diff --git a/.ipynb_checkpoints/CS 4780 Final Project Student Template-checkpoint.ipynb b/.ipynb_checkpoints/CS 4780 Final Project Student Template-checkpoint.ipynb
index 4223a52..5c459fd 100755
--- a/.ipynb_checkpoints/CS 4780 Final Project Student Template-checkpoint.ipynb	
+++ b/.ipynb_checkpoints/CS 4780 Final Project Student Template-checkpoint.ipynb	
@@ -74,6 +74,7 @@
     "import numpy as np\n",
     "#import sklearn as sk\n",
     "from sklearn.preprocessing import StandardScaler\n",
+    "from sklearn.preprocessing import KBinsDiscretizer\n",
     "from sklearn.neighbors import KNeighborsClassifier\n",
     "from sklearn import svm\n",
     "from sklearn.ensemble import AdaBoostClassifier as ABC\n",
@@ -84,7 +85,6 @@
     "from sklearn.model_selection import GridSearchCV\n",
     "from sklearn.model_selection import KFold\n",
     "from sklearn.metrics import balanced_accuracy_score\n",
-    "from sklearn.preprocessing import KBinsDiscretizer\n",
     "from sklearn.compose import ColumnTransformer"
    ]
   },
@@ -159,7 +159,6 @@
     "\n",
     "#Trim the data of irrelevant fields and convert to numpy array\n",
     "data = data[:,4:]\n",
-    "#print(data.dtypes)\n",
     "\n",
     "#Load and format the test_2016_no_label.csv file\n",
     "test = pd.read_csv(\"test_2016_no_label.csv\", sep=',', header=0, encoding='unicode_escape', thousands=\",\")\n",
@@ -248,16 +247,16 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "knn avg weighted accuracy:  0.6203334398310165\n",
-      "dtc avg weighted accuracy:  0.7070879048578987\n",
-      "dtc_t avg weighted accuracy:  0.6931251192737173\n",
-      "adb_dtc avg weighted accuracy:  0.6961593601061467\n",
-      "svc avg weighted accuracy:  0.7004446186230044\n",
-      "lr avg weighted accuracy:  0.6590264805800705\n",
-      "adb_lr avg weighted accuracy:  0.6497330664169988\n",
-      "nbc avg weighted accuracy:  0.538413031775743\n",
-      "adb_nbc avg weighted accuracy:  0.5742813326608044\n",
-      "219.0\n"
+      "knn avg weighted accuracy:  0.6237076582105193\n",
+      "dtc avg weighted accuracy:  0.6979205324376133\n",
+      "dtc_t avg weighted accuracy:  0.6855368083208002\n",
+      "adb_dtc avg weighted accuracy:  0.700633705568728\n",
+      "svc avg weighted accuracy:  0.6971740145905911\n",
+      "lr avg weighted accuracy:  0.6592688655540591\n",
+      "adb_lr avg weighted accuracy:  0.6488225086599588\n",
+      "nbc avg weighted accuracy:  0.5358837627311691\n",
+      "adb_nbc avg weighted accuracy:  0.5720794507211426\n",
+      "250.0\n"
      ]
     }
    ],
@@ -350,69 +349,9 @@
     "You may follow the steps in part 2 again but making innovative changes like creating new features, using new training algorithms, etc. Make sure you explain everything clearly in part 3.2. Note that reaching the 75% creative baseline is only a small portion of this part. Any creative ideas will receive most points as long as they are reasonable and clearly explained."
    ]
   },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "# Make sure you comment your code clearly and you may refer to these comments in the part 3.2\n",
-    "# TODO\n",
-    "\n",
-    "dtc.fit(train, yTr)\n",
-    "dtc_preds = dtc.predict(valid)\n",
-    "print(\"dtc weighted accuracy: \", weighted_accuracy(dtc_preds,yV)) #does pretty well, even with no pruning\n",
-    "\n",
-    "dtc_g.fit(train, yTr)\n",
-    "dtc_g_preds = dtc.predict(valid)\n",
-    "print(\"dtc_g weighted accuracy: \", weighted_accuracy(dtc_g_preds,yV)) #does pretty well, even with no pruning\n",
-    "\n",
-    "adb_dtc.fit(train, yTr)\n",
-    "adb_dtc_preds = adb_dtc.predict(valid)\n",
-    "print(\"adb_dtc weighted accuracy: \", weighted_accuracy(adb_dtc_preds,yV))\n",
-    "\n",
-    "#best validation error so far\n",
-    "svc.fit(train, yTr)\n",
-    "svc_preds = svc.predict(valid)\n",
-    "print(\"svc weighted accuracy: \", weighted_accuracy(svc_preds,yV))\n",
-    "\n",
-    "#does not work very well; not really better than random\n",
-    "nbc.fit(train, yTr)\n",
-    "nbc_preds = nbc.predict(valid)\n",
-    "print(\"nbc weighted accuracy: \", weighted_accuracy(nbc_preds,yV))\n",
-    "\n",
-    "adb_nbc.fit(train, yTr)\n",
-    "adb_nbc_preds = adb_nbc.predict(valid)\n",
-    "print(\"adb_nbc weighted accuracy: \", weighted_accuracy(adb_nbc_preds,yV))\n",
-    "\n",
-    "#Split the data into a train and test set\n",
-    "np.random.shuffle(data)\n",
-    "    \n",
-    "results = np.array(df[:,-1],dtype='bool')\n",
-    "data = df[:,:6]\n",
-    "n_valid = int(len(data)/5)\n",
-    "train = data[:4*n_valid]\n",
-    "yTr = results[:4*n_valid]\n",
-    "valid = data[4*n_valid:]\n",
-    "    yV = results[4*n_valid:]\n",
-    "\n",
-    "n_valid = int(len(data)/5)\n",
-    "j = 100;\n",
-    "\n",
-    "#for k in range(5, 100):\n",
-    "sum_err = 0;\n",
-    "#knn = KNeighborsClassifier(n_neighbors = k)\n",
-    "for i in range(j):\n",
-    "    #randomize training and validation set\n",
-    "    np.random.shuffle(data)\n",
-    "\n",
-    "    #run algorithm\n",
-    "    knn.fit(train, yTr)\n",
-    "    knn_preds = knn.predict(valid)\n",
-    "    sum_err = sum_err + sum(knn_preds != yV)/n_valid"
-   ]
-  },
   {
    "cell_type": "code",
-   "execution_count": 13,
+   "execution_count": 58,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -420,97 +359,48 @@
     "\n",
     "### (3.1) Preprocessing and Feature Extraction ###\n",
     "\n",
-    "###(3.1.a) Load and format the training data set(s)###\n",
-    "data = pd.read_csv(\"train_2016.csv\", sep=',', header=0, encoding='unicode_escape', thousands=\",\")\n",
-    "data_2012 = pd.read_csv(\"train_2012.csv\", sep=\",\", header=0, encoding='unicode_escape', thousands=',')\n",
-    "graph = pd.read_csv(\"graph.csv\", sep=',')\n",
+    "###Helper functions to preprocess data###\n",
     "\n",
     "#Add binary features for the state that the county is in (where each binary feature represents one state)\n",
-    "\n",
-    "###2016###\n",
-    "data[\"State\"] = [x.strip()[-2:] for x in data['County']] #Extract state abbreviations from County\n",
-    "dummies = pd.get_dummies(data['State']) #Create state dummies from State\n",
-    "data = pd.concat([data,dummies],axis=1)\n",
-    "data = data.drop('State', axis=1)\n",
-    "\n",
-    "###2012###\n",
-    "data_2012[\"State\"] = [x.strip()[-2:] for x in data_2012['County']] #Extract state abbreviations from County\n",
-    "dummies_2012 = pd.get_dummies(data_2012['State']) #Create state dummies from State\n",
-    "data_2012 = pd.concat([data_2012,dummies_2012],axis=1)\n",
-    "data_2012 = data_2012.drop(\"State\", axis=1)\n",
-    "\n",
-    "#Define a function which will preprocess the data\n",
-    "def preprocess(data):\n",
-    "    #Normalize features\n",
-    "    #Standardize the data with mean 0 and s.d. 1\n",
+    "def get_states(X):\n",
+    "    X[\"State\"] = [x.strip()[-2:] for x in X['County']]\n",
+    "    dummies_test = pd.get_dummies(X['State'])\n",
+    "    X = pd.concat([X,dummies_test],axis=1)\n",
+    "    X = X.drop('State',axis=1)\n",
+    "    return X\n",
+    "\n",
+    "#Normalizes all variables from the data except for the binary state variables\n",
+    "def normalize(X):\n",
+    "    #Normalize features Standardize the data with mean 0 and s.d. 1\n",
     "    scalar = StandardScaler()\n",
-    "    _,d = np.shape(data)\n",
-    "    data_norm = data[:,[0,1,2,3,4,5,(d-2),(d-1)]] #d-2 and d-1 store the graph score and number of known neighbors\n",
-    "    #data_raw_indexes = np.r_[6:(d-2)]\n",
-    "    data_bin = data[:,6:(d-2)]\n",
-    "    #print('untransformed data', data_bin)\n",
+    "    _,d = np.shape(X)\n",
+    "    data_norm = X[:,[0,1,2,3,4,5,(d-2),(d-1)]] #d-2 and d-1 store the graph score and number of known neighbors\n",
+    "    data_bin = X[:,6:(d-2)]\n",
     "    scalar.fit(data_norm)\n",
     "    data_norm = scalar.transform(data_norm)\n",
-    "    #print('transformed data', data_norm)\n",
-    "    data = np.hstack((data_norm, data_bin))\n",
-    "    #print(data)\n",
-    "\n",
-    "    #scalar = StandardScaler()\n",
-    "    #binning = ['MedianIncome']\n",
-    "    #standardization = ['MigraRate','BirthRate','DeathRate','BachelorRate','UnemploymentRate']\n",
-    "    #n,d = np.shape(counties)\n",
-    "    #print(d)\n",
-    "    #binning = [4]\n",
-    "    #standardization = [*range(5,d)]\n",
-    "    #print(standardization)\n",
-    "    #ct = ColumnTransformer([(\"standardize\", scalar, standardization), \n",
-    "     #                       (\"binning\", KBinsDiscretizer(n_bins=10), binning)])\n",
-    "    #try min maxing if binning doesn't \n",
-    "    #data = ct.fit_transform(counties)\n",
-    "    return data"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "###EXPERIMENTING DON'T RUN YET###\n",
-    "\n",
-    "#Add binary feature for whether a county is on a state border (which may indicate they receive other states' political ads)\n",
-    "fips_allstates = data[[\"FIPS\",\"State\"]]\n",
-    "tmp1 = graph.merge(fips_allstates, left_on=\"SRC\", right_on=\"FIPS\", how=\"left\")\n",
-    "tmp1 = tmp1[[\"SRC\",\"DST\",\"State\"]]\n",
-    "tmp1.rename(columns={'State':'SRC_State'}, inplace=True)\n",
-    "print(tmp1)\n",
+    "    X = np.hstack((data_norm, data_bin))\n",
+    "    return X\n",
     "\n",
-    "tmp2 = tmp1.merge(fips_allstates, left_on=\"DST\", right_on=\"FIPS\", how=\"left\")\n",
-    "tmp2 = tmp2[[\"SRC\",\"DST\",\"SRC_State\",\"State\"]]\n",
-    "tmp2.rename(columns={'State':'DST_State'}, inplace=True)\n",
-    "print(tmp2)\n",
-    "\n",
-    "tmp2[\"SRConBorder\"] = tmp2[\"SRC_State\"].eq(tmp2[\"DST_State\"])\n",
-    "tmp2[\"SRConBorder\"] = tmp2[\"SRConBorder\"].astype(int)\n",
-    "tmp2[\"SRConBorder\"] = tmp2[\"SRConBorder\"].replace({0:1,1:0}) #Make sure =1 iff on border\n",
-    "print(tmp2)\n",
-    "tmp2 = tmp2.dropna() #Because NaNs indicate counties that are not in training data - not sure how to deal with loss of data\n",
-    "tmp2 = tmp2.groupby(['SRC'])['SRConBorder'].sum()\n",
-    "print(tmp2)\n",
+    "###(3.1.a) Load and format the training data set(s)###\n",
+    "data = pd.read_csv(\"train_2016.csv\", sep=',', header=0, encoding='unicode_escape', thousands=\",\")\n",
+    "data_2012 = pd.read_csv(\"train_2012.csv\", sep=\",\", header=0, encoding='unicode_escape', thousands=',')\n",
+    "graph = pd.read_csv(\"graph.csv\", sep=',', header=0, encoding='unicode_escape', thousands=\",\")\n",
     "\n",
-    "data = data.merge(tmp2, left_on=\"FIPS\", right_on=\"SRC\", how=\"left\")\n",
-    "print(data)\n",
+    "###2016###\n",
+    "data = get_states(data)\n",
     "\n",
-    "#Notes: Source counties can be found in test, destination in train (or vice versa - one occurence in either column(?))\n",
-    "#Notes: Get list of neighbors that voted one way or other (check one per column)"
+    "###2012###\n",
+    "data_2012 = get_states(data_2012)\n",
+    "\n"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 14,
+   "execution_count": 59,
    "metadata": {},
    "outputs": [],
    "source": [
     "#Creates a lexicographically ordered list of all the county's neighbors\n",
-    "graph = pd.read_csv(\"graph.csv\", sep=',', header=0, encoding='unicode_escape', thousands=\",\")\n",
     "neighbors = {}\n",
     "for i in range(len(graph)):\n",
     "    src = graph['SRC'][i]\n",
@@ -531,15 +421,12 @@
     "dem = dict(zip(data['FIPS'], dem_perc))\n",
     "votes = data[['DEM','GOP']]\n",
     "votes = list(votes.itertuples(index=False, name=None))\n",
-    "votes = dict(zip(data['FIPS'], votes))\n",
-    "votes_2012 = data_2012[['DEM','GOP']]\n",
-    "votes_2012 = list(votes_2012.itertuples(index=False, name=None))\n",
-    "votes_2012 = dict(zip(data['FIPS'], votes_2012))"
+    "votes = dict(zip(data['FIPS'], votes))"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 15,
+   "execution_count": 60,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -554,7 +441,7 @@
     "            score += dem[i]\n",
     "    return n, score/n if n != 0 else np.nan\n",
     "\n",
-    "#TODO - implement a function that calculates the percentage of all voters in the surrounding counties who voted dem\n",
+    "#Calculates the percentage of all voters from the neighboring counties who voted democratic\n",
     "def aggregate_gscore(neighs, votes):\n",
     "    n = 0\n",
     "    dem = 0\n",
@@ -577,27 +464,37 @@
     "            n = 0\n",
     "        else:\n",
     "            n, score = aggregate_gscore(neighbors[key], votes)\n",
-    "            _, s2012 = aggregate_gscore(neighbors[key], votes_2012)\n",
+    "            total_n = len(neighbors[key])\n",
     "        counties.loc[i, 'GraphScore'] = score\n",
     "        counties.loc[i, 'Neighbors'] = n\n",
-    "        mean = counties['GraphScore'].sum()/len(counties)\n",
-    "    counties['GraphScore'] = counties['GraphScore'].map(lambda x: mean if pd.isnull(x) else x)\n",
+    "        smean = counties['GraphScore'].sum()/len(counties)\n",
+    "    counties['GraphScore'] = counties['GraphScore'].map(lambda x: smean if pd.isnull(x) else x)\n",
     "    return counties"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 16,
+   "execution_count": 61,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "['FIPS', 'County', 'DEM', 'GOP', 'MedianIncome', 'MigraRate', 'BirthRate', 'DeathRate', 'BachelorRate', 'UnemploymentRate', 'AL', 'AR', 'AZ', 'CA', 'CO', 'CT', 'DC', 'DE', 'FL', 'GA', 'HI', 'IA', 'ID', 'IL', 'IN', 'KS', 'KY', 'LA', 'MA', 'MD', 'ME', 'MI', 'MN', 'MO', 'MS', 'MT', 'NC', 'ND', 'NE', 'NH', 'NJ', 'NM', 'NV', 'NY', 'OH', 'OK', 'OR', 'PA', 'RI', 'SC', 'SD', 'TN', 'TX', 'UT', 'VA', 'VT', 'WA', 'WI', 'WV', 'WY', 'GraphScore', 'Neighbors']\n",
+      "['FIPS', 'County', 'DEM', 'GOP', 'MedianIncome', 'MigraRate', 'BirthRate', 'DeathRate', 'BachelorRate', 'UnemploymentRate', 'AL', 'AR', 'AZ', 'CA', 'CO', 'CT', 'DC', 'DE', 'FL', 'GA', 'HI', 'IA', 'ID', 'IL', 'IN', 'KS', 'KY', 'LA', 'MA', 'MD', 'ME', 'MI', 'MN', 'MO', 'MS', 'MT', 'NC', 'ND', 'NE', 'NH', 'NJ', 'NM', 'NV', 'NY', 'OH', 'OK', 'OR', 'PA', 'RI', 'SC', 'SD', 'TN', 'TX', 'UT', 'VA', 'VT', 'WA', 'WI', 'WV', 'WY', 'GraphScore', 'Neighbors']\n"
+     ]
+    }
+   ],
    "source": [
     "###Prep for concatenating###\n",
     "\n",
     "###2016###\n",
-    "#cols = data.columns.tolist()\n",
-    "#cols = cols[0:2] + cols[4:] #Helps make sure columns in test align with those in data\n",
-    "\n",
     "data = add_gscore(data)\n",
+    "#Rearrange the columns to ensure that DC is in the same location of each dataset\n",
+    "cols = data.columns.tolist()\n",
+    "cols = cols[0:2] + cols[4:]\n",
+    "print(list(data.columns))\n",
     "data = data.to_numpy()\n",
     "\n",
     "#These are the training labels for each county, 1 if the county voted Dem and 0 if it voted Rep\n",
@@ -606,11 +503,13 @@
     "\n",
     "#Trim the data of irrelevant fields and convert to numpy array\n",
     "data = data[:,4:]\n",
-    "#print(np.shape(data))\n",
-    "#print(data)\n",
     "\n",
     "###2012###\n",
     "data_2012 = add_gscore(data_2012)\n",
+    "#Rearrange the columns to ensure that DC is in the same location of each dataset\n",
+    "cols_2012 = data_2012.columns.tolist()\n",
+    "cols_2012 = cols_2012[0:2] + cols_2012[4:]\n",
+    "print(list(data_2012.columns))\n",
     "data_2012 = data_2012.to_numpy()\n",
     "\n",
     "#These are the training labels for each county, 1 if the county voted Dem and 0 if it voted Rep\n",
@@ -622,9 +521,18 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 17,
+   "execution_count": 62,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "['FIPS', 'County', 'MedianIncome', 'MigraRate', 'BirthRate', 'DeathRate', 'BachelorRate', 'UnemploymentRate', 'AL', 'AR', 'AZ', 'CA', 'CO', 'CT', 'DC', 'DE', 'FL', 'GA', 'HI', 'IA', 'ID', 'IL', 'IN', 'KS', 'KY', 'LA', 'MA', 'MD', 'ME', 'MI', 'MN', 'MO', 'MS', 'MT', 'NC', 'ND', 'NE', 'NH', 'NJ', 'NM', 'NV', 'NY', 'OH', 'OK', 'OR', 'PA', 'RI', 'SC', 'SD', 'TN', 'TX', 'UT', 'VA', 'VT', 'WA', 'WI', 'WV', 'WY', 'GraphScore', 'Neighbors']\n",
+      "['FIPS', 'County', 'MedianIncome', 'MigraRate', 'BirthRate', 'DeathRate', 'BachelorRate', 'UnemploymentRate', 'AL', 'AR', 'AZ', 'CA', 'CO', 'CT', 'DC', 'DE', 'FL', 'GA', 'HI', 'IA', 'ID', 'IL', 'IN', 'KS', 'KY', 'LA', 'MA', 'MD', 'ME', 'MI', 'MN', 'MO', 'MS', 'MT', 'NC', 'ND', 'NE', 'NH', 'NJ', 'NM', 'NV', 'NY', 'OH', 'OK', 'OR', 'PA', 'RI', 'SC', 'SD', 'TN', 'TX', 'UT', 'VA', 'VT', 'WA', 'WI', 'WV', 'WY', 'GraphScore', 'Neighbors']\n"
+     ]
+    }
+   ],
    "source": [
     "###(3.1.b) Load and format the testing data sets###\n",
     "test = pd.read_csv(\"test_2016_no_label.csv\", sep=',', header=0, encoding='unicode_escape', thousands=\",\")\n",
@@ -633,15 +541,11 @@
     "#Process the testing data set in the same manner as the training data set\n",
     "\n",
     "###2016###\n",
-    "test[\"State\"] = [x.strip()[-2:] for x in test['County']]\n",
-    "dummies_test = pd.get_dummies(test['State'])\n",
-    "#print(dummies.columns.difference(dummies_test.columns)) #Check that DC is the only missing county/state\n",
-    "test = pd.concat([test,dummies_test],axis=1)\n",
+    "test = get_states(test)\n",
     "test['DC'] = 0 #Add in a zero vector for DC (because there is no DC in either test data set)\n",
-    "test = test.drop('State',axis=1)\n",
-    "#print(test)\n",
-    "#test = test[cols] #Make sure columns in test align with those in data\n",
     "test = add_gscore(test)\n",
+    "test = test[cols]\n",
+    "print(list(test.columns))\n",
     "test = test.to_numpy()\n",
     "\n",
     "fips = test[:,0]\n",
@@ -651,21 +555,19 @@
     "data = np.vstack((data, test))\n",
     "\n",
     "#Standardize the (non-binary) features to have mean 0 and s.d. 1\n",
-    "data = preprocess(data)\n",
+    "data = normalize(data)\n",
     "\n",
     "test = data[-n_data:,:]\n",
     "data = data[:n_data,:]\n",
     "data = np.hstack((data, np.transpose(np.array([results]))))\n",
     "\n",
     "###2012###\n",
-    "test_2012[\"State\"] = [x.strip()[-2:] for x in test_2012['County']]\n",
-    "dummies_test_2012 = pd.get_dummies(test_2012['State'])\n",
-    "#print(dummies_2012.columns.difference(dummies_test_2012.columns)) #Check that DC is the only missing county/state\n",
-    "test_2012 = pd.concat([test_2012,dummies_test_2012],axis=1)\n",
+    "test_2012 = get_states(test_2012)\n",
     "test_2012['DC'] = 0 #Add in a zero vector for DC (because there is no DC in either test data set)\n",
-    "test_2012 = test_2012.drop('State',axis=1)\n",
+    "test_2012 = add_gscore(test_2012)\n",
+    "test_2012 = test_2012[cols_2012]\n",
+    "print(list(test_2012.columns))\n",
     "\n",
-    "add_gscore(test_2012)\n",
     "test_2012 = test_2012.to_numpy()\n",
     "\n",
     "fips_2012 = test_2012[:,0]\n",
@@ -674,7 +576,7 @@
     "n_data_2012 = len(data_2012)\n",
     "data_2012 = np.vstack((data_2012, test_2012))\n",
     "\n",
-    "data_2012 = preprocess(data_2012)\n",
+    "data_2012 = normalize(data_2012)\n",
     "test_2012 = data_2012[-n_data_2012:,:]\n",
     "data_2012 = data_2012[:n_data_2012,:]\n",
     "data_2012 = np.hstack((data_2012, np.transpose(np.array([results_2012]))))"
@@ -682,9 +584,20 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 18,
+   "execution_count": 67,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "[ 1.25892541  1.49170935  1.76753662  2.09436625  2.48162892  2.94049911\n",
+      "  3.48421754  4.12847324  4.89185622  5.79639395  6.86818691  8.13816172\n",
+      "  9.64296358 11.42601361 13.5387618  16.04217161 19.00847905 22.52327705\n",
+      " 26.68798528 31.6227766 ]\n"
+     ]
+    }
+   ],
    "source": [
     "### (3.2) Testing algorithms ###\n",
     "\n",
@@ -699,7 +612,9 @@
     "adb_dtc_g = ABC(base_estimator = DTC(criterion = \"entropy\"), n_estimators = 100)\n",
     "\n",
     "#SVM classifier\n",
-    "Cs = np.logspace(1.25,1.25,20)\n",
+    "Cs = np.logspace(.1,1.5,20)\n",
+    "print(Cs)\n",
+    "#Cs = np.logspace(1.25,1.25,20)\n",
     "svc = svm.SVC(gamma='scale', kernel='rbf')\n",
     "c_svc = GridSearchCV(estimator=svc,param_grid=dict(C=Cs))\n",
     "\n",
@@ -716,27 +631,25 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 19,
+   "execution_count": null,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "knn avg weighted accuracy:  0.6329961166249547\n",
-      "knn avg weighted accuracy 2012:  0.6215636493214642\n",
-      "dtc_t avg weighted accuracy:  0.7461445847623741\n",
-      "dtc_t avg weighted accuracy 2012:  0.6661740263653098\n",
-      "adb_dtc avg weighted accuracy:  0.708280844124877\n",
-      "adb_dtc avg weighted accuracy 2012:  0.7105907368805827\n",
-      "dtc_g avg weighted accuracy:  0.7506852491001083\n",
-      "dtc_g avg weighted accuracy 2012:  0.683730011834672\n",
-      "adb_dtc_g avg weighted accuracy:  0.760068618725555\n",
-      "adb_dtc_g avg weighted accuracy 2012:  0.6846362764834062\n",
-      "svc avg weighted accuracy:  0.7974214368589101\n",
-      "svc avg weighted accuracy 2012:  0.7389995155649615\n",
-      "lr avg weighted accuracy:  0.7449232757260804\n",
-      "lr avg weighted accuracy 2012:  0.713245836578974\n"
+      "knn avg weighted accuracy:  0.6330706501088754\n",
+      "knn avg weighted accuracy 2012:  0.6287551049058671\n",
+      "dtc_t avg weighted accuracy:  0.7126628181827105\n",
+      "dtc_t avg weighted accuracy 2012:  0.6654573475411754\n",
+      "adb_dtc avg weighted accuracy:  0.7583268043136182\n",
+      "adb_dtc avg weighted accuracy 2012:  0.6876199289767525\n",
+      "dtc_g avg weighted accuracy:  0.7331652849631233\n",
+      "dtc_g avg weighted accuracy 2012:  0.6960158238672269\n",
+      "adb_dtc_g avg weighted accuracy:  0.7526903758034124\n",
+      "adb_dtc_g avg weighted accuracy 2012:  0.6782594899860952\n",
+      "svc avg weighted accuracy:  0.7850360498781626\n",
+      "svc avg weighted accuracy 2012:  0.7207312417562269\n"
      ]
     }
    ],
@@ -832,42 +745,30 @@
     "print(\"svc avg weighted accuracy: \", x)\n",
     "print(\"svc avg weighted accuracy 2012: \", x_2012)\n",
     "\n",
-    "#print(\"lr avg weighted accuracy: \", kfold(lr, data))\n",
-    "#print(\"lr avg weighted accuracy: \", kfold(c_lr, data, grid=\"Yes\", grid_model=\"lr\")) #.721 using Cs = np.logspace(-1,1,100) and solver = 'liblinear'\n",
-    "#print(\"lr avg weighted accuracy 2012: \", kfold(c_lr, data_2012, grid=\"Yes\", grid_model=\"lr\"))\n",
     "x_lr,y_lr,z_lr = kfold_lr(c_lr, data)\n",
     "x_lr_2012,y_lr_2012,z_lr_2012 = kfold_lr(c_lr, data_2012)\n",
     "print(\"lr avg weighted accuracy: \", x_lr)\n",
     "print(\"lr avg weighted accuracy 2012: \", x_lr_2012)\n",
     "\n",
-    "#print(\"adb_lr avg weighted accuracy: \", kfold(adb_lr, data)) #.713 using solver = 'liblinear'\n",
-    "#print(\"adb_lr avg weighted accuracy 2012: \", kfold(adb_lr, data_2012)) #.713 using solver = 'liblinear'\n",
-    "\n",
-    "#print(\"nbc avg weighted accuracy: \", kfold(nbc, data)) #.653\n",
-    "\n",
-    "#print(\"adb_nbc avg weighted accuracy: \", kfold(adb_nbc, data)) #.5 using n_estimators = 100\n",
-    "\n",
     "#Get final predictions\n",
-    "preds = z.predict(test) #Uses 2016"
+    "preds_creative = z.predict(test) #Uses 2016"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 20,
+   "execution_count": 65,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "223\n",
-      "216\n"
+      "235\n"
      ]
     }
    ],
    "source": [
-    "print(sum(preds))\n",
-    "print(sum(z.predict(data[:,:-1])))"
+    "print(sum(preds_creative))"
    ]
   },
   {
@@ -888,6 +789,13 @@
     "3.2.2 Please explain in detail how you achieved this and what you did specifically and why you tried this."
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "<img src=\"files/Kaggle (Creative) Solution.png\">"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -908,16 +816,27 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 56,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "228\n"
+     ]
+    }
+   ],
    "source": [
     "# TODO\n",
     "output = np.array([fips, preds])\n",
     "output = pd.DataFrame(np.transpose(output))\n",
     "output[1] = output[1].astype(int)\n",
-    "#output.to_csv(\"submission.csv\", header=[\"FIPS\",\"Result\"], index = False)\n",
-    "output.to_csv(\"submission_creative.csv\", header=[\"FIPS\",\"Result\"], index = False)\n",
+    "output_creative = np.array([fips, preds_creative])\n",
+    "output_creative = pd.DataFrame(np.transpose(output_creative))\n",
+    "output_creative[1] = output_creative[1].astype(int)\n",
+    "output.to_csv(\"submission.csv\", header=[\"FIPS\",\"Result\"], index = False)\n",
+    "output_creative.to_csv(\"submission_creative.csv\", header=[\"FIPS\",\"Result\"], index = False)\n",
     "\n",
     "# You may use pandas to generate a dataframe with FIPS and your predictions first \n",
     "# and then use to_csv to generate a CSV file."
diff --git a/CS 4780 Final Project Student Template.ipynb b/CS 4780 Final Project Student Template.ipynb
index c6aadc1..5c459fd 100755
--- a/CS 4780 Final Project Student Template.ipynb	
+++ b/CS 4780 Final Project Student Template.ipynb	
@@ -74,6 +74,7 @@
     "import numpy as np\n",
     "#import sklearn as sk\n",
     "from sklearn.preprocessing import StandardScaler\n",
+    "from sklearn.preprocessing import KBinsDiscretizer\n",
     "from sklearn.neighbors import KNeighborsClassifier\n",
     "from sklearn import svm\n",
     "from sklearn.ensemble import AdaBoostClassifier as ABC\n",
@@ -84,7 +85,6 @@
     "from sklearn.model_selection import GridSearchCV\n",
     "from sklearn.model_selection import KFold\n",
     "from sklearn.metrics import balanced_accuracy_score\n",
-    "from sklearn.preprocessing import KBinsDiscretizer\n",
     "from sklearn.compose import ColumnTransformer"
    ]
   },
@@ -159,7 +159,6 @@
     "\n",
     "#Trim the data of irrelevant fields and convert to numpy array\n",
     "data = data[:,4:]\n",
-    "#print(data.dtypes)\n",
     "\n",
     "#Load and format the test_2016_no_label.csv file\n",
     "test = pd.read_csv(\"test_2016_no_label.csv\", sep=',', header=0, encoding='unicode_escape', thousands=\",\")\n",
@@ -248,16 +247,16 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "knn avg weighted accuracy:  0.6203334398310165\n",
-      "dtc avg weighted accuracy:  0.7070879048578987\n",
-      "dtc_t avg weighted accuracy:  0.6931251192737173\n",
-      "adb_dtc avg weighted accuracy:  0.6961593601061467\n",
-      "svc avg weighted accuracy:  0.7004446186230044\n",
-      "lr avg weighted accuracy:  0.6590264805800705\n",
-      "adb_lr avg weighted accuracy:  0.6497330664169988\n",
-      "nbc avg weighted accuracy:  0.538413031775743\n",
-      "adb_nbc avg weighted accuracy:  0.5742813326608044\n",
-      "219.0\n"
+      "knn avg weighted accuracy:  0.6237076582105193\n",
+      "dtc avg weighted accuracy:  0.6979205324376133\n",
+      "dtc_t avg weighted accuracy:  0.6855368083208002\n",
+      "adb_dtc avg weighted accuracy:  0.700633705568728\n",
+      "svc avg weighted accuracy:  0.6971740145905911\n",
+      "lr avg weighted accuracy:  0.6592688655540591\n",
+      "adb_lr avg weighted accuracy:  0.6488225086599588\n",
+      "nbc avg weighted accuracy:  0.5358837627311691\n",
+      "adb_nbc avg weighted accuracy:  0.5720794507211426\n",
+      "250.0\n"
      ]
     }
    ],
@@ -350,69 +349,9 @@
     "You may follow the steps in part 2 again but making innovative changes like creating new features, using new training algorithms, etc. Make sure you explain everything clearly in part 3.2. Note that reaching the 75% creative baseline is only a small portion of this part. Any creative ideas will receive most points as long as they are reasonable and clearly explained."
    ]
   },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "# Make sure you comment your code clearly and you may refer to these comments in the part 3.2\n",
-    "# TODO\n",
-    "\n",
-    "dtc.fit(train, yTr)\n",
-    "dtc_preds = dtc.predict(valid)\n",
-    "print(\"dtc weighted accuracy: \", weighted_accuracy(dtc_preds,yV)) #does pretty well, even with no pruning\n",
-    "\n",
-    "dtc_g.fit(train, yTr)\n",
-    "dtc_g_preds = dtc.predict(valid)\n",
-    "print(\"dtc_g weighted accuracy: \", weighted_accuracy(dtc_g_preds,yV)) #does pretty well, even with no pruning\n",
-    "\n",
-    "adb_dtc.fit(train, yTr)\n",
-    "adb_dtc_preds = adb_dtc.predict(valid)\n",
-    "print(\"adb_dtc weighted accuracy: \", weighted_accuracy(adb_dtc_preds,yV))\n",
-    "\n",
-    "#best validation error so far\n",
-    "svc.fit(train, yTr)\n",
-    "svc_preds = svc.predict(valid)\n",
-    "print(\"svc weighted accuracy: \", weighted_accuracy(svc_preds,yV))\n",
-    "\n",
-    "#does not work very well; not really better than random\n",
-    "nbc.fit(train, yTr)\n",
-    "nbc_preds = nbc.predict(valid)\n",
-    "print(\"nbc weighted accuracy: \", weighted_accuracy(nbc_preds,yV))\n",
-    "\n",
-    "adb_nbc.fit(train, yTr)\n",
-    "adb_nbc_preds = adb_nbc.predict(valid)\n",
-    "print(\"adb_nbc weighted accuracy: \", weighted_accuracy(adb_nbc_preds,yV))\n",
-    "\n",
-    "#Split the data into a train and test set\n",
-    "np.random.shuffle(data)\n",
-    "    \n",
-    "results = np.array(df[:,-1],dtype='bool')\n",
-    "data = df[:,:6]\n",
-    "n_valid = int(len(data)/5)\n",
-    "train = data[:4*n_valid]\n",
-    "yTr = results[:4*n_valid]\n",
-    "valid = data[4*n_valid:]\n",
-    "    yV = results[4*n_valid:]\n",
-    "\n",
-    "n_valid = int(len(data)/5)\n",
-    "j = 100;\n",
-    "\n",
-    "#for k in range(5, 100):\n",
-    "sum_err = 0;\n",
-    "#knn = KNeighborsClassifier(n_neighbors = k)\n",
-    "for i in range(j):\n",
-    "    #randomize training and validation set\n",
-    "    np.random.shuffle(data)\n",
-    "\n",
-    "    #run algorithm\n",
-    "    knn.fit(train, yTr)\n",
-    "    knn_preds = knn.predict(valid)\n",
-    "    sum_err = sum_err + sum(knn_preds != yV)/n_valid"
-   ]
-  },
   {
    "cell_type": "code",
-   "execution_count": 34,
+   "execution_count": 58,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -420,97 +359,48 @@
     "\n",
     "### (3.1) Preprocessing and Feature Extraction ###\n",
     "\n",
-    "###(3.1.a) Load and format the training data set(s)###\n",
-    "data = pd.read_csv(\"train_2016.csv\", sep=',', header=0, encoding='unicode_escape', thousands=\",\")\n",
-    "data_2012 = pd.read_csv(\"train_2012.csv\", sep=\",\", header=0, encoding='unicode_escape', thousands=',')\n",
-    "graph = pd.read_csv(\"graph.csv\", sep=',')\n",
+    "###Helper functions to preprocess data###\n",
     "\n",
     "#Add binary features for the state that the county is in (where each binary feature represents one state)\n",
-    "\n",
-    "###2016###\n",
-    "data[\"State\"] = [x.strip()[-2:] for x in data['County']] #Extract state abbreviations from County\n",
-    "dummies = pd.get_dummies(data['State']) #Create state dummies from State\n",
-    "data = pd.concat([data,dummies],axis=1)\n",
-    "data = data.drop('State', axis=1)\n",
-    "\n",
-    "###2012###\n",
-    "data_2012[\"State\"] = [x.strip()[-2:] for x in data_2012['County']] #Extract state abbreviations from County\n",
-    "dummies_2012 = pd.get_dummies(data_2012['State']) #Create state dummies from State\n",
-    "data_2012 = pd.concat([data_2012,dummies_2012],axis=1)\n",
-    "data_2012 = data_2012.drop(\"State\", axis=1)\n",
-    "\n",
-    "#Define a function which will preprocess the data\n",
-    "def preprocess(data):\n",
-    "    #Normalize features\n",
-    "    #Standardize the data with mean 0 and s.d. 1\n",
+    "def get_states(X):\n",
+    "    X[\"State\"] = [x.strip()[-2:] for x in X['County']]\n",
+    "    dummies_test = pd.get_dummies(X['State'])\n",
+    "    X = pd.concat([X,dummies_test],axis=1)\n",
+    "    X = X.drop('State',axis=1)\n",
+    "    return X\n",
+    "\n",
+    "#Normalizes all variables from the data except for the binary state variables\n",
+    "def normalize(X):\n",
+    "    #Normalize features Standardize the data with mean 0 and s.d. 1\n",
     "    scalar = StandardScaler()\n",
-    "    _,d = np.shape(data)\n",
-    "    data_norm = data[:,[0,1,2,3,4,5,(d-3),(d-2),(d-1)]] #d-2 and d-1 store the graph score and number of known neighbors\n",
-    "    #data_raw_indexes = np.r_[6:(d-2)]\n",
-    "    data_bin = data[:,6:(d-3)]\n",
-    "    #print('untransformed data', data_bin)\n",
+    "    _,d = np.shape(X)\n",
+    "    data_norm = X[:,[0,1,2,3,4,5,(d-2),(d-1)]] #d-2 and d-1 store the graph score and number of known neighbors\n",
+    "    data_bin = X[:,6:(d-2)]\n",
     "    scalar.fit(data_norm)\n",
     "    data_norm = scalar.transform(data_norm)\n",
-    "    #print('transformed data', data_norm)\n",
-    "    data = np.hstack((data_norm, data_bin))\n",
-    "    #print(data)\n",
-    "\n",
-    "    #scalar = StandardScaler()\n",
-    "    #binning = ['MedianIncome']\n",
-    "    #standardization = ['MigraRate','BirthRate','DeathRate','BachelorRate','UnemploymentRate']\n",
-    "    #n,d = np.shape(counties)\n",
-    "    #print(d)\n",
-    "    #binning = [4]\n",
-    "    #standardization = [*range(5,d)]\n",
-    "    #print(standardization)\n",
-    "    #ct = ColumnTransformer([(\"standardize\", scalar, standardization), \n",
-    "     #                       (\"binning\", KBinsDiscretizer(n_bins=10), binning)])\n",
-    "    #try min maxing if binning doesn't \n",
-    "    #data = ct.fit_transform(counties)\n",
-    "    return data"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "###EXPERIMENTING DON'T RUN YET###\n",
-    "\n",
-    "#Add binary feature for whether a county is on a state border (which may indicate they receive other states' political ads)\n",
-    "fips_allstates = data[[\"FIPS\",\"State\"]]\n",
-    "tmp1 = graph.merge(fips_allstates, left_on=\"SRC\", right_on=\"FIPS\", how=\"left\")\n",
-    "tmp1 = tmp1[[\"SRC\",\"DST\",\"State\"]]\n",
-    "tmp1.rename(columns={'State':'SRC_State'}, inplace=True)\n",
-    "print(tmp1)\n",
-    "\n",
-    "tmp2 = tmp1.merge(fips_allstates, left_on=\"DST\", right_on=\"FIPS\", how=\"left\")\n",
-    "tmp2 = tmp2[[\"SRC\",\"DST\",\"SRC_State\",\"State\"]]\n",
-    "tmp2.rename(columns={'State':'DST_State'}, inplace=True)\n",
-    "print(tmp2)\n",
+    "    X = np.hstack((data_norm, data_bin))\n",
+    "    return X\n",
     "\n",
-    "tmp2[\"SRConBorder\"] = tmp2[\"SRC_State\"].eq(tmp2[\"DST_State\"])\n",
-    "tmp2[\"SRConBorder\"] = tmp2[\"SRConBorder\"].astype(int)\n",
-    "tmp2[\"SRConBorder\"] = tmp2[\"SRConBorder\"].replace({0:1,1:0}) #Make sure =1 iff on border\n",
-    "print(tmp2)\n",
-    "tmp2 = tmp2.dropna() #Because NaNs indicate counties that are not in training data - not sure how to deal with loss of data\n",
-    "tmp2 = tmp2.groupby(['SRC'])['SRConBorder'].sum()\n",
-    "print(tmp2)\n",
+    "###(3.1.a) Load and format the training data set(s)###\n",
+    "data = pd.read_csv(\"train_2016.csv\", sep=',', header=0, encoding='unicode_escape', thousands=\",\")\n",
+    "data_2012 = pd.read_csv(\"train_2012.csv\", sep=\",\", header=0, encoding='unicode_escape', thousands=',')\n",
+    "graph = pd.read_csv(\"graph.csv\", sep=',', header=0, encoding='unicode_escape', thousands=\",\")\n",
     "\n",
-    "data = data.merge(tmp2, left_on=\"FIPS\", right_on=\"SRC\", how=\"left\")\n",
-    "print(data)\n",
+    "###2016###\n",
+    "data = get_states(data)\n",
     "\n",
-    "#Notes: Source counties can be found in test, destination in train (or vice versa - one occurence in either column(?))\n",
-    "#Notes: Get list of neighbors that voted one way or other (check one per column)"
+    "###2012###\n",
+    "data_2012 = get_states(data_2012)\n",
+    "\n"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 35,
+   "execution_count": 59,
    "metadata": {},
    "outputs": [],
    "source": [
     "#Creates a lexicographically ordered list of all the county's neighbors\n",
-    "graph = pd.read_csv(\"graph.csv\", sep=',', header=0, encoding='unicode_escape', thousands=\",\")\n",
     "neighbors = {}\n",
     "for i in range(len(graph)):\n",
     "    src = graph['SRC'][i]\n",
@@ -531,15 +421,12 @@
     "dem = dict(zip(data['FIPS'], dem_perc))\n",
     "votes = data[['DEM','GOP']]\n",
     "votes = list(votes.itertuples(index=False, name=None))\n",
-    "votes = dict(zip(data['FIPS'], votes))\n",
-    "votes_2012 = data_2012[['DEM','GOP']]\n",
-    "votes_2012 = list(votes_2012.itertuples(index=False, name=None))\n",
-    "votes_2012 = dict(zip(data['FIPS'], votes_2012))"
+    "votes = dict(zip(data['FIPS'], votes))"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 36,
+   "execution_count": 60,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -554,7 +441,7 @@
     "            score += dem[i]\n",
     "    return n, score/n if n != 0 else np.nan\n",
     "\n",
-    "#TODO - implement a function that calculates the percentage of all voters in the surrounding counties who voted dem\n",
+    "#Calculates the percentage of all voters from the neighboring counties who voted democratic\n",
     "def aggregate_gscore(neighs, votes):\n",
     "    n = 0\n",
     "    dem = 0\n",
@@ -569,40 +456,45 @@
     "#Function that takes a list of counties and appends their graph_score to the end of their vector:\n",
     "def add_gscore(counties):\n",
     "    counties['GraphScore'] = 0\n",
-    "    counties['2012GScore'] = 0\n",
     "    counties['Neighbors'] = 0\n",
     "    for i in range(len(counties)):\n",
     "        key = int(counties['FIPS'][i])\n",
     "        if neighbors.get(key) == None:\n",
     "            score = np.nan\n",
-    "            s2012 = np.nan\n",
     "            n = 0\n",
     "        else:\n",
     "            n, score = aggregate_gscore(neighbors[key], votes)\n",
-    "            _, s2012 = aggregate_gscore(neighbors[key], votes_2012)\n",
+    "            total_n = len(neighbors[key])\n",
     "        counties.loc[i, 'GraphScore'] = score\n",
-    "        counties.loc[i, '2012GScore'] = s2012\n",
     "        counties.loc[i, 'Neighbors'] = n\n",
     "        smean = counties['GraphScore'].sum()/len(counties)\n",
-    "        s2012_mean = counties['2012GScore'].sum()/len(counties)\n",
     "    counties['GraphScore'] = counties['GraphScore'].map(lambda x: smean if pd.isnull(x) else x)\n",
-    "    counties['2012GScore'] = counties['2012GScore'].map(lambda x: s2012_mean if pd.isnull(x) else x)\n",
     "    return counties"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 37,
+   "execution_count": 61,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "['FIPS', 'County', 'DEM', 'GOP', 'MedianIncome', 'MigraRate', 'BirthRate', 'DeathRate', 'BachelorRate', 'UnemploymentRate', 'AL', 'AR', 'AZ', 'CA', 'CO', 'CT', 'DC', 'DE', 'FL', 'GA', 'HI', 'IA', 'ID', 'IL', 'IN', 'KS', 'KY', 'LA', 'MA', 'MD', 'ME', 'MI', 'MN', 'MO', 'MS', 'MT', 'NC', 'ND', 'NE', 'NH', 'NJ', 'NM', 'NV', 'NY', 'OH', 'OK', 'OR', 'PA', 'RI', 'SC', 'SD', 'TN', 'TX', 'UT', 'VA', 'VT', 'WA', 'WI', 'WV', 'WY', 'GraphScore', 'Neighbors']\n",
+      "['FIPS', 'County', 'DEM', 'GOP', 'MedianIncome', 'MigraRate', 'BirthRate', 'DeathRate', 'BachelorRate', 'UnemploymentRate', 'AL', 'AR', 'AZ', 'CA', 'CO', 'CT', 'DC', 'DE', 'FL', 'GA', 'HI', 'IA', 'ID', 'IL', 'IN', 'KS', 'KY', 'LA', 'MA', 'MD', 'ME', 'MI', 'MN', 'MO', 'MS', 'MT', 'NC', 'ND', 'NE', 'NH', 'NJ', 'NM', 'NV', 'NY', 'OH', 'OK', 'OR', 'PA', 'RI', 'SC', 'SD', 'TN', 'TX', 'UT', 'VA', 'VT', 'WA', 'WI', 'WV', 'WY', 'GraphScore', 'Neighbors']\n"
+     ]
+    }
+   ],
    "source": [
     "###Prep for concatenating###\n",
     "\n",
     "###2016###\n",
-    "#cols = data.columns.tolist()\n",
-    "#cols = cols[0:2] + cols[4:] #Helps make sure columns in test align with those in data\n",
-    "\n",
     "data = add_gscore(data)\n",
+    "#Rearrange the columns to ensure that DC is in the same location of each dataset\n",
+    "cols = data.columns.tolist()\n",
+    "cols = cols[0:2] + cols[4:]\n",
+    "print(list(data.columns))\n",
     "data = data.to_numpy()\n",
     "\n",
     "#These are the training labels for each county, 1 if the county voted Dem and 0 if it voted Rep\n",
@@ -611,11 +503,13 @@
     "\n",
     "#Trim the data of irrelevant fields and convert to numpy array\n",
     "data = data[:,4:]\n",
-    "#print(np.shape(data))\n",
-    "#print(data)\n",
     "\n",
     "###2012###\n",
     "data_2012 = add_gscore(data_2012)\n",
+    "#Rearrange the columns to ensure that DC is in the same location of each dataset\n",
+    "cols_2012 = data_2012.columns.tolist()\n",
+    "cols_2012 = cols_2012[0:2] + cols_2012[4:]\n",
+    "print(list(data_2012.columns))\n",
     "data_2012 = data_2012.to_numpy()\n",
     "\n",
     "#These are the training labels for each county, 1 if the county voted Dem and 0 if it voted Rep\n",
@@ -627,9 +521,18 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 38,
+   "execution_count": 62,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "['FIPS', 'County', 'MedianIncome', 'MigraRate', 'BirthRate', 'DeathRate', 'BachelorRate', 'UnemploymentRate', 'AL', 'AR', 'AZ', 'CA', 'CO', 'CT', 'DC', 'DE', 'FL', 'GA', 'HI', 'IA', 'ID', 'IL', 'IN', 'KS', 'KY', 'LA', 'MA', 'MD', 'ME', 'MI', 'MN', 'MO', 'MS', 'MT', 'NC', 'ND', 'NE', 'NH', 'NJ', 'NM', 'NV', 'NY', 'OH', 'OK', 'OR', 'PA', 'RI', 'SC', 'SD', 'TN', 'TX', 'UT', 'VA', 'VT', 'WA', 'WI', 'WV', 'WY', 'GraphScore', 'Neighbors']\n",
+      "['FIPS', 'County', 'MedianIncome', 'MigraRate', 'BirthRate', 'DeathRate', 'BachelorRate', 'UnemploymentRate', 'AL', 'AR', 'AZ', 'CA', 'CO', 'CT', 'DC', 'DE', 'FL', 'GA', 'HI', 'IA', 'ID', 'IL', 'IN', 'KS', 'KY', 'LA', 'MA', 'MD', 'ME', 'MI', 'MN', 'MO', 'MS', 'MT', 'NC', 'ND', 'NE', 'NH', 'NJ', 'NM', 'NV', 'NY', 'OH', 'OK', 'OR', 'PA', 'RI', 'SC', 'SD', 'TN', 'TX', 'UT', 'VA', 'VT', 'WA', 'WI', 'WV', 'WY', 'GraphScore', 'Neighbors']\n"
+     ]
+    }
+   ],
    "source": [
     "###(3.1.b) Load and format the testing data sets###\n",
     "test = pd.read_csv(\"test_2016_no_label.csv\", sep=',', header=0, encoding='unicode_escape', thousands=\",\")\n",
@@ -638,15 +541,11 @@
     "#Process the testing data set in the same manner as the training data set\n",
     "\n",
     "###2016###\n",
-    "test[\"State\"] = [x.strip()[-2:] for x in test['County']]\n",
-    "dummies_test = pd.get_dummies(test['State'])\n",
-    "#print(dummies.columns.difference(dummies_test.columns)) #Check that DC is the only missing county/state\n",
-    "test = pd.concat([test,dummies_test],axis=1)\n",
+    "test = get_states(test)\n",
     "test['DC'] = 0 #Add in a zero vector for DC (because there is no DC in either test data set)\n",
-    "test = test.drop('State',axis=1)\n",
-    "#print(test)\n",
-    "#test = test[cols] #Make sure columns in test align with those in data\n",
     "test = add_gscore(test)\n",
+    "test = test[cols]\n",
+    "print(list(test.columns))\n",
     "test = test.to_numpy()\n",
     "\n",
     "fips = test[:,0]\n",
@@ -656,21 +555,19 @@
     "data = np.vstack((data, test))\n",
     "\n",
     "#Standardize the (non-binary) features to have mean 0 and s.d. 1\n",
-    "data = preprocess(data)\n",
+    "data = normalize(data)\n",
     "\n",
     "test = data[-n_data:,:]\n",
     "data = data[:n_data,:]\n",
     "data = np.hstack((data, np.transpose(np.array([results]))))\n",
     "\n",
     "###2012###\n",
-    "test_2012[\"State\"] = [x.strip()[-2:] for x in test_2012['County']]\n",
-    "dummies_test_2012 = pd.get_dummies(test_2012['State'])\n",
-    "#print(dummies_2012.columns.difference(dummies_test_2012.columns)) #Check that DC is the only missing county/state\n",
-    "test_2012 = pd.concat([test_2012,dummies_test_2012],axis=1)\n",
+    "test_2012 = get_states(test_2012)\n",
     "test_2012['DC'] = 0 #Add in a zero vector for DC (because there is no DC in either test data set)\n",
-    "test_2012 = test_2012.drop('State',axis=1)\n",
+    "test_2012 = add_gscore(test_2012)\n",
+    "test_2012 = test_2012[cols_2012]\n",
+    "print(list(test_2012.columns))\n",
     "\n",
-    "add_gscore(test_2012)\n",
     "test_2012 = test_2012.to_numpy()\n",
     "\n",
     "fips_2012 = test_2012[:,0]\n",
@@ -679,7 +576,7 @@
     "n_data_2012 = len(data_2012)\n",
     "data_2012 = np.vstack((data_2012, test_2012))\n",
     "\n",
-    "data_2012 = preprocess(data_2012)\n",
+    "data_2012 = normalize(data_2012)\n",
     "test_2012 = data_2012[-n_data_2012:,:]\n",
     "data_2012 = data_2012[:n_data_2012,:]\n",
     "data_2012 = np.hstack((data_2012, np.transpose(np.array([results_2012]))))"
@@ -687,9 +584,20 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 39,
+   "execution_count": 67,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "[ 1.25892541  1.49170935  1.76753662  2.09436625  2.48162892  2.94049911\n",
+      "  3.48421754  4.12847324  4.89185622  5.79639395  6.86818691  8.13816172\n",
+      "  9.64296358 11.42601361 13.5387618  16.04217161 19.00847905 22.52327705\n",
+      " 26.68798528 31.6227766 ]\n"
+     ]
+    }
+   ],
    "source": [
     "### (3.2) Testing algorithms ###\n",
     "\n",
@@ -704,7 +612,9 @@
     "adb_dtc_g = ABC(base_estimator = DTC(criterion = \"entropy\"), n_estimators = 100)\n",
     "\n",
     "#SVM classifier\n",
-    "Cs = np.logspace(1.25,1.25,20)\n",
+    "Cs = np.logspace(.1,1.5,20)\n",
+    "print(Cs)\n",
+    "#Cs = np.logspace(1.25,1.25,20)\n",
     "svc = svm.SVC(gamma='scale', kernel='rbf')\n",
     "c_svc = GridSearchCV(estimator=svc,param_grid=dict(C=Cs))\n",
     "\n",
@@ -728,18 +638,18 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "knn avg weighted accuracy:  0.6307393036738488\n",
-      "knn avg weighted accuracy 2012:  0.6547367701142874\n",
-      "dtc_t avg weighted accuracy:  0.745723590078064\n",
-      "dtc_t avg weighted accuracy 2012:  0.6877354733446237\n",
-      "adb_dtc avg weighted accuracy:  0.7285094502910128\n",
-      "adb_dtc avg weighted accuracy 2012:  0.7070487212085304\n",
-      "dtc_g avg weighted accuracy:  0.7504774822113874\n",
-      "dtc_g avg weighted accuracy 2012:  0.6999027843035145\n",
-      "adb_dtc_g avg weighted accuracy:  0.7373869781842327\n",
-      "adb_dtc_g avg weighted accuracy 2012:  0.6992326684962192\n",
-      "svc avg weighted accuracy:  0.7840850521161004\n",
-      "svc avg weighted accuracy 2012:  0.7782245706696347\n"
+      "knn avg weighted accuracy:  0.6330706501088754\n",
+      "knn avg weighted accuracy 2012:  0.6287551049058671\n",
+      "dtc_t avg weighted accuracy:  0.7126628181827105\n",
+      "dtc_t avg weighted accuracy 2012:  0.6654573475411754\n",
+      "adb_dtc avg weighted accuracy:  0.7583268043136182\n",
+      "adb_dtc avg weighted accuracy 2012:  0.6876199289767525\n",
+      "dtc_g avg weighted accuracy:  0.7331652849631233\n",
+      "dtc_g avg weighted accuracy 2012:  0.6960158238672269\n",
+      "adb_dtc_g avg weighted accuracy:  0.7526903758034124\n",
+      "adb_dtc_g avg weighted accuracy 2012:  0.6782594899860952\n",
+      "svc avg weighted accuracy:  0.7850360498781626\n",
+      "svc avg weighted accuracy 2012:  0.7207312417562269\n"
      ]
     }
    ],
@@ -835,33 +745,30 @@
     "print(\"svc avg weighted accuracy: \", x)\n",
     "print(\"svc avg weighted accuracy 2012: \", x_2012)\n",
     "\n",
-    "#print(\"lr avg weighted accuracy: \", kfold(lr, data))\n",
-    "#print(\"lr avg weighted accuracy: \", kfold(c_lr, data, grid=\"Yes\", grid_model=\"lr\")) #.721 using Cs = np.logspace(-1,1,100) and solver = 'liblinear'\n",
-    "#print(\"lr avg weighted accuracy 2012: \", kfold(c_lr, data_2012, grid=\"Yes\", grid_model=\"lr\"))\n",
     "x_lr,y_lr,z_lr = kfold_lr(c_lr, data)\n",
     "x_lr_2012,y_lr_2012,z_lr_2012 = kfold_lr(c_lr, data_2012)\n",
     "print(\"lr avg weighted accuracy: \", x_lr)\n",
     "print(\"lr avg weighted accuracy 2012: \", x_lr_2012)\n",
     "\n",
-    "#print(\"adb_lr avg weighted accuracy: \", kfold(adb_lr, data)) #.713 using solver = 'liblinear'\n",
-    "#print(\"adb_lr avg weighted accuracy 2012: \", kfold(adb_lr, data_2012)) #.713 using solver = 'liblinear'\n",
-    "\n",
-    "#print(\"nbc avg weighted accuracy: \", kfold(nbc, data)) #.653\n",
-    "\n",
-    "#print(\"adb_nbc avg weighted accuracy: \", kfold(adb_nbc, data)) #.5 using n_estimators = 100\n",
-    "\n",
     "#Get final predictions\n",
-    "preds = z.predict(test) #Uses 2016"
+    "preds_creative = z.predict(test) #Uses 2016"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 65,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "235\n"
+     ]
+    }
+   ],
    "source": [
-    "print(sum(preds))\n",
-    "print(sum(z.predict(data[:,:-1])))"
+    "print(sum(preds_creative))"
    ]
   },
   {
@@ -882,6 +789,13 @@
     "3.2.2 Please explain in detail how you achieved this and what you did specifically and why you tried this."
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "<img src=\"files/Kaggle (Creative) Solution.png\">"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -902,16 +816,27 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 56,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "228\n"
+     ]
+    }
+   ],
    "source": [
     "# TODO\n",
     "output = np.array([fips, preds])\n",
     "output = pd.DataFrame(np.transpose(output))\n",
     "output[1] = output[1].astype(int)\n",
-    "#output.to_csv(\"submission.csv\", header=[\"FIPS\",\"Result\"], index = False)\n",
-    "output.to_csv(\"submission_creative.csv\", header=[\"FIPS\",\"Result\"], index = False)\n",
+    "output_creative = np.array([fips, preds_creative])\n",
+    "output_creative = pd.DataFrame(np.transpose(output_creative))\n",
+    "output_creative[1] = output_creative[1].astype(int)\n",
+    "output.to_csv(\"submission.csv\", header=[\"FIPS\",\"Result\"], index = False)\n",
+    "output_creative.to_csv(\"submission_creative.csv\", header=[\"FIPS\",\"Result\"], index = False)\n",
     "\n",
     "# You may use pandas to generate a dataframe with FIPS and your predictions first \n",
     "# and then use to_csv to generate a CSV file."
diff --git a/submission.csv b/submission.csv
index a59fce4..e2da191 100644
--- a/submission.csv
+++ b/submission.csv
@@ -6,21 +6,21 @@ FIPS,Result
 39039,0
 30083,0
 46047,0
-45013,1
+45013,0
 18119,0
 48333,0
-53005,1
+53005,0
 5097,0
 48297,0
 51141,0
 54107,0
-12103,1
+12103,0
 51077,0
 25015,1
 47043,0
 48363,0
 31077,0
-21043,1
+21043,0
 28047,0
 19193,0
 18165,0
@@ -30,33 +30,33 @@ FIPS,Result
 21229,0
 17097,1
 29159,0
-28149,1
+28149,0
 48467,0
 5053,0
 31093,0
 19143,0
 13181,0
 47127,0
-21153,0
+21153,1
 20143,0
 22119,0
 19119,0
 47083,0
 42095,0
-6081,1
+6081,0
 13313,0
 5123,0
 28039,0
-51760,0
+51760,1
 48245,0
-41053,1
+41053,0
 8025,0
 5079,0
 49043,0
-48489,0
+48489,1
 22057,0
 1021,0
-29185,1
+29185,0
 29211,0
 54081,0
 6003,0
@@ -67,7 +67,7 @@ FIPS,Result
 37097,0
 48283,0
 46083,0
-31055,1
+31055,0
 51580,0
 37157,0
 42033,0
@@ -76,7 +76,7 @@ FIPS,Result
 13075,0
 21123,0
 48185,0
-13245,1
+13245,0
 13001,0
 31139,0
 38025,0
@@ -90,25 +90,25 @@ FIPS,Result
 13183,0
 53035,1
 48479,1
-21063,0
+21063,1
 20101,0
 48415,0
 51185,0
-51043,1
-36013,1
+51043,0
+36013,0
 55119,0
-45075,1
+45075,0
 47101,0
 26089,1
 48367,0
-37131,0
-21205,1
+37131,1
+21205,0
 47179,0
 48093,0
-5101,1
+5101,0
 1093,0
 31101,0
-5069,1
+5069,0
 8119,0
 54037,0
 13213,0
@@ -121,39 +121,39 @@ FIPS,Result
 19045,0
 21167,0
 31127,0
-5015,1
+5015,0
 51159,0
-28133,0
+28133,1
 5045,0
 39069,0
-48381,1
+48381,0
 13129,0
 37041,0
 51057,0
 25005,0
 22059,0
-38017,1
+38017,0
 26109,0
-49037,0
+49037,1
 48265,0
 39129,0
 20055,0
-21121,1
+21121,0
 31173,0
-22035,1
+22035,0
 8011,0
 39027,0
 17107,0
-42021,1
+42021,0
 37023,0
 54003,0
 31097,0
 17141,0
 48203,0
 20011,0
-23031,1
+23031,0
 16027,0
-48003,1
+48003,0
 51600,0
 13027,0
 21021,0
@@ -168,7 +168,7 @@ FIPS,Result
 36015,0
 1063,1
 13279,0
-42125,1
+42125,0
 42105,0
 20207,0
 13285,0
@@ -196,7 +196,7 @@ FIPS,Result
 46119,0
 12045,0
 40141,0
-41015,1
+41015,0
 48129,0
 29083,0
 17117,0
@@ -205,7 +205,7 @@ FIPS,Result
 27009,0
 19195,0
 19039,0
-18005,1
+18005,0
 37083,0
 39077,0
 19147,0
@@ -227,12 +227,12 @@ FIPS,Result
 26121,0
 21203,0
 21211,0
-44005,1
+44005,0
 35005,0
 30017,0
 31107,0
 20003,0
-22103,1
+22103,0
 5117,0
 47107,0
 20047,0
@@ -241,15 +241,15 @@ FIPS,Result
 54027,0
 13147,0
 30055,0
-20203,1
+20203,0
 1051,0
 18153,0
 29043,0
-8065,0
+8065,1
 29135,0
 31083,0
 17065,0
-12097,1
+12097,0
 28111,0
 1113,0
 48501,0
@@ -261,7 +261,7 @@ FIPS,Result
 12125,0
 51073,0
 47149,0
-46031,1
+46031,0
 40085,0
 40127,0
 28015,0
@@ -274,8 +274,8 @@ FIPS,Result
 13319,0
 20135,0
 48021,0
-28121,1
-51678,1
+28121,0
+51678,0
 12027,0
 30005,0
 17113,1
@@ -287,26 +287,26 @@ FIPS,Result
 27111,0
 38039,0
 20153,0
-48361,1
+48361,0
 20195,0
 40101,0
 21199,0
 55117,0
-34039,0
+34039,1
 16019,0
 30009,0
 19155,0
 29105,0
 23009,0
 30101,0
-8059,0
+8059,1
 17069,0
 29175,0
 31051,0
 56039,1
-26131,1
+26131,0
 48177,0
-8027,1
+8027,0
 23007,0
 51145,0
 51087,1
@@ -333,9 +333,9 @@ FIPS,Result
 26081,0
 26155,0
 17143,1
-53067,1
+53067,0
 21105,0
-24009,0
+24009,1
 1025,0
 27067,0
 19075,0
@@ -345,21 +345,21 @@ FIPS,Result
 1105,1
 16071,0
 17199,0
-48061,0
+48061,1
 1039,0
 8097,1
 27091,0
 42099,0
 42117,0
-13077,1
+13077,0
 51520,0
 36051,0
 54025,0
-30049,1
+30049,0
 1079,0
 37117,0
 48133,0
-42041,1
+42041,0
 49007,0
 47143,0
 55039,0
@@ -370,16 +370,16 @@ FIPS,Result
 53041,0
 31169,0
 37185,0
-26099,1
+26099,0
 47155,0
 6111,1
 29029,0
 13081,0
 5111,0
-51133,1
-6057,1
+51133,0
+6057,0
 21225,0
-48141,0
+48141,1
 28073,0
 27135,0
 37129,0
@@ -390,8 +390,8 @@ FIPS,Result
 46129,0
 28003,0
 1123,0
-39017,1
-6097,0
+39017,0
+6097,1
 4021,0
 38031,0
 48405,0
@@ -414,13 +414,13 @@ FIPS,Result
 36117,0
 48043,1
 45063,0
-16057,0
+16057,1
 48157,0
 12073,1
-41067,0
+41067,1
 29187,0
 8049,0
-40133,1
+40133,0
 37183,1
 54005,0
 18045,0
@@ -430,9 +430,9 @@ FIPS,Result
 54057,0
 21065,0
 48341,0
-47077,1
-28135,0
-1013,0
+47077,0
+28135,1
+1013,1
 40145,0
 1111,0
 37111,0
@@ -440,7 +440,7 @@ FIPS,Result
 12129,0
 17103,0
 1077,0
-45089,1
+45089,0
 51187,0
 13013,0
 31087,0
@@ -448,11 +448,11 @@ FIPS,Result
 13293,0
 48461,0
 28107,0
-8041,1
+8041,0
 5049,0
 27073,0
-36055,0
-13259,1
+36055,1
+13259,0
 51155,0
 35001,1
 47025,0
@@ -462,11 +462,11 @@ FIPS,Result
 48199,0
 19037,0
 13311,0
-51165,1
-17077,0
+51165,0
+17077,1
 29001,0
 54011,0
-56035,1
+56035,0
 51135,0
 25001,0
 21175,0
@@ -492,14 +492,14 @@ FIPS,Result
 6089,0
 13227,0
 51011,0
-36113,1
+36113,0
 54061,1
 18135,0
 21221,0
 17011,0
-29169,1
+29169,0
 29081,0
-12087,1
+12087,0
 17001,0
 21117,0
 47035,0
@@ -508,7 +508,7 @@ FIPS,Result
 28069,0
 56015,0
 21159,0
-17203,1
+17203,0
 37135,1
 5011,0
 28151,1
@@ -521,12 +521,12 @@ FIPS,Result
 20021,0
 26129,0
 48507,1
-21037,1
+21037,0
 13083,0
 17137,0
 48053,0
 18041,0
-17167,0
+17167,1
 42023,0
 42081,0
 17041,0
@@ -546,9 +546,9 @@ FIPS,Result
 20023,0
 26107,0
 13255,0
-5093,0
-53063,1
-45067,1
+5093,1
+53063,0
+45067,0
 46123,0
 38007,0
 29117,0
@@ -556,9 +556,9 @@ FIPS,Result
 37139,0
 29127,0
 13123,0
-13283,1
+13283,0
 31073,0
-51103,1
+51103,0
 36119,1
 13229,0
 48281,0
@@ -571,10 +571,10 @@ FIPS,Result
 17179,0
 1055,0
 55141,0
-25007,0
+25007,1
 24015,0
 29227,0
-48031,1
+48031,0
 20175,0
 41059,0
 27087,0
@@ -601,7 +601,7 @@ FIPS,Result
 21049,0
 54019,0
 21135,0
-29223,1
+29223,0
 24031,1
 55109,0
 55001,0
@@ -609,13 +609,13 @@ FIPS,Result
 55053,0
 47011,0
 13009,0
-13003,1
+13003,0
 48471,0
 12085,0
 21047,0
 51685,0
 5139,0
-48271,1
+48271,0
 56007,0
 54015,0
 6115,0
@@ -658,36 +658,36 @@ FIPS,Result
 31069,0
 26037,0
 23003,0
-13063,0
+13063,1
 19109,0
 42079,0
 39107,0
 19103,1
 39047,0
-50005,0
+50005,1
 47051,0
 39105,0
 38091,0
-54089,1
+54089,0
 42051,0
 51081,0
-22055,0
+22055,1
 51029,0
 51093,0
 22025,0
 51750,1
-35011,1
-49005,1
+35011,0
+49005,0
 48197,0
 31125,0
 1107,0
 45085,0
-51061,1
+51061,0
 13091,0
 5131,0
 49045,0
 39159,0
-19053,1
+19053,0
 37193,0
 13149,0
 18105,1
@@ -695,17 +695,17 @@ FIPS,Result
 20089,0
 28099,0
 48373,0
-51157,1
+51157,0
 31081,0
 27125,0
-1035,1
-13163,1
+1035,0
+13163,0
 31057,0
 13257,0
 18109,0
 28053,1
 13211,0
-22017,1
+22017,0
 12039,0
 13043,0
 18075,0
@@ -715,19 +715,19 @@ FIPS,Result
 29033,0
 24047,0
 6085,1
-45079,0
-46011,1
+45079,1
+46011,0
 48145,0
 49039,0
 45051,0
 39113,0
-12086,0
+12086,1
 21223,0
 39057,0
 18157,0
 18133,0
 28021,1
-53061,0
+53061,1
 6019,1
 45087,0
 37085,0
@@ -738,21 +738,21 @@ FIPS,Result
 5003,0
 18029,0
 48231,0
-24043,1
+24043,0
 51017,0
-8021,1
+8021,0
 27053,1
 18145,0
 37031,0
-18087,1
+18087,0
 5063,0
 31017,0
 51710,1
 45047,0
 30039,0
 31035,0
-13039,1
-12061,1
+13039,0
+12061,0
 18123,0
 39119,0
 26071,0
@@ -763,7 +763,7 @@ FIPS,Result
 25021,1
 42133,0
 8077,0
-13021,1
+13021,0
 6075,1
 37093,0
 37055,0
@@ -785,21 +785,21 @@ FIPS,Result
 51015,0
 45071,0
 19031,0
-31109,1
+31109,0
 55137,0
 39063,0
 5017,0
 54035,0
-26061,0
+26061,1
 28101,0
 8087,0
 31043,0
 21039,0
 39009,0
-13269,1
+13269,0
 30067,0
 48445,0
-41017,1
+41017,0
 41057,0
 48473,0
 51021,0
@@ -812,29 +812,29 @@ FIPS,Result
 28093,0
 48171,0
 48139,0
-50017,0
+50017,1
 38081,0
-35017,1
+35017,0
 51117,0
-51161,1
+51161,0
 19157,0
 26157,0
 47135,0
-13153,1
+13153,0
 48431,0
 20075,0
 20069,0
 38051,0
-13061,1
+13061,0
 16053,0
-50025,0
+50025,1
 47085,0
-9007,0
+9007,1
 27061,0
 27081,0
 51153,1
-46071,1
-50015,0
+46071,0
+50015,1
 39061,0
 19035,0
 1133,0
@@ -842,18 +842,18 @@ FIPS,Result
 19145,0
 36043,0
 42039,0
-19113,1
-48247,0
+19113,0
+48247,1
 21171,0
 47133,0
 46097,0
 35013,1
-22071,0
+22071,1
 22015,0
 35019,0
 37145,0
-37081,0
-41069,1
+37081,1
+41069,0
 29189,1
 4015,0
 21227,0
@@ -864,14 +864,14 @@ FIPS,Result
 37059,0
 42111,0
 36017,0
-35053,1
+35053,0
 48067,0
 18057,1
 6101,0
 48465,0
 51740,0
 20037,0
-17093,0
+17093,1
 40055,0
 26117,0
 27071,0
@@ -889,25 +889,25 @@ FIPS,Result
 54013,0
 26073,0
 48291,0
-51660,0
+51660,1
 39007,0
 37077,0
 41003,1
 26125,1
 25017,1
-17099,1
+17099,0
 37013,0
 20013,0
 27107,0
 28077,0
-21235,1
+21235,0
 31071,0
 19027,0
 28159,0
 39109,0
 6041,1
 30077,0
-48419,1
+48419,0
 20073,0
 37067,0
 55013,0
@@ -924,7 +924,7 @@ FIPS,Result
 8121,0
 19167,0
 18049,0
-1125,1
+1125,0
 20169,0
 36003,0
 20129,0
@@ -936,9 +936,9 @@ FIPS,Result
 48039,0
 20139,0
 12095,0
-51650,1
+51650,0
 1087,1
-54069,1
+54069,0
 39133,0
 21001,0
 17073,0
@@ -948,7 +948,7 @@ FIPS,Result
 40137,0
 40047,0
 48435,0
-49017,1
+49017,0
 40105,0
 30027,0
 26025,0
@@ -957,13 +957,13 @@ FIPS,Result
 17071,0
 47041,0
 21067,1
-12065,1
-42043,1
+12065,0
+42043,0
 19187,0
 27093,0
 48045,0
 21003,0
-45033,1
+45033,0
 55055,0
 8091,0
 27063,0
@@ -974,27 +974,27 @@ FIPS,Result
 22001,0
 20017,0
 13029,0
-48323,0
+48323,1
 46085,0
 54093,0
-36019,0
+36019,1
 19165,0
 45003,0
 51810,1
-26053,1
+26053,0
 22045,0
 38003,0
-12099,1
+12099,0
 40115,0
-1091,0
+1091,1
 51023,0
-13085,1
+13085,0
 36007,0
 45041,0
 22023,0
 51053,0
 40131,0
-6001,0
+6001,1
 19091,0
 19097,0
 13131,0
@@ -1014,12 +1014,12 @@ FIPS,Result
 20137,0
 13189,0
 6055,1
-22065,1
+22065,0
 55033,0
-1081,0
+1081,1
 31001,0
 25009,1
-13215,0
+13215,1
 20131,0
 8073,0
 17039,0
@@ -1033,7 +1033,7 @@ FIPS,Result
 9015,0
 4017,0
 29145,0
-5099,1
+5099,0
 39147,0
 48261,0
 28071,1
@@ -1049,8 +1049,8 @@ FIPS,Result
 47073,0
 13071,0
 37091,0
-27045,1
-8013,0
+27045,0
+8013,1
 13197,0
 32005,0
 17161,0
@@ -1068,7 +1068,7 @@ FIPS,Result
 23011,0
 17193,0
 22005,0
-48249,0
+48249,1
 34031,1
 18183,0
 51001,0
@@ -1095,17 +1095,17 @@ FIPS,Result
 19169,1
 49051,0
 55125,0
-26145,1
+26145,0
 4009,0
-47037,1
-36077,1
+47037,0
+36077,0
 54105,0
 27017,0
 12053,0
 39059,0
 21095,0
 32023,0
-13199,0
+13199,1
 55069,0
 47161,0
 29015,0
@@ -1124,7 +1124,7 @@ FIPS,Result
 46043,0
 18003,0
 17013,0
-51700,0
+51700,1
 29097,0
 54021,0
 55093,0
@@ -1132,7 +1132,7 @@ FIPS,Result
 5085,0
 46015,0
 6043,0
-13235,1
+13235,0
 48025,0
 18013,0
 35006,0
@@ -1144,7 +1144,7 @@ FIPS,Result
 48075,0
 26147,0
 26063,0
-53077,0
+53077,1
 21077,0
 21093,0
 18065,0
@@ -1153,7 +1153,7 @@ FIPS,Result
 27059,0
 32003,0
 21197,0
-49049,1
+49049,0
 1115,0
 40075,0
 51143,0
@@ -1161,18 +1161,18 @@ FIPS,Result
 16011,0
 6017,0
 37011,0
-29133,1
+29133,0
 55075,0
 36059,1
 30097,0
-26065,0
+26065,1
 40039,0
 18043,0
-18127,0
+18127,1
 48485,0
 23001,0
-36021,1
-32510,1
+36021,0
+32510,0
 51036,0
 38103,0
 18091,0
@@ -1180,23 +1180,23 @@ FIPS,Result
 13117,0
 47053,0
 47023,0
-18081,1
+18081,0
 20117,0
 33007,0
 40027,0
 50009,0
 48097,0
-48041,0
+48041,1
 36053,0
 16079,0
 36005,1
-47097,1
+47097,0
 37175,0
 21191,0
 51097,0
 5027,0
 21035,0
-17019,0
+17019,1
 37009,0
 48313,0
 20059,0
@@ -1207,8 +1207,8 @@ FIPS,Result
 41033,0
 28049,1
 19177,0
-28043,1
-13165,1
+28043,0
+13165,0
 53057,0
 51610,1
 27139,0
@@ -1222,27 +1222,27 @@ FIPS,Result
 1029,0
 22085,0
 30041,0
-1041,1
-46041,0
+1041,0
+46041,1
 41051,1
-47131,1
+47131,0
 51169,0
 18071,0
 51830,1
 34023,1
 8055,0
-28105,0
+28105,1
 48309,0
-6105,0
+6105,1
 21057,0
 18073,0
 28031,0
-4019,0
-56033,1
+4019,1
+56033,0
 40107,0
 48335,0
 17025,0
-1065,1
+1065,0
 51051,0
 31145,0
 28027,1
@@ -1292,10 +1292,10 @@ FIPS,Result
 44003,0
 48317,0
 38089,0
-48505,0
+48505,1
 13069,0
-28087,1
-29085,1
+28087,0
+29085,0
 12105,0
 54023,0
 51101,0
@@ -1311,7 +1311,7 @@ FIPS,Result
 1061,0
 41063,0
 40135,0
-13145,1
+13145,0
 29077,0
 30079,0
 16035,0
@@ -1319,38 +1319,38 @@ FIPS,Result
 48023,0
 40021,0
 48267,0
-54067,1
-5119,1
+54067,0
+5119,0
 40083,0
 48251,0
 21183,0
 27027,0
-23013,1
+23013,0
 13139,0
 28009,0
 51025,0
-13095,0
-41027,0
+13095,1
+41027,1
 39033,0
-35031,0
+35031,1
 38047,0
 36105,0
 51690,0
 30023,0
 29047,0
 12071,0
-48191,1
-21075,1
+48191,0
+21075,0
 19057,0
 18063,0
 40113,0
-22101,1
+22101,0
 38087,0
 13289,0
 37121,0
 45017,0
 36071,0
-46095,1
+46095,0
 48441,0
 47071,0
 13109,0
@@ -1364,20 +1364,20 @@ FIPS,Result
 45065,0
 17021,0
 47087,0
-36079,1
+36079,0
 8115,0
 10001,0
 45039,0
 22091,0
-24041,0
+24041,1
 21165,0
 25013,0
-36067,0
+36067,1
 31089,0
 38001,0
 46075,0
 29053,0
-27127,1
+27127,0
 6073,1
 8069,1
 51175,0
@@ -1385,7 +1385,7 @@ FIPS,Result
 13143,0
 18051,0
 54051,0
-6065,0
+6065,1
 17183,0
 35043,1
 21089,0
@@ -1398,14 +1398,14 @@ FIPS,Result
 40097,0
 51127,0
 1009,0
-37147,0
+37147,1
 37103,0
 47065,0
 55101,0
 29137,0
 31151,0
-27109,1
-36027,1
+27109,0
+36027,0
 6045,0
 54039,0
 8001,0
@@ -1431,11 +1431,11 @@ FIPS,Result
 29037,0
 16085,0
 38075,0
-21189,1
+21189,0
 29181,0
 19085,0
 37161,0
-46121,1
+46121,0
 37195,0
 48417,0
 37113,0
@@ -1448,7 +1448,7 @@ FIPS,Result
 55099,0
 36041,0
 31079,0
-44009,1
+44009,0
 16049,0
 39049,1
 20209,0
@@ -1459,9 +1459,9 @@ FIPS,Result
 5021,0
 35045,0
 21031,0
-21025,1
+21025,0
 48163,0
-46017,1
+46017,0
 48095,0
 55123,0
 12077,0
@@ -1477,7 +1477,7 @@ FIPS,Result
 21041,0
 48503,0
 27173,0
-55063,1
+55063,0
 29115,0
 27005,0
 29125,0
@@ -1494,7 +1494,7 @@ FIPS,Result
 38027,0
 5105,0
 51033,0
-54047,1
+54047,0
 28155,0
 28097,0
 13321,0
@@ -1508,7 +1508,7 @@ FIPS,Result
 39127,0
 5071,0
 47125,0
-8037,0
+8037,1
 38045,0
 17153,0
 29067,0
@@ -1525,9 +1525,9 @@ FIPS,Result
 17095,0
 18095,0
 29027,0
-51540,0
+51540,1
 37027,0
-54055,1
+54055,0
 51059,1
 48189,0
 37179,0
@@ -1536,7 +1536,7 @@ FIPS,Result
 21127,0
 47045,0
 51119,0
-1085,0
+1085,1
 13037,0
 47089,0
 40067,0
diff --git a/submission_creative.csv b/submission_creative.csv
index 20a59f6..e4517d3 100644
--- a/submission_creative.csv
+++ b/submission_creative.csv
@@ -43,17 +43,17 @@ FIPS,Result
 19119,0
 47083,0
 42095,0
-6081,1
+6081,0
 13313,0
-5123,1
+5123,0
 28039,0
 51760,1
 48245,0
-41053,0
+41053,1
 8025,0
 5079,0
 49043,0
-48489,0
+48489,1
 22057,0
 1021,0
 29185,0
@@ -61,13 +61,13 @@ FIPS,Result
 54081,0
 6003,0
 29225,0
-33015,0
+33015,1
 42071,0
 20065,0
 37097,0
 48283,0
 46083,0
-31055,0
+31055,1
 51580,0
 37157,0
 42033,0
@@ -76,7 +76,7 @@ FIPS,Result
 13075,0
 21123,0
 48185,0
-13245,1
+13245,0
 13001,0
 31139,0
 38025,0
@@ -85,12 +85,12 @@ FIPS,Result
 48101,0
 42037,0
 19079,0
-12049,0
+12049,1
 21131,0
 13183,0
 53035,1
 48479,1
-21063,1
+21063,0
 20101,0
 48415,0
 51185,0
@@ -99,7 +99,7 @@ FIPS,Result
 55119,0
 45075,1
 47101,0
-26089,1
+26089,0
 48367,0
 37131,1
 21205,0
@@ -109,7 +109,7 @@ FIPS,Result
 1093,0
 31101,0
 5069,0
-8119,0
+8119,1
 54037,0
 13213,0
 29089,0
@@ -134,13 +134,13 @@ FIPS,Result
 22059,0
 38017,0
 26109,0
-49037,0
+49037,1
 48265,0
 39129,0
 20055,0
 21121,0
 31173,1
-22035,1
+22035,0
 8011,0
 39027,0
 17107,0
@@ -159,7 +159,7 @@ FIPS,Result
 21021,0
 42123,0
 32015,0
-15009,0
+15009,1
 42057,0
 53065,0
 48211,0
@@ -191,7 +191,7 @@ FIPS,Result
 48083,0
 45031,0
 4023,1
-47095,0
+47095,1
 12081,0
 46119,0
 12045,0
@@ -222,12 +222,12 @@ FIPS,Result
 28081,0
 29093,0
 46003,0
-4027,1
-53025,1
+4027,0
+53025,0
 26121,0
 21203,0
 21211,0
-44005,1
+44005,0
 35005,0
 30017,0
 31107,0
@@ -307,7 +307,7 @@ FIPS,Result
 26131,0
 48177,0
 8027,0
-23007,0
+23007,1
 51145,0
 51087,1
 20157,0
@@ -335,8 +335,8 @@ FIPS,Result
 17143,1
 53067,1
 21105,0
-24009,0
-1025,1
+24009,1
+1025,0
 27067,0
 19075,0
 12037,0
@@ -351,15 +351,15 @@ FIPS,Result
 27091,0
 42099,0
 42117,0
-13077,1
-51520,1
+13077,0
+51520,0
 36051,0
 54025,0
 30049,0
 1079,0
-37117,0
+37117,1
 48133,0
-42041,1
+42041,0
 49007,0
 47143,0
 55039,0
@@ -376,11 +376,11 @@ FIPS,Result
 29029,0
 13081,0
 5111,0
-51133,0
+51133,1
 6057,1
 21225,0
-48141,1
-28073,1
+48141,0
+28073,0
 27135,0
 37129,0
 37025,0
@@ -407,19 +407,19 @@ FIPS,Result
 53001,0
 19019,0
 39043,0
-34001,0
+34001,1
 31113,0
 27145,0
 47105,0
 36117,0
 48043,1
-45063,1
+45063,0
 16057,1
 48157,0
 12073,1
 41067,1
 29187,0
-8049,0
+8049,1
 40133,0
 37183,1
 54005,0
@@ -432,7 +432,7 @@ FIPS,Result
 48341,0
 47077,0
 28135,1
-1013,0
+1013,1
 40145,0
 1111,0
 37111,0
@@ -446,9 +446,9 @@ FIPS,Result
 31087,0
 17091,0
 13293,0
-48461,0
+48461,1
 28107,1
-8041,0
+8041,1
 5049,0
 27073,0
 36055,1
@@ -479,7 +479,7 @@ FIPS,Result
 27033,0
 22047,0
 37171,0
-51003,1
+51003,0
 54041,0
 17155,0
 5075,0
@@ -493,7 +493,7 @@ FIPS,Result
 13227,0
 51011,0
 36113,0
-54061,0
+54061,1
 18135,0
 21221,0
 17011,0
@@ -538,7 +538,7 @@ FIPS,Result
 51071,0
 26113,0
 8015,0
-19061,0
+19061,1
 27049,0
 31011,0
 48357,0
@@ -558,8 +558,8 @@ FIPS,Result
 13123,0
 13283,0
 31073,0
-51103,0
-36119,1
+51103,1
+36119,0
 13229,0
 48281,0
 6039,0
@@ -577,7 +577,7 @@ FIPS,Result
 48031,0
 20175,0
 41059,0
-27087,1
+27087,0
 41061,0
 48307,0
 56005,0
@@ -640,23 +640,23 @@ FIPS,Result
 32027,0
 19181,0
 13267,0
-17007,1
+17007,0
 48137,0
-8035,1
+8035,0
 34025,1
 18007,0
-4013,0
+4013,1
 54073,0
 26045,0
 29011,0
-55135,0
+55135,1
 30071,0
 12093,0
 6107,0
-1011,0
+1011,1
 29035,0
 31069,0
-26037,1
+26037,0
 23003,0
 13063,1
 19109,0
@@ -668,20 +668,20 @@ FIPS,Result
 47051,0
 39105,0
 38091,0
-54089,1
+54089,0
 42051,0
 51081,1
-22055,1
+22055,0
 51029,0
 51093,0
 22025,0
 51750,1
 35011,0
-49005,1
+49005,0
 48197,0
 31125,0
 1107,0
-45085,1
+45085,0
 51061,0
 13091,0
 5131,0
@@ -698,7 +698,7 @@ FIPS,Result
 51157,0
 31081,0
 27125,0
-1035,1
+1035,0
 13163,0
 31057,0
 13257,0
@@ -716,14 +716,14 @@ FIPS,Result
 24047,0
 6085,1
 45079,1
-46011,0
+46011,1
 48145,0
 49039,0
 45051,0
 39113,0
-12086,1
+12086,0
 21223,0
-39057,1
+39057,0
 18157,1
 18133,0
 28021,1
@@ -740,7 +740,7 @@ FIPS,Result
 48231,0
 24043,0
 51017,0
-8021,1
+8021,0
 27053,1
 18145,0
 37031,0
@@ -762,8 +762,8 @@ FIPS,Result
 5037,0
 25021,1
 42133,0
-8077,0
-13021,1
+8077,1
+13021,0
 6075,1
 37093,1
 37055,0
@@ -776,7 +776,7 @@ FIPS,Result
 34029,0
 20099,0
 22079,0
-13239,1
+13239,0
 51079,0
 31105,0
 13249,0
@@ -788,9 +788,9 @@ FIPS,Result
 31109,0
 55137,0
 39063,0
-5017,1
+5017,0
 54035,0
-26061,1
+26061,0
 28101,0
 8087,0
 31043,0
@@ -815,12 +815,12 @@ FIPS,Result
 50017,1
 38081,0
 35017,0
-51117,0
+51117,1
 51161,0
 19157,0
 26157,0
 47135,0
-13153,1
+13153,0
 48431,0
 20075,0
 20069,0
@@ -832,7 +832,7 @@ FIPS,Result
 9007,0
 27061,0
 27081,0
-51153,1
+51153,0
 46071,0
 50015,1
 39061,1
@@ -848,7 +848,7 @@ FIPS,Result
 47133,0
 46097,0
 35013,1
-22071,1
+22071,0
 22015,0
 35019,0
 37145,0
@@ -866,11 +866,11 @@ FIPS,Result
 36017,0
 35053,0
 48067,0
-18057,0
+18057,1
 6101,0
 48465,1
 51740,1
-20037,1
+20037,0
 17093,1
 40055,0
 26117,0
@@ -887,7 +887,7 @@ FIPS,Result
 26051,0
 29171,0
 54013,0
-26073,1
+26073,0
 48291,0
 51660,1
 39007,0
@@ -909,7 +909,7 @@ FIPS,Result
 30077,0
 48419,0
 20073,0
-37067,0
+37067,1
 55013,0
 30013,0
 42065,0
@@ -930,13 +930,13 @@ FIPS,Result
 20129,0
 18023,0
 51031,0
-22097,1
+22097,0
 16051,0
 27147,0
 48039,0
 20139,0
 12095,1
-51650,0
+51650,1
 1087,1
 54069,0
 39133,0
@@ -977,11 +977,11 @@ FIPS,Result
 48323,1
 46085,0
 54093,0
-36019,1
+36019,0
 19165,0
 45003,0
 51810,0
-26053,0
+26053,1
 22045,0
 38003,0
 12099,0
@@ -989,8 +989,8 @@ FIPS,Result
 1091,1
 51023,0
 13085,0
-36007,0
-45041,1
+36007,1
+45041,0
 22023,0
 51053,0
 40131,0
@@ -1014,7 +1014,7 @@ FIPS,Result
 20137,0
 13189,0
 6055,1
-22065,1
+22065,0
 55033,0
 1081,1
 31001,0
@@ -1025,12 +1025,12 @@ FIPS,Result
 17039,0
 1083,0
 21213,0
-13219,1
+13219,0
 17157,0
 48153,0
 54001,0
 54083,0
-9015,0
+9015,1
 4017,1
 29145,0
 5099,0
@@ -1041,7 +1041,7 @@ FIPS,Result
 13187,0
 40149,0
 20171,0
-5035,1
+5035,0
 13231,0
 40009,0
 38069,0
@@ -1051,9 +1051,9 @@ FIPS,Result
 37091,0
 27045,0
 8013,1
-13197,0
+13197,1
 32005,0
-17161,1
+17161,0
 12023,0
 40059,0
 42127,0
@@ -1067,7 +1067,7 @@ FIPS,Result
 13047,0
 23011,0
 17193,0
-22005,1
+22005,0
 48249,1
 34031,1
 18183,0
@@ -1075,7 +1075,7 @@ FIPS,Result
 19007,0
 13295,0
 34019,1
-55079,0
+55079,1
 17027,0
 41029,0
 25027,1
@@ -1089,7 +1089,7 @@ FIPS,Result
 27103,0
 31033,0
 47175,0
-1047,1
+1047,0
 40095,0
 55077,0
 19169,1
@@ -1097,8 +1097,8 @@ FIPS,Result
 55125,0
 26145,0
 4009,1
-47037,0
-36077,0
+47037,1
+36077,1
 54105,0
 27017,0
 12053,0
@@ -1122,7 +1122,7 @@ FIPS,Result
 22027,0
 48047,1
 46043,0
-18003,0
+18003,1
 17013,0
 51700,1
 29097,0
@@ -1135,7 +1135,7 @@ FIPS,Result
 13235,0
 48025,0
 18013,0
-35006,1
+35006,0
 36037,0
 27129,0
 23017,0
@@ -1151,24 +1151,24 @@ FIPS,Result
 27023,0
 29139,0
 27059,0
-32003,0
+32003,1
 21197,0
 49049,0
 1115,0
 40075,0
 51143,0
-16039,1
+16039,0
 16011,0
-6017,0
+6017,1
 37011,0
 29133,0
 55075,0
-36059,1
+36059,0
 30097,0
 26065,1
 40039,0
-18043,0
-18127,0
+18043,1
+18127,1
 48485,0
 23001,0
 36021,0
@@ -1209,12 +1209,12 @@ FIPS,Result
 19177,0
 28043,0
 13165,0
-53057,1
+53057,0
 51610,1
 27139,0
 16009,0
 22093,0
-38005,0
+38005,1
 48147,0
 48423,0
 54099,0
@@ -1232,7 +1232,7 @@ FIPS,Result
 34023,1
 8055,1
 28105,1
-48309,0
+48309,1
 6105,1
 21057,0
 18073,0
@@ -1287,9 +1287,9 @@ FIPS,Result
 49035,0
 20063,0
 18079,0
-13125,0
+13125,1
 17191,0
-44003,1
+44003,0
 48317,0
 38089,0
 48505,1
@@ -1328,7 +1328,7 @@ FIPS,Result
 23013,0
 13139,0
 28009,0
-51025,0
+51025,1
 13095,1
 41027,1
 39033,0
@@ -1367,11 +1367,11 @@ FIPS,Result
 36079,0
 8115,0
 10001,0
-45039,0
+45039,1
 22091,0
 24041,0
 21165,0
-25013,0
+25013,1
 36067,1
 31089,0
 38001,0
@@ -1387,7 +1387,7 @@ FIPS,Result
 54051,0
 6065,1
 17183,0
-35043,1
+35043,0
 21089,0
 16077,0
 22043,0
@@ -1406,7 +1406,7 @@ FIPS,Result
 31151,0
 27109,0
 36027,0
-6045,0
+6045,1
 54039,0
 8001,0
 21027,0
@@ -1415,7 +1415,7 @@ FIPS,Result
 29123,0
 46079,0
 31121,0
-5107,1
+5107,0
 34015,1
 18121,0
 1099,0
@@ -1436,11 +1436,11 @@ FIPS,Result
 19085,0
 37161,0
 46121,1
-37195,0
+37195,1
 48417,0
 37113,0
 48007,0
-51179,1
+51179,0
 12015,0
 4012,0
 29063,0
@@ -1448,20 +1448,20 @@ FIPS,Result
 55099,0
 36041,0
 31079,0
-44009,1
+44009,0
 16049,0
 39049,1
-20209,1
-47069,0
+20209,0
+47069,1
 29205,0
-5007,1
+5007,0
 6029,1
 5021,0
 35045,1
 21031,0
 21025,0
-48163,0
-46017,0
+48163,1
+46017,1
 48095,0
 55123,0
 12077,0
@@ -1477,7 +1477,7 @@ FIPS,Result
 21041,0
 48503,0
 27173,0
-55063,0
+55063,1
 29115,0
 27005,0
 29125,0
@@ -1531,7 +1531,7 @@ FIPS,Result
 51059,1
 48189,0
 37179,0
-53055,1
+53055,0
 48063,0
 21127,0
 47045,0
@@ -1541,7 +1541,7 @@ FIPS,Result
 47089,0
 40067,0
 8101,0
-17119,1
+17119,0
 26059,0
 12047,0
 18021,0