diff --git a/nncw.ipynb b/nncw.ipynb
index 8f034ee..a4ed2f5 100644
--- a/nncw.ipynb
+++ b/nncw.ipynb
@@ -59,7 +59,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 2,
    "metadata": {
     "colab": {
      "base_uri": "https://localhost:8080/",
@@ -245,7 +245,7 @@
        "max          1.000000         1.000000    1.000000  "
       ]
      },
-     "execution_count": 4,
+     "execution_count": 2,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -417,7 +417,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 3,
    "metadata": {
     "executionInfo": {
      "elapsed": 2604,
@@ -450,7 +450,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 4,
    "metadata": {
     "executionInfo": {
      "elapsed": 2598,
@@ -1598,7 +1598,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 89,
+   "execution_count": 6,
    "metadata": {
     "tags": [
      "exp2",
@@ -1607,8 +1607,7 @@
    },
    "outputs": [],
    "source": [
-    "# num_models=[1, 3, 9, 15, 25]\n",
-    "num_models=[1, 3, 9]\n",
+    "num_models=[1, 3, 9, 15, 25]\n",
     "\n",
     "def evaluate_ensemble_vote(hidden_nodes=16, \n",
     "                           epochs=50, \n",
@@ -1678,17 +1677,19 @@
     "        ########################\n",
     "        ##  FEEDFORWARD TEST\n",
     "        ########################\n",
+    "        # TEST DATA PREDICTIONS\n",
     "        response[\"predictions\"] = [model(dtest.to_numpy()) for model in models]\n",
+    "        # TEST LABEL TENSOR\n",
+    "        ltest_tensor = tf.constant(ltest.to_numpy())\n",
     "\n",
     "        ########################\n",
     "        ##  ENSEMBLE ACCURACY\n",
     "        ########################\n",
     "        ensem_sum_rounded = sum(tf.math.round(pred) for pred in response[\"predictions\"])\n",
-    "        ensem_sum = sum(pred for pred in response[\"predictions\"])\n",
+    "        ensem_sum = sum(response[\"predictions\"])\n",
     "        # round predictions to onehot vectors and sum over all ensemble models\n",
     "        # take argmax for ensemble predicted class\n",
     "        \n",
-    "        ltest_tensor = tf.constant(ltest.to_numpy()) # transform test labels into tensor\n",
     "        correct = 0 # number of correct ensemble predictions\n",
     "        correct_num_models = 0 # when correctly predicted ensembley, proportion of models correctly classifying\n",
     "        individual_accuracy = 0 # proportion of models correctly classifying\n",
@@ -1702,11 +1703,12 @@
     "            else:\n",
     "                pred_val = pc\n",
     "            \n",
-    "            individual_accuracy += pcr[gt_argmax] / m\n",
+    "            correct_models = pcr[gt_argmax] / m # use rounded value so will divide nicely\n",
+    "            individual_accuracy += correct_models\n",
     "            \n",
     "            if tf.math.argmax(pred_val) == gt_argmax: # ENSEMBLE EVALUATE HERE\n",
     "                correct += 1\n",
-    "                correct_num_models += pcr[gt_argmax] / m\n",
+    "                correct_num_models += correct_models\n",
     "                \n",
     "#                 print(pc.numpy(), pcr.numpy(), gt.numpy(), (pcr[gt_argmax] / m).numpy(), True) # debug\n",
     "#             else:\n",
@@ -1838,7 +1840,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 116,
+   "execution_count": 40,
    "metadata": {
     "tags": [
      "exp2"
@@ -1849,22 +1851,24 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Iteration 1/3\n",
-      "Iteration 2/3\n",
-      "Iteration 3/3\n"
+      "Iteration 1/5\n",
+      "Iteration 2/5\n",
+      "Iteration 3/5\n",
+      "Iteration 4/5\n",
+      "Iteration 5/5\n"
      ]
     }
    ],
    "source": [
     "multi_ensem_results = list()\n",
-    "multi_ensem_iterations = 3\n",
+    "multi_ensem_iterations = 5\n",
     "for i in range(multi_ensem_iterations):\n",
     "    print(f\"Iteration {i+1}/{multi_ensem_iterations}\")\n",
     "    data_train, data_test, labels_train, labels_test = train_test_split(data, labels, test_size=0.5, stratify=labels)\n",
     "    multi_ensem_results.append(list(evaluate_ensemble_vote(epochs=(1, 100),\n",
     "                                                           hidden_nodes=16,\n",
-    "                                                           nmodels=[1, 3, 5, 7, 9],\n",
-    "                                                           optimizer=lambda: tf.keras.optimizers.SGD(learning_rate=0.05, momentum=0.02),\n",
+    "                                                           nmodels=[1, 3, 7, 11, 15],\n",
+    "                                                           optimizer=lambda: tf.keras.optimizers.SGD(learning_rate=0.05, momentum=0.01),\n",
     "                                                           weight_init=lambda: 'random_uniform',\n",
     "                                                           batch_size=35,\n",
     "                                                           dtrain=data_train, \n",
@@ -1898,7 +1902,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 117,
+   "execution_count": 41,
    "metadata": {
     "tags": [
      "exp2"
@@ -1909,12 +1913,12 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "3 Tests\n",
-      "Models: [1, 3, 5, 7, 9]\n",
+      "5 Tests\n",
+      "Models: [1, 3, 7, 11, 15]\n",
       "\n",
       "Loss: categorical_crossentropy\n",
       "LR: 0.05\n",
-      "Momentum: 0.02\n"
+      "Momentum: 0.01\n"
      ]
     }
    ],
@@ -2005,7 +2009,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 118,
+   "execution_count": 42,
    "metadata": {
     "tags": [
      "exp2"
@@ -2016,7 +2020,7 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Models: 1, 8e+01% Accurate\n"
+      "Models: 11, 74.3% Accurate\n"
      ]
     }
    ],
@@ -2025,7 +2029,7 @@
     "best_ensem_accuracy = mean_ensem_accuracy[best_ensem_accuracy_idx]\n",
     "best_ensem_accuracy_models = multi_ensem_models[best_ensem_accuracy_idx[1]]\n",
     "\n",
-    "print(f'Models: {best_ensem_accuracy_models}, {best_ensem_accuracy * 100:.1}% Accurate')"
+    "print(f'Models: {best_ensem_accuracy_models}, {best_ensem_accuracy * 100:.3}% Accurate')"
    ]
   },
   {
@@ -2036,12 +2040,12 @@
     ]
    },
    "source": [
-    "### Test/Train Accuracy Over Model Numbers"
+    "### Test/Train Error Over Model Numbers"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 139,
+   "execution_count": 43,
    "metadata": {
     "tags": [
      "exp2"
@@ -2050,7 +2054,7 @@
    "outputs": [
     {
      "data": {
-      "image/png": "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\n",
+      "image/png": "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\n",
       "text/plain": [
        "<Figure size 560x350 with 1 Axes>"
       ]
@@ -2065,10 +2069,15 @@
     "fig = plt.figure(figsize=(8, 5))\n",
     "fig.set_dpi(fig_dpi)\n",
     "\n",
-    "plt.errorbar(multi_ensem_models, 1 - mean_ensem_accuracy[0, :], yerr=std_ensem_accuracy[0, :], capsize=2, label='Ensemble Test')\n",
-    "plt.errorbar(multi_ensem_models, 1 - mean_ensem_accuracy[2, :], yerr=std_ensem_accuracy[2, :], capsize=2, label='Individual Test')\n",
-    "plt.errorbar(multi_ensem_models, 1 - mean_ensem_accuracy[1, :], yerr=std_ensem_accuracy[1, :], capsize=2, label='Individual Train')\n",
-    "plt.errorbar(multi_ensem_models, 1 - mean_ensem_accuracy[3, :], yerr=std_ensem_accuracy[3, :], capsize=2, label='Anti-agreement')\n",
+    "# plt.plot(multi_ensem_models, 1 - mean_ensem_accuracy[0, :], 'x-', label='Ensemble Test')\n",
+    "# plt.plot(multi_ensem_models, 1 - mean_ensem_accuracy[2, :], 'x-', label='Individual Test')\n",
+    "# plt.plot(multi_ensem_models, 1 - mean_ensem_accuracy[1, :], 'x-', label='Individual Train')\n",
+    "# plt.plot(multi_ensem_models, 1 - mean_ensem_accuracy[3, :], 'x-', label='Agreement')\n",
+    "\n",
+    "plt.errorbar(multi_ensem_models, 1 - mean_ensem_accuracy[0, :], yerr=std_ensem_accuracy[0, :], capsize=4, label='Ensemble Test')\n",
+    "plt.errorbar(multi_ensem_models, 1 - mean_ensem_accuracy[2, :], yerr=std_ensem_accuracy[2, :], capsize=4, label='Individual Test')\n",
+    "plt.errorbar(multi_ensem_models, 1 - mean_ensem_accuracy[1, :], yerr=std_ensem_accuracy[1, :], capsize=4, label='Individual Train')\n",
+    "plt.errorbar(multi_ensem_models, 1 - mean_ensem_accuracy[3, :], yerr=std_ensem_accuracy[3, :], capsize=4, label='Disagreement')\n",
     "\n",
     "plt.title(f\"Error Rate for Horizontal Ensemble Models\")\n",
     "plt.ylim(0, 1)\n",
@@ -2079,62 +2088,6 @@
     "plt.show()"
    ]
   },
-  {
-   "cell_type": "markdown",
-   "metadata": {
-    "tags": [
-     "exp2"
-    ]
-   },
-   "source": [
-    "### Ensemble Model Statistics"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 140,
-   "metadata": {
-    "tags": [
-     "exp2"
-    ]
-   },
-   "outputs": [
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 560x350 with 1 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "fig = plt.figure(figsize=(8, 5))\n",
-    "fig.set_dpi(fig_dpi)\n",
-    "\n",
-    "# plt.plot(multi_ensem_models, mean_ensem_accuracy[0, :], 'x-', label='Ensemble')\n",
-    "# plt.plot(multi_ensem_models, mean_ensem_accuracy[2, :], 'x-', label='Individual Test')\n",
-    "# plt.plot(multi_ensem_models, mean_ensem_accuracy[1, :], 'x-', label='Individual Train')\n",
-    "# plt.plot(multi_ensem_models, mean_ensem_accuracy[3, :], 'x-', label='Agreement')\n",
-    "\n",
-    "plt.errorbar(multi_ensem_models, mean_ensem_accuracy[0, :], yerr=std_ensem_accuracy[0, :], capsize=2, label='Ensemble Test')\n",
-    "plt.errorbar(multi_ensem_models, mean_ensem_accuracy[2, :], yerr=std_ensem_accuracy[2, :], capsize=2, label='Individual Test')\n",
-    "plt.errorbar(multi_ensem_models, mean_ensem_accuracy[1, :], yerr=std_ensem_accuracy[1, :], capsize=2, label='Individual Train')\n",
-    "plt.errorbar(multi_ensem_models, mean_ensem_accuracy[3, :], yerr=std_ensem_accuracy[3, :], capsize=2, label='Agreement')\n",
-    "\n",
-    "plt.title(\"Accuracy for Horizontal Model Ensembles\")\n",
-    "plt.ylim(0, 1)\n",
-    "plt.grid()\n",
-    "plt.legend()\n",
-    "plt.ylabel(\"Accuracy\")\n",
-    "plt.xlabel(\"Number of Models\")\n",
-    "plt.show()"
-   ]
-  },
   {
    "cell_type": "markdown",
    "metadata": {
@@ -2177,9 +2130,9 @@
    "version": "3.8.8"
   },
   "toc-autonumbering": false,
-  "toc-showcode": true,
+  "toc-showcode": false,
   "toc-showmarkdowntxt": false,
-  "toc-showtags": true
+  "toc-showtags": false
  },
  "nbformat": 4,
  "nbformat_minor": 4