From 48cc920ae7cb433d26710e8e0d10d66e657debb9 Mon Sep 17 00:00:00 2001
From: Kamil Hryniewicz <kamil@magnet.coop>
Date: Thu, 11 Jun 2020 08:50:20 +0200
Subject: [PATCH] Run notebook on subset

---
 liberia.ipynb | 551 ++++++++++++++++++++++++++++++++++++++++++++++++--
 1 file changed, 536 insertions(+), 15 deletions(-)

diff --git a/liberia.ipynb b/liberia.ipynb
index 45c1a93de..5d072805c 100644
--- a/liberia.ipynb
+++ b/liberia.ipynb
@@ -2,7 +2,39 @@
  "cells": [
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n",
+    "df = pd.read_csv(\"testdata/liberia.csv\")\n",
+    "df.drop([\"quality\", \"problems\"], axis=1, inplace=True)\n",
+    "df[:1000].to_csv(\"testdata/liberia_noq.csv\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "11724"
+      ]
+     },
+     "execution_count": 2,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "len(df)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -21,9 +53,17 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 4,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "06:44:25 - [DEBUG] - initiating session with parameters: {'db_user': 'holocleanuser', 'db_pwd': 'abcd1234', 'db_host': 'localhost', 'db_name': 'holo', 'threads': 1, 'timeout': 180000, 'seed': 45, 'layer_sizes': [1], 'learning_rate': 0.001, 'optimizer': 'adam', 'epochs': 20, 'weight_decay': 0.01, 'momentum': 0.0, 'batch_size': 1, 'weak_label_thresh': 0.99, 'domain_thresh_1': 0.0, 'domain_thresh_2': 0.0, 'max_domain': 10000, 'cor_strength': 0.6, 'nb_cor_strength': 0.8, 'feature_norm': False, 'weight_norm': False, 'estimator_type': 'NaiveBayes', 'estimator_epochs': 10, 'estimator_batch_size': 32, 'estimator_embedding_size': 10, 'train_attrs': None, 'infer_mode': 'dk', 'verbose': True, 'bias': False, 'print_fw': True, 'debug_mode': False}\n"
+     ]
+    }
+   ],
    "source": [
     "hc = holoclean.HoloClean(\n",
     "    db_name='holo',\n",
@@ -54,11 +94,41 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 5,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "06:44:25 - [WARNING] - Dropping the following null column from the dataset: 'Region'\n",
+      "06:44:25 - [WARNING] - Dropping the following null column from the dataset: 'Stroke test'\n",
+      "06:44:25 - [ INFO] - Loaded 1000 rows with 19000 cells\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [ INFO] - DONE Loading liberia_noq.csv\n",
+      "06:44:25 - [DEBUG] - Time to load dataset: 0.22 secs\n"
+     ]
+    }
+   ],
    "source": [
-    "hc.load_data(\"liberia\", \"testdata/liberia.csv\")"
+    "hc.load_data(\"liberia\", \"testdata/liberia_noq.csv\")"
    ]
   },
   {
@@ -99,9 +169,21 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 6,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "06:44:25 - [DEBUG] - DONE with Error Detector: NullDetector in 0.04 secs\n",
+      "06:44:25 - [ INFO] - detected 483 potentially erroneous cells\n",
+      "06:44:25 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:25 - [ INFO] - DONE with error detection.\n",
+      "06:44:25 - [DEBUG] - Time to detect errors: 0.08 secs\n"
+     ]
+    }
+   ],
    "source": [
     "detectors = [NullDetector()]\n",
     "hc.detect_errors(detectors)"
@@ -116,31 +198,470 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 7,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "06:44:25 - [DEBUG] - Computing correlations...\n",
+      "06:44:32 - [DEBUG] - correlations:\n",
+      "attr             Unnamed: 0  Instance  Submission Date  Submitter  settlement  \\\n",
+      "cond_attr                                                                       \n",
+      "CMMoney            0.178756  0.178756         0.182726   0.168557    0.192597   \n",
+      "Functionality      0.109320  0.109320         0.111965   0.087271    0.113627   \n",
+      "Geo Code           0.988325  0.988325         0.995203   0.998560    0.997882   \n",
+      "Instance           1.000000  1.000000         1.000000   1.000000    1.000000   \n",
+      "Latitude           0.993098  0.993098         1.000000   1.000000    1.000000   \n",
+      "Longitude          0.993098  0.993098         1.000000   1.000000    1.000000   \n",
+      "Submission Date    0.992775  0.992775         1.000000   0.999674    0.999793   \n",
+      "Submitter          0.585167  0.585167         0.590139   1.000000    0.613940   \n",
+      "Unnamed: 0         1.000000  1.000000         1.000000   1.000000    1.000000   \n",
+      "Wptype             0.026905  0.026905         0.029122   0.088595    0.040718   \n",
+      "YearCons           0.354940  0.354940         0.358721   0.223316    0.356973   \n",
+      "committee          0.109022  0.109022         0.111665   0.101341    0.120695   \n",
+      "drinking?          0.083260  0.083260         0.085769   0.076938    0.089179   \n",
+      "handpump           0.099625  0.099625         0.102220   0.126282    0.109051   \n",
+      "partner            0.729834  0.729834         0.735533   0.723274    0.742956   \n",
+      "photo              0.992976  0.992976         1.000000   1.000000    1.000000   \n",
+      "quantity           0.173257  0.173257         0.176234   0.122849    0.177834   \n",
+      "settlement         0.957531  0.957531         0.964371   0.995318    1.000000   \n",
+      "\n",
+      "attr               Wptype  handpump  Functionality  YearCons   partner  \\\n",
+      "cond_attr                                                                \n",
+      "CMMoney          0.951049  0.744687       0.521480  0.314553  0.246513   \n",
+      "Functionality    0.931390  0.733954       1.000000  0.298901  0.198882   \n",
+      "Geo Code         0.999436  0.996619       0.994949  0.996656  0.996934   \n",
+      "Instance         1.000000  1.000000       1.000000  1.000000  1.000000   \n",
+      "Latitude         1.000000  1.000000       1.000000  1.000000  1.000000   \n",
+      "Longitude        1.000000  1.000000       1.000000  1.000000  1.000000   \n",
+      "Submission Date  1.000000  1.000000       1.000000  0.999586  0.999762   \n",
+      "Submitter        0.969971  0.831470       0.535768  0.482146  0.627726   \n",
+      "Unnamed: 0       1.000000  1.000000       1.000000  1.000000  1.000000   \n",
+      "Wptype           1.000000  0.761506       0.466353  0.302570  0.166037   \n",
+      "YearCons         0.959124  0.784619       0.498981  1.000000  0.380487   \n",
+      "committee        0.934076  0.732856       0.466171  0.291265  0.199898   \n",
+      "drinking?        0.930546  0.729746       0.651284  0.290177  0.180482   \n",
+      "handpump         0.963661  1.000000       0.474810  0.322976  0.221755   \n",
+      "partner          0.983062  0.916437       0.698272  0.691148  1.000000   \n",
+      "photo            1.000000  1.000000       1.000000  1.000000  1.000000   \n",
+      "quantity         0.933570  0.732716       0.535806  0.297367  0.230158   \n",
+      "settlement       0.997421  0.979733       0.937511  0.955632  0.974529   \n",
+      "\n",
+      "attr             drinking?  quantity  committee   CMMoney     photo  Latitude  \\\n",
+      "cond_attr                                                                       \n",
+      "CMMoney           0.328969  0.219604   0.000000  1.000000  0.182726  0.182596   \n",
+      "Functionality     0.562279  0.217212   0.328256  0.430005  0.111897  0.112065   \n",
+      "Geo Code          0.985859  0.992548   0.996176  0.998757  0.995204  0.995204   \n",
+      "Instance          1.000000  1.000000   1.000000  1.000000  1.000000  1.000000   \n",
+      "Latitude          1.000000  1.000000   1.000000  1.000000  1.000000  1.000000   \n",
+      "Longitude         1.000000  1.000000   1.000000  1.000000  1.000000  1.000000   \n",
+      "Submission Date   1.000000  1.000000   1.000000  1.000000  0.999798  0.999796   \n",
+      "Submitter         0.267137  0.324643   0.466151  0.674407  0.590202  0.591021   \n",
+      "Unnamed: 0        1.000000  1.000000   1.000000  1.000000  1.000000  1.000000   \n",
+      "Wptype            0.189388  0.151665   0.334559  0.423152  0.029067  0.029304   \n",
+      "YearCons          0.218376  0.176586   0.345997  0.475619  0.358819  0.359399   \n",
+      "committee         0.202467  0.149695   1.000000  0.415637  0.111597  0.111808   \n",
+      "drinking?         1.000000  0.214351   0.334826  0.422561  0.085705  0.085732   \n",
+      "handpump          0.194244  0.153998   0.336673  0.429497  0.102153  0.102168   \n",
+      "partner           0.501919  0.513555   0.626205  0.738250  0.735574  0.735927   \n",
+      "photo             1.000000  1.000000   1.000000  1.000000  1.000000  1.000000   \n",
+      "quantity          0.325599  1.000000   0.330927  0.425682  0.176156  0.176641   \n",
+      "settlement        0.898126  0.929603   0.966176  0.987533  0.964377  0.964249   \n",
+      "\n",
+      "attr             Longitude  Geo Code  \n",
+      "cond_attr                             \n",
+      "CMMoney           0.182596  0.183674  \n",
+      "Functionality     0.112065  0.112835  \n",
+      "Geo Code          0.995204  1.000000  \n",
+      "Instance          1.000000  1.000000  \n",
+      "Latitude          1.000000  1.000000  \n",
+      "Longitude         1.000000  1.000000  \n",
+      "Submission Date   0.999796  0.999795  \n",
+      "Submitter         0.591021  0.593579  \n",
+      "Unnamed: 0        1.000000  1.000000  \n",
+      "Wptype            0.029304  0.030612  \n",
+      "YearCons          0.359399  0.360306  \n",
+      "committee         0.111808  0.112987  \n",
+      "drinking?         0.085732  0.086003  \n",
+      "handpump          0.102168  0.102700  \n",
+      "partner           0.735927  0.737075  \n",
+      "photo             1.000000  1.000000  \n",
+      "quantity          0.176641  0.177145  \n",
+      "settlement        0.964249  0.966866  \n",
+      "06:44:32 - [DEBUG] - summary of correlations:\n",
+      "attr   Unnamed: 0   Instance  Submission Date  Submitter  settlement  \\\n",
+      "count   18.000000  18.000000        18.000000  18.000000   18.000000   \n",
+      "mean     0.575994   0.575994         0.580204   0.595110    0.586402   \n",
+      "std      0.418003   0.418003         0.419176   0.439011    0.418665   \n",
+      "min      0.026905   0.026905         0.029122   0.076938    0.040718   \n",
+      "25%      0.125304   0.125304         0.128032   0.123707    0.134980   \n",
+      "50%      0.657501   0.657501         0.662836   0.859296    0.678448   \n",
+      "75%      0.992926   0.992926         1.000000   1.000000    1.000000   \n",
+      "max      1.000000   1.000000         1.000000   1.000000    1.000000   \n",
+      "\n",
+      "attr      Wptype   handpump  Functionality   YearCons    partner  drinking?  \\\n",
+      "count  18.000000  18.000000      18.000000  18.000000  18.000000  18.000000   \n",
+      "mean    0.975184   0.885797       0.765632   0.680165   0.634620   0.648576   \n",
+      "std     0.028479   0.122744       0.240640   0.336962   0.386167   0.361597   \n",
+      "min     0.930546   0.729746       0.466171   0.290177   0.166037   0.189388   \n",
+      "25%     0.953068   0.748892       0.525052   0.305566   0.223856   0.281752   \n",
+      "50%     0.990242   0.948085       0.817891   0.823390   0.801128   0.730202   \n",
+      "75%     1.000000   1.000000       1.000000   1.000000   1.000000   1.000000   \n",
+      "max     1.000000   1.000000       1.000000   1.000000   1.000000   1.000000   \n",
+      "\n",
+      "attr    quantity  committee    CMMoney      photo   Latitude  Longitude  \\\n",
+      "count  18.000000  18.000000  18.000000  18.000000  18.000000  18.000000   \n",
+      "mean    0.613526   0.670330   0.745617   0.580182   0.580328   0.580328   \n",
+      "std     0.397393   0.353261   0.273618   0.419189   0.419106   0.419106   \n",
+      "min     0.149695   0.000000   0.415637   0.029067   0.029304   0.029304   \n",
+      "25%     0.215066   0.335288   0.429624   0.127962   0.128209   0.128209   \n",
+      "50%     0.721579   0.796191   0.862891   0.662888   0.663474   0.663474   \n",
+      "75%     1.000000   1.000000   1.000000   0.999950   0.999949   0.999949   \n",
+      "max     1.000000   1.000000   1.000000   1.000000   1.000000   1.000000   \n",
+      "\n",
+      "attr    Geo Code  \n",
+      "count  18.000000  \n",
+      "mean    0.581310  \n",
+      "std     0.419157  \n",
+      "min     0.030612  \n",
+      "25%     0.129027  \n",
+      "50%     0.665327  \n",
+      "75%     1.000000  \n",
+      "max     1.000000  \n",
+      "06:44:32 - [DEBUG] - computing frequency and co-occurrence statistics from raw data...\n",
+      "06:44:32 - [DEBUG] - Collecting single/pair-wise statistics...\n",
+      "06:44:34 - [DEBUG] - DONE computing statistics in 2.41s\n",
+      "06:44:34 - [DEBUG] - preparing pruned co-occurring statistics...\n",
+      "100%|██████████| 18/18 [00:00<00:00, 49.43it/s]\n",
+      "06:44:34 - [DEBUG] - DONE with pruned co-occurring statistics in 0.37 secs\n",
+      "06:44:34 - [DEBUG] - generating initial set of un-pruned domain values...\n",
+      "100%|██████████| 1000/1000 [00:00<00:00, 1472.86it/s]\n",
+      "06:44:35 - [DEBUG] - domain size stats: count    6970.000000\n",
+      "mean       11.580631\n",
+      "std        15.005491\n",
+      "min         2.000000\n",
+      "25%         4.000000\n",
+      "50%         6.000000\n",
+      "75%        12.000000\n",
+      "max        80.000000\n",
+      "Name: domain_size, dtype: float64\n",
+      "06:44:35 - [DEBUG] - domain count by attr: partner      999\n",
+      "Latitude     998\n",
+      "Geo Code     998\n",
+      "Longitude    998\n",
+      "YearCons     993\n",
+      "CMMoney      992\n",
+      "Submitter    992\n",
+      "Name: attribute, dtype: int64\n",
+      "06:44:35 - [DEBUG] - DONE generating initial set of domain values in 0.72s\n",
+      "06:44:36 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:36 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:36 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:36 - [DEBUG] - Time to create table: 0.00 secs\n",
+      "06:44:37 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:37 - [ INFO] - DONE with domain preparation.\n",
+      "06:44:37 - [DEBUG] - Time to generate the domain: 11.21 secs\n"
+     ]
+    }
+   ],
    "source": [
     "hc.generate_domain()"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 8,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "06:44:37 - [DEBUG] - training estimator for estimating domain value probabilities...\n",
+      "06:44:37 - [DEBUG] - using estimator: NaiveBayes\n",
+      "06:44:37 - [DEBUG] - DONE training estimator in 0.01s\n",
+      "06:44:37 - [DEBUG] - predicting domain value probabilities from estimator...\n",
+      "06:44:37 - [DEBUG] - DONE predictions in 0.00 secs, re-constructing cell domain...\n",
+      "06:44:37 - [DEBUG] - re-assembling final cell domain table...\n",
+      "100%|██████████| 6970/6970 [00:02<00:00, 3115.74it/s]\n",
+      "100%|██████████| 6970/6970 [00:00<00:00, 142325.85it/s]\n",
+      "06:44:39 - [DEBUG] - DONE assembling cell domain table in 2.52s\n",
+      "06:44:39 - [ INFO] - number of (additional) weak labels assigned from estimator: 144\n",
+      "06:44:39 - [DEBUG] - DONE generating domain and weak labels\n",
+      "06:44:40 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:40 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:40 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:44:40 - [DEBUG] - Time to create table: 0.00 secs\n",
+      "06:44:40 - [DEBUG] - Time to create index: 0.00 secs\n"
+     ]
+    }
+   ],
    "source": [
     "hc.run_estimator()"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 9,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "06:44:40 - [DEBUG] - Time to execute query: 0.00 secs\n",
+      "06:44:40 - [DEBUG] - Time to execute query: 0.00 secs\n",
+      "06:44:40 - [DEBUG] - Time to execute query: 0.00 secs\n",
+      "06:44:40 - [DEBUG] - featurizing training data...\n",
+      "06:44:40 - [DEBUG] - Time to execute query: 0.01 secs\n",
+      "100%|██████████| 6970/6970 [00:12<00:00, 559.48it/s]\n",
+      "06:44:54 - [DEBUG] - DONE featurization. Feature tensor size: torch.Size([6970, 80, 144])\n",
+      "06:44:54 - [DEBUG] - generating weak labels...\n",
+      "06:44:54 - [DEBUG] - Time to execute query: 0.00 secs\n",
+      "100%|██████████| 6661/6661 [00:00<00:00, 71341.86it/s]\n",
+      "06:44:54 - [DEBUG] - DONE generating weak labels.\n",
+      "06:44:54 - [DEBUG] - generating mask...\n",
+      "06:44:54 - [DEBUG] - Time to execute query: 0.00 secs\n",
+      "100%|██████████| 6970/6970 [00:00<00:00, 103917.20it/s]\n",
+      "06:44:54 - [DEBUG] - DONE generating mask.\n",
+      "06:44:54 - [ INFO] - DONE setting up featurized dataset.\n",
+      "06:44:54 - [DEBUG] - Time to featurize data: 13.99 secs\n",
+      "06:44:54 - [DEBUG] - training model with first layer size: ['torch.Size([18, 1])', 'torch.Size([126, 1])']\n",
+      "06:44:54 - [ INFO] - DONE setting up repair model.\n",
+      "06:44:54 - [DEBUG] - Time to setup repair model: 13.99 secs\n",
+      "06:44:54 - [ INFO] - training with 6661 training examples (cells)\n",
+      "  0%|          | 0/20 [00:00<?, ?it/s]06:44:56 - [DEBUG] - Epoch 1, cost = 0.252406, acc = 97.97%\n",
+      "  5%|▌         | 1/20 [00:02<00:41,  2.20s/it]06:44:58 - [DEBUG] - Epoch 2, cost = 0.063995, acc = 98.05%\n",
+      " 10%|█         | 2/20 [00:04<00:39,  2.19s/it]06:45:00 - [DEBUG] - Epoch 3, cost = 0.063603, acc = 98.05%\n",
+      " 15%|█▌        | 3/20 [00:06<00:35,  2.10s/it]06:45:02 - [DEBUG] - Epoch 4, cost = 0.063573, acc = 98.05%\n",
+      " 20%|██        | 4/20 [00:08<00:32,  2.02s/it]06:45:04 - [DEBUG] - Epoch 5, cost = 0.063571, acc = 98.05%\n",
+      " 25%|██▌       | 5/20 [00:09<00:29,  1.97s/it]06:45:06 - [DEBUG] - Epoch 6, cost = 0.063570, acc = 98.05%\n",
+      " 30%|███       | 6/20 [00:11<00:27,  1.94s/it]06:45:08 - [DEBUG] - Epoch 7, cost = 0.063570, acc = 98.05%\n",
+      " 35%|███▌      | 7/20 [00:13<00:24,  1.91s/it]06:45:10 - [DEBUG] - Epoch 8, cost = 0.063570, acc = 98.05%\n",
+      " 40%|████      | 8/20 [00:15<00:23,  1.93s/it]06:45:12 - [DEBUG] - Epoch 9, cost = 0.063570, acc = 98.05%\n",
+      " 45%|████▌     | 9/20 [00:17<00:21,  1.92s/it]06:45:14 - [DEBUG] - Epoch 10, cost = 0.063570, acc = 98.05%\n",
+      " 50%|█████     | 10/20 [00:19<00:19,  1.91s/it]06:45:16 - [DEBUG] - Epoch 11, cost = 0.063570, acc = 98.05%\n",
+      " 55%|█████▌    | 11/20 [00:21<00:17,  2.00s/it]06:45:18 - [DEBUG] - Epoch 12, cost = 0.063570, acc = 98.05%\n",
+      " 60%|██████    | 12/20 [00:23<00:16,  2.06s/it]06:45:20 - [DEBUG] - Epoch 13, cost = 0.063570, acc = 98.05%\n",
+      " 65%|██████▌   | 13/20 [00:26<00:14,  2.10s/it]06:45:22 - [DEBUG] - Epoch 14, cost = 0.063570, acc = 98.05%\n",
+      " 70%|███████   | 14/20 [00:28<00:12,  2.13s/it]06:45:25 - [DEBUG] - Epoch 15, cost = 0.063570, acc = 98.05%\n",
+      " 75%|███████▌  | 15/20 [00:30<00:10,  2.15s/it]06:45:27 - [DEBUG] - Epoch 16, cost = 0.063570, acc = 98.05%\n",
+      " 80%|████████  | 16/20 [00:32<00:08,  2.17s/it]06:45:29 - [DEBUG] - Epoch 17, cost = 0.063570, acc = 98.05%\n",
+      " 85%|████████▌ | 17/20 [00:34<00:06,  2.07s/it]06:45:30 - [DEBUG] - Epoch 18, cost = 0.063570, acc = 98.05%\n",
+      " 90%|█████████ | 18/20 [00:36<00:04,  2.02s/it]06:45:32 - [DEBUG] - Epoch 19, cost = 0.063570, acc = 98.05%\n",
+      " 95%|█████████▌| 19/20 [00:38<00:01,  1.98s/it]06:45:34 - [DEBUG] - Epoch 20, cost = 0.063570, acc = 98.05%\n",
+      "100%|██████████| 20/20 [00:40<00:00,  1.97s/it]\n",
+      "06:45:34 - [ INFO] - DONE training repair model.\n",
+      "06:45:34 - [DEBUG] - Time to fit repair model: 41.25 secs\n",
+      "06:45:34 - [ INFO] - inferring on 453 examples (cells)\n",
+      "06:45:34 - [DEBUG] - Time to execute query: 0.04 secs\n",
+      "06:45:35 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:45:35 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:45:35 - [ INFO] - DONE inferring repairs.\n",
+      "06:45:35 - [DEBUG] - Time to infer correct cell values: 0.17 secs\n",
+      "06:45:35 - [DEBUG] - Time to create table: 0.00 secs\n",
+      "06:45:35 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:45:35 - [DEBUG] - Time to create index: 0.00 secs\n",
+      "06:45:35 - [ INFO] - DONE collecting the inferred values.\n",
+      "06:45:35 - [DEBUG] - Time to collect inferred values: 0.02 secs\n",
+      "06:45:35 - [ INFO] - DONE generating repaired dataset\n",
+      "06:45:35 - [DEBUG] - Time to store repaired dataset: 0.08 secs\n",
+      "06:45:35 - [ INFO] - featurizer FreqFeaturizer,size 18,max 0.0073,min -0.2639,avg -0.0143,abs_avg 0.0152,weights:\n",
+      "Unnamed: 0 -0.0\n",
+      "Instance -0.0\n",
+      "Submission Date 0.0\n",
+      "Submitter -0.0\n",
+      "settlement 0.0\n",
+      "Wptype 0.0\n",
+      "handpump 0.0\n",
+      "Functionality -0.0\n",
+      "YearCons 0.007\n",
+      "partner -0.001\n",
+      "drinking? 0.0\n",
+      "quantity -0.0\n",
+      "committee 0.0\n",
+      "CMMoney -0.264\n",
+      "photo 0.0\n",
+      "Latitude 0.0\n",
+      "Longitude 0.0\n",
+      "Geo Code 0.0\n",
+      "featurizer OccurAttrFeaturizer,size 126,max 1.2106,min -0.2599,avg 0.1930,abs_avg 0.2070,weights:\n",
+      "CMMoney | Unnamed: 0 0.262\n",
+      "CMMoney | Instance 0.262\n",
+      "CMMoney | Submission Date 0.262\n",
+      "CMMoney | Submitter -0.166\n",
+      "CMMoney | settlement 0.257\n",
+      "CMMoney | Wptype -0.26\n",
+      "CMMoney | handpump -0.26\n",
+      "CMMoney | Functionality 0.505\n",
+      "CMMoney | YearCons 0.514\n",
+      "CMMoney | partner 0.642\n",
+      "CMMoney | drinking? -0.201\n",
+      "CMMoney | quantity 0.344\n",
+      "CMMoney | committee 0.046\n",
+      "CMMoney | CMMoney 0.004\n",
+      "CMMoney | photo 0.921\n",
+      "CMMoney | Latitude 0.916\n",
+      "CMMoney | Longitude 0.928\n",
+      "CMMoney | Geo Code 0.274\n",
+      "Geo Code | Unnamed: 0 0.485\n",
+      "Geo Code | Instance 0.014\n",
+      "Geo Code | Submission Date 0.004\n",
+      "Geo Code | Submitter 1.196\n",
+      "Geo Code | settlement 1.205\n",
+      "Geo Code | Wptype 1.211\n",
+      "Geo Code | handpump 0.025\n",
+      "Geo Code | Functionality 0.329\n",
+      "Geo Code | YearCons 0.008\n",
+      "Geo Code | partner 0.004\n",
+      "Geo Code | drinking? 0.973\n",
+      "Geo Code | quantity 0.33\n",
+      "Geo Code | committee 0.984\n",
+      "Geo Code | CMMoney 0.618\n",
+      "Geo Code | photo 0.312\n",
+      "Geo Code | Latitude 0.003\n",
+      "Geo Code | Longitude 0.007\n",
+      "Geo Code | Geo Code 0.936\n",
+      "Latitude | Unnamed: 0 0.943\n",
+      "Latitude | Instance 0.551\n",
+      "Latitude | Submission Date 0.461\n",
+      "Latitude | Submitter 0.415\n",
+      "Latitude | settlement 0.013\n",
+      "Latitude | Wptype 0.021\n",
+      "Latitude | handpump 0.941\n",
+      "Latitude | Functionality 0.913\n",
+      "Latitude | YearCons 0.98\n",
+      "Latitude | partner 0.313\n",
+      "Latitude | drinking? 0.537\n",
+      "Latitude | quantity 0.016\n",
+      "Latitude | committee 0.031\n",
+      "Latitude | CMMoney 0.503\n",
+      "Latitude | photo 0.503\n",
+      "Latitude | Latitude 0.476\n",
+      "Latitude | Longitude 0.467\n",
+      "Latitude | Geo Code 0.001\n",
+      "Longitude | Unnamed: 0 0.0\n",
+      "Longitude | Instance 0.003\n",
+      "Longitude | Submission Date 0.618\n",
+      "Longitude | Submitter 0.575\n",
+      "Longitude | settlement 0.575\n",
+      "Longitude | Wptype 0.563\n",
+      "Longitude | handpump -0.0\n",
+      "Longitude | Functionality -0.0\n",
+      "Longitude | YearCons 0.0\n",
+      "Longitude | partner 0.0\n",
+      "Longitude | drinking? 0.0\n",
+      "Longitude | quantity -0.0\n",
+      "Longitude | committee -0.0\n",
+      "Longitude | CMMoney 0.0\n",
+      "Longitude | photo -0.0\n",
+      "Longitude | Latitude 0.0\n",
+      "Longitude | Longitude 0.0\n",
+      "Longitude | Geo Code 0.0\n",
+      "Submitter | Unnamed: 0 -0.0\n",
+      "Submitter | Instance 0.0\n",
+      "Submitter | Submission Date 0.0\n",
+      "Submitter | Submitter -0.0\n",
+      "Submitter | settlement -0.0\n",
+      "Submitter | Wptype 0.0\n",
+      "Submitter | handpump 0.0\n",
+      "Submitter | Functionality -0.0\n",
+      "Submitter | YearCons 0.0\n",
+      "Submitter | partner -0.0\n",
+      "Submitter | drinking? 0.0\n",
+      "Submitter | quantity 0.0\n",
+      "Submitter | committee -0.0\n",
+      "Submitter | CMMoney -0.0\n",
+      "Submitter | photo 0.0\n",
+      "Submitter | Latitude 0.0\n",
+      "Submitter | Longitude -0.0\n",
+      "Submitter | Geo Code -0.0\n",
+      "YearCons | Unnamed: 0 0.0\n",
+      "YearCons | Instance 0.0\n",
+      "YearCons | Submission Date -0.0\n",
+      "YearCons | Submitter 0.0\n",
+      "YearCons | settlement 0.0\n",
+      "YearCons | Wptype 0.0\n",
+      "YearCons | handpump -0.0\n",
+      "YearCons | Functionality 0.0\n",
+      "YearCons | YearCons 0.0\n",
+      "YearCons | partner 0.0\n",
+      "YearCons | drinking? 0.0\n",
+      "YearCons | quantity -0.0\n",
+      "YearCons | committee -0.0\n",
+      "YearCons | CMMoney -0.0\n",
+      "YearCons | photo 0.0\n",
+      "YearCons | Latitude 0.0\n",
+      "YearCons | Longitude -0.0\n",
+      "YearCons | Geo Code -0.0\n",
+      "partner | Unnamed: 0 -0.0\n",
+      "partner | Instance -0.0\n",
+      "partner | Submission Date 0.0\n",
+      "partner | Submitter 0.0\n",
+      "partner | settlement -0.0\n",
+      "partner | Wptype 0.0\n",
+      "partner | handpump 0.0\n",
+      "partner | Functionality 0.0\n",
+      "partner | YearCons 0.0\n",
+      "partner | partner 0.0\n",
+      "partner | drinking? 0.0\n",
+      "partner | quantity -0.0\n",
+      "partner | committee 0.0\n",
+      "partner | CMMoney -0.0\n",
+      "partner | photo -0.0\n",
+      "partner | Latitude -0.0\n",
+      "partner | Longitude 0.0\n",
+      "partner | Geo Code -0.0\n",
+      "\n",
+      "06:45:35 - [DEBUG] - Time to store featurizer weights: 0.00 secs\n"
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "'featurizer FreqFeaturizer,size 18,max 0.0073,min -0.2639,avg -0.0143,abs_avg 0.0152,weights:\\nUnnamed: 0 -0.0\\nInstance -0.0\\nSubmission Date 0.0\\nSubmitter -0.0\\nsettlement 0.0\\nWptype 0.0\\nhandpump 0.0\\nFunctionality -0.0\\nYearCons 0.007\\npartner -0.001\\ndrinking? 0.0\\nquantity -0.0\\ncommittee 0.0\\nCMMoney -0.264\\nphoto 0.0\\nLatitude 0.0\\nLongitude 0.0\\nGeo Code 0.0\\nfeaturizer OccurAttrFeaturizer,size 126,max 1.2106,min -0.2599,avg 0.1930,abs_avg 0.2070,weights:\\nCMMoney | Unnamed: 0 0.262\\nCMMoney | Instance 0.262\\nCMMoney | Submission Date 0.262\\nCMMoney | Submitter -0.166\\nCMMoney | settlement 0.257\\nCMMoney | Wptype -0.26\\nCMMoney | handpump -0.26\\nCMMoney | Functionality 0.505\\nCMMoney | YearCons 0.514\\nCMMoney | partner 0.642\\nCMMoney | drinking? -0.201\\nCMMoney | quantity 0.344\\nCMMoney | committee 0.046\\nCMMoney | CMMoney 0.004\\nCMMoney | photo 0.921\\nCMMoney | Latitude 0.916\\nCMMoney | Longitude 0.928\\nCMMoney | Geo Code 0.274\\nGeo Code | Unnamed: 0 0.485\\nGeo Code | Instance 0.014\\nGeo Code | Submission Date 0.004\\nGeo Code | Submitter 1.196\\nGeo Code | settlement 1.205\\nGeo Code | Wptype 1.211\\nGeo Code | handpump 0.025\\nGeo Code | Functionality 0.329\\nGeo Code | YearCons 0.008\\nGeo Code | partner 0.004\\nGeo Code | drinking? 0.973\\nGeo Code | quantity 0.33\\nGeo Code | committee 0.984\\nGeo Code | CMMoney 0.618\\nGeo Code | photo 0.312\\nGeo Code | Latitude 0.003\\nGeo Code | Longitude 0.007\\nGeo Code | Geo Code 0.936\\nLatitude | Unnamed: 0 0.943\\nLatitude | Instance 0.551\\nLatitude | Submission Date 0.461\\nLatitude | Submitter 0.415\\nLatitude | settlement 0.013\\nLatitude | Wptype 0.021\\nLatitude | handpump 0.941\\nLatitude | Functionality 0.913\\nLatitude | YearCons 0.98\\nLatitude | partner 0.313\\nLatitude | drinking? 0.537\\nLatitude | quantity 0.016\\nLatitude | committee 0.031\\nLatitude | CMMoney 0.503\\nLatitude | photo 0.503\\nLatitude | Latitude 0.476\\nLatitude | Longitude 0.467\\nLatitude | Geo Code 0.001\\nLongitude | Unnamed: 0 0.0\\nLongitude | Instance 0.003\\nLongitude | Submission Date 0.618\\nLongitude | Submitter 0.575\\nLongitude | settlement 0.575\\nLongitude | Wptype 0.563\\nLongitude | handpump -0.0\\nLongitude | Functionality -0.0\\nLongitude | YearCons 0.0\\nLongitude | partner 0.0\\nLongitude | drinking? 0.0\\nLongitude | quantity -0.0\\nLongitude | committee -0.0\\nLongitude | CMMoney 0.0\\nLongitude | photo -0.0\\nLongitude | Latitude 0.0\\nLongitude | Longitude 0.0\\nLongitude | Geo Code 0.0\\nSubmitter | Unnamed: 0 -0.0\\nSubmitter | Instance 0.0\\nSubmitter | Submission Date 0.0\\nSubmitter | Submitter -0.0\\nSubmitter | settlement -0.0\\nSubmitter | Wptype 0.0\\nSubmitter | handpump 0.0\\nSubmitter | Functionality -0.0\\nSubmitter | YearCons 0.0\\nSubmitter | partner -0.0\\nSubmitter | drinking? 0.0\\nSubmitter | quantity 0.0\\nSubmitter | committee -0.0\\nSubmitter | CMMoney -0.0\\nSubmitter | photo 0.0\\nSubmitter | Latitude 0.0\\nSubmitter | Longitude -0.0\\nSubmitter | Geo Code -0.0\\nYearCons | Unnamed: 0 0.0\\nYearCons | Instance 0.0\\nYearCons | Submission Date -0.0\\nYearCons | Submitter 0.0\\nYearCons | settlement 0.0\\nYearCons | Wptype 0.0\\nYearCons | handpump -0.0\\nYearCons | Functionality 0.0\\nYearCons | YearCons 0.0\\nYearCons | partner 0.0\\nYearCons | drinking? 0.0\\nYearCons | quantity -0.0\\nYearCons | committee -0.0\\nYearCons | CMMoney -0.0\\nYearCons | photo 0.0\\nYearCons | Latitude 0.0\\nYearCons | Longitude -0.0\\nYearCons | Geo Code -0.0\\npartner | Unnamed: 0 -0.0\\npartner | Instance -0.0\\npartner | Submission Date 0.0\\npartner | Submitter 0.0\\npartner | settlement -0.0\\npartner | Wptype 0.0\\npartner | handpump 0.0\\npartner | Functionality 0.0\\npartner | YearCons 0.0\\npartner | partner 0.0\\npartner | drinking? 0.0\\npartner | quantity -0.0\\npartner | committee 0.0\\npartner | CMMoney -0.0\\npartner | photo -0.0\\npartner | Latitude -0.0\\npartner | Longitude 0.0\\npartner | Geo Code -0.0\\n'"
+      ]
+     },
+     "execution_count": 9,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
    "source": [
     "featurizers = [FreqFeaturizer(), OccurAttrFeaturizer()]\n",
     "hc.repair_errors(featurizers)"
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df2 = hc.ds.get_raw_data()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 16,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "True"
+      ]
+     },
+     "execution_count": 16,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "df.equals(df)"
+   ]
   }
  ],
  "metadata": {
@@ -159,7 +680,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.6.10"
+   "version": "3.6.8"
   }
  },
  "nbformat": 4,