append(None)\n",
    "        lemma.append(None)\n",
    "        pos.append(None)\n",
    "\n",
    "df['tokens_takeaway'] = tokens\n",
    "df['lemmas_takeaway'] = lemma\n",
    "df['pos_takeaway'] = pos\n",
    "\n",
    "#df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:00.646247Z",
     "start_time": "2019-01-24T22:13:00.641680Z"
    }
   },
   "outputs": [],
   "source": [
    "df['lemmas_takeaway'] = df['lemmas_takeaway'].apply(lambda x: [t for t in x if t != \"-PRON-\"])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['lemmas_takeaway'] = df['lemmas_takeaway'].apply(lambda x: [t for t in x if t != \"the\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['lemmas_takeaway'] = df['lemmas_takeaway'].apply(lambda x: [t for t in x if t != \"like\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['lemmas_takeaway'] = df['lemmas_takeaway'].apply(lambda x: [t for t in x if t != \"not\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Create new df with just groups, word counts, and lemmas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:00.663465Z",
     "start_time": "2019-01-24T22:13:00.648848Z"
    }
   },
   "outputs": [],
   "source": [
    "takeaway = df[['emp_group',\n",
    "'totalwords_takeaway',\n",
    " 'lemmas_takeaway']]\n",
    "#takeaway.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### drop na values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:00.678964Z",
     "start_time": "2019-01-24T22:13:00.665986Z"
    }
   },
   "outputs": [],
   "source": [
    "takeaway = takeaway.dropna(how='any')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Make a tfidf function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:00.683859Z",
     "start_time": "2019-01-24T22:13:00.681398Z"
    }
   },
   "outputs": [],
   "source": [
    "vect = sklearn.feature_extraction.text.TfidfVectorizer(max_features=11)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Apply tfidf function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:00.698505Z",
     "start_time": "2019-01-24T22:13:00.686454Z"
    }
   },
   "outputs": [],
   "source": [
    "lemmas = takeaway['lemmas_takeaway'].apply(lambda x: ' '.join(x))\n",
    "features = vect.fit_transform(lemmas)\n",
    "\n",
    "feature_names = vect.get_feature_names()\n",
    "dense = features.todense()\n",
    "denselist = dense.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:00.761062Z",
     "start_time": "2019-01-24T22:13:00.700424Z"
    }
   },
   "outputs": [],
   "source": [
    "takeaway['all'] = 'all'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Visualize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:00.776820Z",
     "start_time": "2019-01-24T22:13:00.763462Z"
    }
   },
   "outputs": [],
   "source": [
    "dfs = top_feats_by_class(features, takeaway['emp_group'], feature_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:00.783073Z",
     "start_time": "2019-01-24T22:13:00.778583Z"
    }
   },
   "outputs": [],
   "source": [
    "df_all = top_feats_by_class(features, takeaway['all'], feature_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:01.554307Z",
     "start_time": "2019-01-24T22:13:00.785060Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1224x720 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_tfidf_classfeats_h(dfs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:01.816735Z",
     "start_time": "2019-01-24T22:13:01.556147Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1224x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_tfidf_classfeats_h(df_all)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "_Takeaways:_<br>\n",
    " - The majority of participant responses focused on **bias** as a central theme of the trainings, which aligns with the program’s goal of making employees more aware of implicit and explicit bias. \n",
    " - Many respondents identified their **facilitator** as an important factor in their experience, which is good feedback for our team, as we invest time and resources into training for our facilitators.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T21:19:07.335076Z",
     "start_time": "2019-01-24T21:19:07.332627Z"
    }
   },
   "source": [
    "##### kmeans"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:01.844181Z",
     "start_time": "2019-01-24T22:13:01.818731Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top terms per cluster:\n",
      "Cluster 0:\n",
      " training\n",
      " facilitator\n",
      " work\n",
      " learn\n",
      " white\n",
      " issue\n",
      " bias\n",
      " people\n",
      " racism\n",
      " session\n",
      "Cluster 1:\n",
      " bias\n",
      " facilitator\n",
      " work\n",
      " learn\n",
      " white\n",
      " session\n",
      " training\n",
      " understand\n",
      " racism\n",
      " people\n",
      "Cluster 2:\n",
      " understand\n",
      " bias\n",
      " work\n",
      " racism\n",
      " people\n",
      " issue\n",
      " session\n",
      " learn\n",
      " facilitator\n",
      " white\n",
      "Cluster 3:\n",
      " people\n",
      " issue\n",
      " white\n",
      " facilitator\n",
      " work\n",
      " bias\n",
      " racism\n",
      " understand\n",
      " session\n",
      " learn\n",
      "Cluster 4:\n",
      " racism\n",
      " bias\n",
      " work\n",
      " white\n",
      " issue\n",
      " training\n",
      " learn\n",
      " understand\n",
      " session\n",
      " people\n",
      "Cluster 5:\n",
      " session\n",
      " bias\n",
      " issue\n",
      " people\n",
      " learn\n",
      " facilitator\n",
      " work\n",
      " training\n",
      " white\n",
      " understand\n"
     ]
    }
   ],
   "source": [
    "X = vect.fit_transform(takeaway['lemmas_takeaway'].apply(lambda x: ' '.join(x)))\n",
    "\n",
    "true_k = 6\n",
    "model = KMeans(n_clusters=true_k, init='k-means++', max_iter=100, n_init=1)\n",
    "model.fit(X)\n",
    "\n",
    "print(\"Top terms per cluster:\")\n",
    "order_centroids = model.cluster_centers_.argsort()[:, ::-1]\n",
    "terms = vect.get_feature_names()\n",
    "for i in range(true_k):\n",
    "    print(\"Cluster %d:\" % i),\n",
    "    for ind in order_centroids[i, :10]:\n",
    "        print(' %s' % terms[ind]),\n",
    "    print"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:03:39.951816Z",
     "start_time": "2019-01-24T22:03:39.949145Z"
    }
   },
   "source": [
    "#### Analyse improvement"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Get tokens, lemmas, and pos tags for the improvement suggestions text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:04.527515Z",
     "start_time": "2019-01-24T22:13:01.846157Z"
    }
   },
   "outputs": [],
   "source": [
    "tokens = []\n",
    "lemma = []\n",
    "pos = []\n",
    "\n",
    "for doc in nlp.pipe(df['Whatwouldhavemadethesessionmorevaluable\\xa0'].astype('unicode').values, batch_size=30,\n",
    "                        n_threads=3):\n",
    "    if doc.is_parsed:\n",
    "        tokens.append([n.text for n in doc if not n.is_punct and not n.is_stop and not n.is_space])\n",
    "        lemma.append([n.lemma_ for n in doc if not n.is_punct and not n.is_stop and not n.is_space])\n",
    "        pos.append([n.pos_ for n in doc if not n.is_punct and not n.is_stop and not n.is_space])\n",
    "    else:\n",
    "        tokens.append(None)\n",
    "        lemma.append(None)\n",
    "        pos.append(None)\n",
    "\n",
    "df['tokens_improvement'] = tokens\n",
    "df['lemmas_improvement'] = lemma\n",
    "df['pos_improvement'] = pos\n",
    "\n",
    "#df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:04.534430Z",
     "start_time": "2019-01-24T22:13:04.529765Z"
    }
   },
   "outputs": [],
   "source": [
    "df['lemmas_improvement'] = df['lemmas_improvement'].apply(lambda x: [t for t in x if t != \"-PRON-\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['lemmas_improvement'] = df['lemmas_improvement'].apply(lambda x: [t for t in x if t != \"like\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['lemmas_improvement'] = df['lemmas_improvement'].apply(lambda x: [t for t in x if t != \"the\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['lemmas_improvement'] = df['lemmas_improvement'].apply(lambda x: [t for t in x if t != \"not\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Create new df with just groups, word counts, and lemmas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:04.550253Z",
     "start_time": "2019-01-24T22:13:04.537036Z"
    }
   },
   "outputs": [],
   "source": [
    "improvement = df[['emp_group',\n",
    "'totalwords_improvement',\n",
    " 'lemmas_improvement']]\n",
    "#improvement.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### drop na values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:04.565768Z",
     "start_time": "2019-01-24T22:13:04.552505Z"
    }
   },
   "outputs": [],
   "source": [
    "improvement = improvement.dropna(how='any')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Make a tfidf function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:04.570546Z",
     "start_time": "2019-01-24T22:13:04.567769Z"
    }
   },
   "outputs": [],
   "source": [
    "vect = sklearn.feature_extraction.text.TfidfVectorizer(max_features=11)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Apply tfidf function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:04.585745Z",
     "start_time": "2019-01-24T22:13:04.572888Z"
    }
   },
   "outputs": [],
   "source": [
    "lemmas = improvement['lemmas_improvement'].apply(lambda x: ' '.join(x))\n",
    "features = vect.fit_transform(lemmas)\n",
    "\n",
    "feature_names = vect.get_feature_names()\n",
    "dense = features.todense()\n",
    "denselist = dense.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:04.652161Z",
     "start_time": "2019-01-24T22:13:04.587513Z"
    }
   },
   "outputs": [],
   "source": [
    "improvement['all'] = 'all'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:04.665955Z",
     "start_time": "2019-01-24T22:13:04.653662Z"
    }
   },
   "outputs": [],
   "source": [
    "dfs = top_feats_by_class(features, improvement['emp_group'], feature_names)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:04.671959Z",
     "start_time": "2019-01-24T22:13:04.667779Z"
    }
   },
   "outputs": [],
   "source": [
    "df_all = top_feats_by_class(features, improvement['all'], feature_names)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Visualize TF-IDF"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:07.027833Z",
     "start_time": "2019-01-24T22:13:04.673730Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1224x720 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_tfidf_classfeats_h(dfs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:07.259838Z",
     "start_time": "2019-01-24T22:13:07.030172Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1224x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_tfidf_classfeats_h(df_all)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "_Improvements:_<br>\n",
    " - The majority of participants mentioned **time**, though the context varies a bit between participants. Ex. _\"I think this was a very good 1-day training. But these are not 1-day topics. It takes time to get aligned on a vision for what EDI needs to be internally in the DOE, and to build trust within the group to be able to honestly reflect on what it is today / why it is the way it is today. Many participants at my table were coming from different places, and at different levels of readiness for these conversations.\"_ vs _\"More time spent on how to bring back what we discussed and learned to our work place. That piece felt rushed at the end.\"_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T21:19:07.335076Z",
     "start_time": "2019-01-24T21:19:07.332627Z"
    }
   },
   "source": [
    "##### kmeans"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-01-24T22:13:07.288574Z",
     "start_time": "2019-01-24T22:13:07.261979Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top terms per cluster:\n",
      "Cluster 0:\n",
      " bias\n",
      " people\n",
      " training\n",
      " think\n",
      " work\n",
      " doe\n",
      " session\n",
      " time\n",
      " group\n",
      " feel\n",
      "Cluster 1:\n",
      " time\n",
      " session\n",
      " think\n",
      " work\n",
      " people\n",
      " doe\n",
      " bias\n",
      " feel\n",
      " group\n",
      " training\n",
      "Cluster 2:\n",
      " day\n",
      " time\n",
      " session\n",
      " training\n",
      " work\n",
      " think\n",
      " feel\n",
      " people\n",
      " bias\n",
      " group\n",
      "Cluster 3:\n",
      " session\n",
      " people\n",
      " bias\n",
      " doe\n",
      " think\n",
      " training\n",
      " time\n",
      " feel\n",
      " group\n",
      " work\n",
      "Cluster 4:\n",
      " group\n",
      " session\n",
      " time\n",
      " work\n",
      " bias\n",
      " think\n",
      " doe\n",
      " training\n",
      " people\n",
      " feel\n",
      "Cluster 5:\n",
      " feel\n",
      " training\n",
      " work\n",
      " time\n",
      " session\n",
      " doe\n",
      " people\n",
      " bias\n",
      " group\n",
      " think\n"
     ]
    }
   ],
   "source": [
    "X = vect.fit_transform(improvement['lemmas_improvement'].apply(lambda x: ' '.join(x)))\n",
    "\n",
    "true_k = 6\n",
    "model = KMeans(n_clusters=true_k, init='k-means++', max_iter=100, n_init=1)\n",
    "model.fit(X)\n",
    "\n",
    "print(\"Top terms per cluster:\")\n",
    "order_centroids = model.cluster_centers_.argsort()[:, ::-1]\n",
    "terms = vect.get_feature_names()\n",
    "for i in range(true_k):\n",
    "    print(\"Cluster %d:\" % i),\n",
    "    for ind in order_centroids[i, :10]:\n",
    "        print(' %s' % terms[ind]),\n",
    "    print"
   ]
  }
 ],
 "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": "3.7.0"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}