Prohack_Solution_to_2nd_Task_(Optimization).ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Prohack: Solution to 2nd Task (Optimization).ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eOnyIES5Ko2H",
        "colab_type": "text"
      },
      "source": [
        "# Prohack: 2nd Task"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bBgZqL9_KtSG",
        "colab_type": "text"
      },
      "source": [
        "We had to allocate energy for Galaxies knowing that : \n",
        "* Every galaxy has a certain limited potential for improvement in the index described by the following function: Potential for increase in the Index = -np.log(Index+0.01)+3 \n",
        "* Likely index increase dependent on potential for improvement and on extra energy availability is described by the following function: Likely increase in the Index = extra energy * Potential for increase in the Index **2 / 1000\n",
        "* **There are several constraints:**\n",
        "  * In total there are 50000 zillion DSML available for allocation.\n",
        "  * No galaxy should be allocated more than 100 zillion DSML or less than 0 zillion DSML .\n",
        "  * galaxies with low existence expectancy index below 0.7 should be allocated at least 10% of the total energy available.\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "v0Lb59iAIkTT",
        "colab_type": "text"
      },
      "source": [
        "#Importing the preds"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xKrHL_oM6sPW",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sM6aRI866ykq",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "ss=pd.read_csv('new_preds.csv')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "R0AMd3QE7Gxw",
        "colab_type": "text"
      },
      "source": [
        "**Defining the Potential Increase and Likely Increase in the index** "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QwrwH6Ys7V9W",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#The index represent the y_pred \n",
        "index = ss['pred']\n",
        "pot_inc = -np.log(index+0.01)+3\n",
        "p2= pot_inc**2\n",
        "ss[\"p2\"] = p2"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "asblBcxr8TCf",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Sorting using Likelyincreasing index\n",
        "ss=ss.sort_values('p2',ascending=False)\n",
        "#Droping The old index \n",
        "ss=ss.reset_index(drop=True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eO0P2moWBFAN",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 402
        },
        "outputId": "791a9b3b-7604-4927-a7ab-9ba4293de3f7"
      },
      "source": [
        "ss"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Index</th>\n",
              "      <th>pred</th>\n",
              "      <th>opt_pred</th>\n",
              "      <th>eei</th>\n",
              "      <th>p2</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>49</td>\n",
              "      <td>0.017404</td>\n",
              "      <td>0</td>\n",
              "      <td>0.671839</td>\n",
              "      <td>43.521325</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>228</td>\n",
              "      <td>0.018274</td>\n",
              "      <td>0</td>\n",
              "      <td>0.837479</td>\n",
              "      <td>43.109723</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>138</td>\n",
              "      <td>0.018336</td>\n",
              "      <td>0</td>\n",
              "      <td>0.775467</td>\n",
              "      <td>43.081108</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>186</td>\n",
              "      <td>0.018623</td>\n",
              "      <td>0</td>\n",
              "      <td>0.806444</td>\n",
              "      <td>42.948887</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>201</td>\n",
              "      <td>0.018673</td>\n",
              "      <td>0</td>\n",
              "      <td>0.816966</td>\n",
              "      <td>42.925849</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>885</th>\n",
              "      <td>800</td>\n",
              "      <td>0.443243</td>\n",
              "      <td>0</td>\n",
              "      <td>1.085557</td>\n",
              "      <td>14.374152</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>886</th>\n",
              "      <td>801</td>\n",
              "      <td>0.443628</td>\n",
              "      <td>0</td>\n",
              "      <td>1.086098</td>\n",
              "      <td>14.367718</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>887</th>\n",
              "      <td>705</td>\n",
              "      <td>0.446518</td>\n",
              "      <td>0</td>\n",
              "      <td>1.038657</td>\n",
              "      <td>14.319613</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>888</th>\n",
              "      <td>583</td>\n",
              "      <td>0.470094</td>\n",
              "      <td>0</td>\n",
              "      <td>0.996074</td>\n",
              "      <td>13.941070</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>889</th>\n",
              "      <td>730</td>\n",
              "      <td>0.488002</td>\n",
              "      <td>0</td>\n",
              "      <td>1.051112</td>\n",
              "      <td>13.668922</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>890 rows × 5 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "     Index      pred  opt_pred       eei         p2\n",
              "0       49  0.017404         0  0.671839  43.521325\n",
              "1      228  0.018274         0  0.837479  43.109723\n",
              "2      138  0.018336         0  0.775467  43.081108\n",
              "3      186  0.018623         0  0.806444  42.948887\n",
              "4      201  0.018673         0  0.816966  42.925849\n",
              "..     ...       ...       ...       ...        ...\n",
              "885    800  0.443243         0  1.085557  14.374152\n",
              "886    801  0.443628         0  1.086098  14.367718\n",
              "887    705  0.446518         0  1.038657  14.319613\n",
              "888    583  0.470094         0  0.996074  13.941070\n",
              "889    730  0.488002         0  1.051112  13.668922\n",
              "\n",
              "[890 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 75
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uKZiCv3zIs8l",
        "colab_type": "text"
      },
      "source": [
        "# Optimisation"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4abcAu2Q75Lt",
        "colab_type": "text"
      },
      "source": [
        "We'll start talking about our optimisation, which might not be the best approach since we believe that it overfitted the public LB: \n",
        "* We've noticed that giving max energy (100 Zillion DSML) for the first 400 rows would give a good score because there's a huge drop in the values of  y_preds, as we've seen in the 1st notebook, *y is highly skewed*\n",
        "* Then we started decreasing linearly (and the overfitting started here )\n",
        "* We took random decreasing steps and random intervals and we were trying to maximize the sum of galaxies with low existence expectancy index below 0.7 because we knew that it had to be more than 5000 Zillion DSML.\n",
        "\n",
        "These points have been taken empicially and through intuition mostly by analyzing the y. Taking different points would result in choosing different slopes for the linear decrease.\n",
        "\n",
        "**How to find these slopes ?**\n",
        "\n",
        "You need to pick some points and compute a given slope between each two points while checking the sum of allocated energy in order to make it equal or almost equal to 50000 units.\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PLcSAwb9K3ty",
        "colab_type": "text"
      },
      "source": [
        "**Linear Programming**\n",
        "* We've tried solving this optimisation task using scipy.optimize using constraints and objective function but we got a better score while using the idea of decreasing linearly, so we didn't focus on it ."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jHNtp0AdMCea",
        "colab_type": "text"
      },
      "source": [
        "**Analyzing the y_preds**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jQKIIzDs99LY",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 616
        },
        "outputId": "079aa19c-bfe7-4751-b640-838b624d4f0f"
      },
      "source": [
        "y_pred=ss.pred\n",
        "x = np.arange(y_pred.min(),y_pred.max(),0.01)\n",
        "plt.figure(figsize=(25,10))\n",
        "plt.hist(y_pred,bins=x)\n",
        "plt.title('Y_pred Histogram',fontsize=20)\n",
        "plt.xlabel('y_pred',fontsize=15)\n",
        "plt.ylabel('Count',fontsize=15)\n",
        "plt.xticks(x)\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1800x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WBBbkid5-uXi",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 202
        },
        "outputId": "760c9e0c-f5f7-4c72-db36-3a0648ce9580"
      },
      "source": [
        "\n",
        "n = 340\n",
        "#Giving the max of Energy to the 340 first element (ordered using the likely Increasing Index)\n",
        "ss.opt_pred[:n]=100\n",
        "ss.opt_pred[n:] = 0\n",
        "c=100\n",
        "alpha = 0.62685\n",
        "for i in range(n,374):\n",
        "  if c>=alpha: \n",
        "    c=c-alpha\n",
        "    ss.loc[i,'opt_pred'] =c\n",
        "alpha=0.067345\n",
        "for i in range(374,455):\n",
        "  if c>=alpha: \n",
        "    c=c-alpha\n",
        "    ss.loc[i,'opt_pred'] =c\n",
        "  else:\n",
        "    ss.loc[i,'opt_pred'] = 0\n",
        "alpha = 0.03\n",
        "for i in range(455,465):\n",
        "  if c>=alpha: \n",
        "    c=c-alpha\n",
        "    ss.loc[i,'opt_pred'] =c\n",
        "alpha=0.4339465\n",
        "for i in range(465,890):\n",
        "  if c>=alpha: \n",
        "    c=c-alpha\n",
        "    ss.loc[i,'opt_pred'] =c\n",
        "  else:\n",
        "    ss.loc[i,'opt_pred'] = 0\n",
        "\n",
        "print(ss.opt_pred.sum())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "49999.995680999906\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  after removing the cwd from sys.path.\n",
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:5: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  \"\"\"\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Q17VLWKp9vtp",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 67
        },
        "outputId": "2a579fd9-d131-4718-f8f1-6727169a87a3"
      },
      "source": [
        "#Checking if the sum of opt_pred in rows having eei<0.7 is >5000\n",
        "print(\"sum\",ss.opt_pred.sum())\n",
        "print(\"left\", (50000-ss.opt_pred.sum())) \n",
        "print(\"eei Sum\",ss[ss.eei<0.7]['opt_pred'].sum())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "sum 49999.995680999906\n",
            "left 0.004319000094255898\n",
            "eei Sum 6520.424461499999\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QHS0cOYN96u8",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "outputId": "31e529e1-354d-4f4e-a4f0-056a97996681"
      },
      "source": [
        "# Plot opt pred vs y\n",
        "plt.plot(ss.pred)\n",
        "plt.plot(ss.opt_pred/100,color=\"red\")\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAD4CAYAAAD8Zh1EAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nO3deXxc9Xnv8c+jmdFmybJsecObMDYYG5vFAmwghCXIQFq7DYtttjihjFtKbnvrlMKFcltCk5A2obTh5uKEJoGwBAgNvikJCeCyJDZYZvWKhcEb3iVZtrbZfvePM7LlDcnyjM4s3/frpdfMnDma88x4+HL0e845P3POISIi2a/A7wJERCQ1FOgiIjlCgS4ikiMU6CIiOUKBLiKSI4J+bbiqqspVV1f7tXkRkay0fPnyXc65wUd6zrdAr66upq6uzq/Ni4hkJTPbcLTnNOQiIpIjFOgiIjlCgS4ikiMU6CIiOUKBLiKSI7oNdDP7DzPbYWYrjvK8mdm/mVm9mb1vZmelvkwREelOT/bQfwJc/hnPXwGMT/6EgR8cf1kiInKsuj0O3Tn3mplVf8Yqs4BHnXcd3qVmNsDMhjvntqaoxoO98Qb89rdpeWnxydChcOutYOZ3JSJZLRUnFo0ANnV5vDm57LBAN7Mw3l48o0eP7t3WliyB++7r3e9K5um8Hv+550JNjb+1iGS5Pm2KOucWOudqnHM1gwcf8czV7v3t30IioZ9c+dm+3ft3ffHF1H3RRPJUKgJ9CzCqy+ORyWUi3RsyBM46S4EukgKpCPRFwE3Jo12mAXvSNn4uuWnGDG8orbnZ70pEslpPDlt8ElgCnGJmm83sZjP7czP78+QqLwDrgXrgh8CtaatWclNtLcRisHix35WIZLWeHOUyt5vnHfCXKatI8s9550FZmTfsMmuW39WIZC2dKSr+KyyEiy/WOLrIcVKgS2aorYX16+Gjj/yuRCRrKdAlM8yY4d1qL12k1xTokhnGjYMTT1SgixwHBbpkBjNvL/2VVyAa9bsakaykQJfMUVsL+/Z5x6SLyDFToEvmuOQSCAQ07CLSSwp0yRwVFTB9ugJdpJcU6JJZZsyAt9+GnTv9rkQk6yjQJbPU1nqX1H3pJb8rEck6CnTJLFOnwsCBGnYR6QUFumSWQAAuu8yblapz8gsR6REFumSe2lrYuhVWHHFechE5CgW6ZJ7aWu9Wwy4ix0SBLpln5EiYNEmBLnKMFOiSmWbMgNdfh9ZWvysRyRoKdMlMtbXQ0QGvveZ3JSJZQ4EumenCC6G4WMMuIsdAgS6ZqaTEC3UFukiPKdAlc82YAatXw6ZNflcikhUU6JK5Og9f/O1v/a1DJEso0CVzTZoEI0Zo2EWkhxTokrnMvL30l16CeNzvakQyngJdMlttLTQ2Ql2d35WIZDwFumS2yy7z9tQ17CLSLQW6ZLZBg6CmRoEu0gMKdMl8M2bAm29CU5PflYhkNAW6ZL7aWq8p+sorflciktEU6JL5pk2D8nINu4h0Q4EumS8Ugksv9QJdsxiJHJUCXbLDjBmwYQOsW+d3JSIZq0eBbmaXm9laM6s3szuO8PxoM1tsZu+Y2ftmdmXqS5W8plmMRLrVbaCbWQB4CLgCmAjMNbOJh6x2N/C0c+5MYA7wf1JdqOS5sWNh3DgFushn6Mke+jlAvXNuvXMuAjwFzDpkHQf0T96vAD5NXYkiSTNmwOLF3sQXInKYngT6CKDr9Us3J5d19Q/ADWa2GXgB+NqRXsjMwmZWZ2Z1O3fu7EW5ktdqa70p6f7wB78rEclIqWqKzgV+4pwbCVwJPGZmh722c26hc67GOVczePDgFG1a8sbFF0MwCM89p6NdRI6gJ4G+BRjV5fHI5LKubgaeBnDOLQGKgapUFCiyX3k5XHklfP/7MGEC3H8/bN3qd1UiGaMngb4MGG9mJ5pZIV7Tc9Eh62wELgUws1PxAl1jKpJ6Tz4JP/4xDBkCd9wBo0bBzJnw/PMQjfpdnYivug1051wMuA14EViNdzTLSjO718xmJldbANxiZu8BTwLznNPfxJIGpaUwbx68/jqsXQtf/zosWwZ/8ideuP/d33nLRfKQ+ZW7NTU1rk7XuJZUiEbh17+GRx6B//ov77ov558PN98M11wDZWV+VyiSMma23DlXc6TndKaoZL9Q6MCwy6ZN3tj6zp3w1a/C8OFwyy2wdKkaqZLzFOiSW4YPh9tvhzVrvGGZq6+GJ56A6dPhtNPge9/zwl4kBynQJTeZwQUXeA3UrVth4ULvKJkFC7yJp6++Gl54QXOVSk5RoEvu69//wLDLihXwta/Bq6/CF78IY8bA3XfD+vV+Vyly3BTokl8mTYLvfhe2bIFnn4UpU+Bb34KTToJLLoHHH4e2Nr+rFOkVBbrkp8JCuOoqb9hlwwa47z7v9oYbvHH4W2+F5cvVSJWsosMWRTolEt5QzCOPwC9+Ae3tcPrp3qGPU6bA5MkwejQUaD9I/PNZhy0q0EWOpKnJOyv1kUe8PfVOF1wAr73mNV1FfKDj0EWO1YAB8Bd/AXV1sGePd4XHBQvgjTe8S/iKZCAFukh3+vf3jmP/xjegstI7BFIkAynQRXqqpARuusm7fK9OTpIMpEAXORbhsHftmJ/+1O9KRA6jQBc5FhMneo3RhQt1SKNkHAW6yLEKh2HdOu8QR5EMokAXOVZXX+0dBfPww35XInIQBbrIsVJzVDKUAl2kN265BSIRePRRvysR2U+BLtIbp50G552n5qhkFAW6SG/Nnw8ffuhdCkAkAyjQRXrrmmvUHJWMokAX6a2SErjxRu/KjLt2+V2NiAJd5LioOSoZRIEucjwmT/Yu3KXmqGQABbrI8QqHYe1aeP11vyuRPKdAFzle114LFRW6rK74ToEucrxKS73m6LPPwu7dflcjeUyBLpIK4TB0dKg5Kr5SoIukwuTJMG2amqPiKwW6SKqEw7BmjTfvqIgPFOgiqTJ7tpqj4qseBbqZXW5ma82s3szuOMo615rZKjNbaWZPpLZMkSxQWgo33ADPPKPmqPii20A3swDwEHAFMBGYa2YTD1lnPHAncL5zbhLw12moVSTz3XKL1xx97DG/K5E81JM99HOAeufceudcBHgKmHXIOrcADznnGgGccztSW6ZIljj9dDj3XDVHxRc9CfQRwKYujzcnl3V1MnCymf3ezJaa2eVHeiEzC5tZnZnV7dRML5KrwmFYvRp+/3u/K5E8k6qmaBAYD1wEzAV+aGYDDl3JObfQOVfjnKsZPHhwijYtkmFmz4b+/dUclT7Xk0DfAozq8nhkcllXm4FFzrmoc+5j4EO8gBfJP/36ec3Rp5+Ghga/q5E80pNAXwaMN7MTzawQmAMsOmSdX+LtnWNmVXhDMOtTWKdIduk8c1TNUelD3Qa6cy4G3Aa8CKwGnnbOrTSze81sZnK1F4HdZrYKWAz8rXNOx21J/jr9dDjnHDVHpU+Z8+nLVlNT4+rq6nzZtkifeOQR+LM/884cPf98v6uRHGFmy51zNUd6TmeKiqTL7NlQXq7mqPQZBbpIupSVHWiONjb6XY3kAQW6SDqFw9DeDj/7md+VSB5QoIuk0xlnwNlnw8MPqzkqaadAF0m3cBhWroQlS/yuRHKcAl0k3ebM8cbT1RyVNFOgi6RbWRlcfz38/OdqjkpaKdBF+sL8+V5z9PHH/a5EcpgCXaQvnHkm1NSoOSpppUAX6SvhMKxYAUuX+l2J5CgFukhfUXNU0kyBLtJXysvhuuu85mhTk9/VSA5SoIv0pXAY2trUHJW0UKCL9KWpU70fNUclDRToIn0tHIYPPoA33/S7EskxCnSRvjZ3rjdNnZqjeScWT/CbFdvY0tSWltdXoIv0tc7m6FNPwZ49flcjfWh3S4Q//9ly/nvtjrS8vgJdxA9qjualXfs6ABjUrzAtr69AF/FDTQ2cdZaao3mmoSUCwKCyorS8vgJdxC/hMLz/Prz1lt+VSB/pDPSB2kMXyTFqjuadXfuSe+gKdJEc07+/F+pqjuaNhpYOggVG/+JQWl5fgS7ip3AYWlvhiSf8rkT6QENLhMp+hRQUWFpeX4Eu4qeaGm/eUTVH88KufZG0DbeAAl3EX2be5BfvvQd1dX5XI2m0ubGVdzY2UZWmI1xAgS7iv+uug9JSby9dctaCp99j174Obpo+Jm3bUKCL+K2zOfrkk9Dc7Hc1kiYbG1r50pkjqJ00LG3bUKCLZAI1R3NeY2uEqvL0DbeAAl0kM5x9Npx+upqjOaotEqc9mqCyNH0NUVCgi2SGzubou+/C8uV+VyMp1tDqnVBUWZqe4887KdBFMoWaozlrw64WAEZWlqZ1Oz0KdDO73MzWmlm9md3xGetdZWbOzGpSV6JInqio8CaSVnM056zetheACcPL07qdbgPdzALAQ8AVwERgrplNPMJ65cBfAZqGRaS3wmFoafFCXXLGmq3NVJUVpfUYdOjZHvo5QL1zbr1zLgI8Bcw6wnrfAO4H2lNYn0h+OeccmDJFF+zKMWu27WXCsPTunUPPAn0EsKnL483JZfuZ2VnAKOfcf33WC5lZ2MzqzKxu586dx1ysSM4z8/bS335bzdEcsmF3CydW9Uv7do67KWpmBcD3gAXdreucW+icq3HO1QwePPh4Ny2Sm264AUpKtJeeI5rbozS3xxhZWZL2bfUk0LcAo7o8Hplc1qkcOA34bzP7BJgGLFJjVKSXOpujTzwBe/f6XY0cpy2N3oTQ6T7CBXoW6MuA8WZ2opkVAnOARZ1POuf2OOeqnHPVzrlqYCkw0zmnKw2J9FY4DPv2qTmaAzYnA31EJuyhO+diwG3Ai8Bq4Gnn3Eozu9fMZqa7QJG8dO65MHmyhl1ywObGVoA+GXIJ9mQl59wLwAuHLLvnKOtedPxlieS5zubo177mNUenTvW7IumlLY1tFIcK0nod9E46U1QkU3U2R3/4Q78rkeOwubGNkZWlmKVnlqKuFOgimWrAAJg9Gx5/3BtPl6y0qbG1T4ZbQIEuktnUHM1qDS0R1mzby8Th/ftkewp0kUw2bRqcdpqao1loT1uUq37wB+IJx8wzTuiTbSrQRTJZZ3O0rs47e1SyxtefeY+Pd7Vw76xJTBimPXQRAa85Wlys5mgWeevjBn63ajvX1ozkpunVfbZdBbpIpqusVHM0y3z716spLwqyoPaUPt2uAl0kG4TD3mUAnnrK70qkG2u37eXtjU389WUnM7R/cZ9uW4Eukg2mT4dJk9QczQJ/+GgXAJefNqzPt61AF8kGnc3RZcvgnXf8rkaOYktTG9/69RqG9i/ihIq+3TsHBbpI9lBzNKM557jvV6uIxBI8dN1ZfXJm6KEU6CLZYuBAuOYa+NnPvGnqJGOs276Xz31nMb9esY1zTxxITfVAX+pQoItkk/nzveboz3/udyWS9NjSDVzx4OvsaYvyjzMn8fifnetbLQp0kWxy3nkwcSI8/LDflQjw8urt/P0vV3DhyYNZ/PWL+PJ51QQD/sWqAl0km3Q2R996C9591+9q8t6yTxoJBYyHb5xKVVmR3+Uo0EWyzo03QlGRmqMZYGNDCyMrSwn5uFfeVWZUISI9p+ZoRnju7c38btV2poys8LuU/RToItlo/nxoboann/a7krz1/VfqicYd//uPJ/ldyn4KdJFsdP75cOqpao76JBZPsLGhlVsvOomBfTC1XE8p0EWyUWdz9M034b33/K4m77y0egexhGPckDK/SzmIAl0kW6k56pvfrtxGZWmIL04Z7ncpB1Ggi2SrQYPg6qvhscegtdXvavJGezTOqx/uZPpJgygKBvwu5yAKdJFsFg6rOdrHHl3yCbtbItw4rdrvUg4T9LsAETkOn/scTJjgXVZ33jy/q8lZ0XiC+361it+t2s6ne9r5wqlDmDbWn+u1fBbtoYtks87m6JIl8MEHfleTk5xzXP/DN/npkg2cMqycu648lQfnnOnL1RS7o0AXyXY33QSFhZr8Ik1WbGnmrU8a+JvLTubHXzmHWy4cS7+izBzcUKCLZDs1R9OmLRLn319ZRyhg3DR9jN/ldEuBLpILwmHYsweeecbvSnJGIuG4/kdL+e2q7dx8wVgGlGbOCURHo0AXyQUXXginnKJhlxToiMV5b1MTt//ifd7e2MS9syZxxxUT/C6rRzJzIEhEjk1nc3TBAlixAk47ze+Ksk40nuDJtzbyz79Zy96OGIXBAm6YNpo5Z4/2u7Qe69EeupldbmZrzazezO44wvN/Y2arzOx9M3vZzDJ/sEkk16g52mtNrREu/e6r3PP8SkZUlvCvs89gyR2XcN+fTKYwmD0DGd3uoZtZAHgIuAzYDCwzs0XOuVVdVnsHqHHOtZrZXwDfAWano2AROYqqKrjqKq85+u1vQ2mp3xVlvE92tfAfv/+YN9btYmNDK9+/7kyuOG04gYLMOySxJ3ryv55zgHrn3HrnXAR4CpjVdQXn3GLnXGd7fSkwMrVlikiPhMPQ1ATPPut3JRkrFk/w4sptfPOF1cxZuJRHl2ygf0mI+6+azB9NOSFrwxx6NoY+AtjU5fFm4LNmQb0Z+PWRnjCzMBAGGD06e8alRLLG5z8PJ5/sDbvcdJPf1WSU1kiM//XcByxZv5vtzR2EAsbkERX84IazOHN0pd/lpURKm6JmdgNQA3z+SM875xYCCwFqampcKrctIhxojn7967ByJUzKnMkX/NQaiTH/seW8Ub+LL5w6lEsmDOGLU4bTvzjkd2kp1ZMhly3AqC6PRyaXHcTMvgDcBcx0znWkpjwROWZf/rKao11s3N3Kpd99ldfX7eI7V03hhzfVMPec0TkX5tCzQF8GjDezE82sEJgDLOq6gpmdCTyMF+Y7Ul+miPRYVRV86Uvw6KPQ1uZ3Nb568KV1XPjPi9m6p52FN07lmppR3f9SFus20J1zMeA24EVgNfC0c26lmd1rZjOTq/0zUAY8Y2bvmtmio7yciPSFPG+O7t7XwfPvbuGBlz5k/JAyFt12PrWThvldVtqZc/4MZdfU1Li6ujpfti2S85zzmqPDhsHrr/tdTdptbmzlpVXbeWXtTj7Y3ERjaxSAytIQLy+4KKPm/TxeZrbcOVdzpOd0pqhILupsjt5+O6xaBRMn+l1RWjjnWPDMe/znO1twDk4a3I8Zk4YxbkgZ44eWc8aoAVSU5N5Y+dEo0EVy1bx5cNdd3pyjDzzgdzVp8cDvPuS5t7dwbc1Ibr1oHNVV/fwuyVfZc06riBybwYO95uhPf5qTzdF3NzXxb6/UUzOmkm9/aUrehzko0EVyWzgMjY3wi1/4XUlKtUXifOuF1QB8+6rJFGTx2Z2ppEAXyWUXXQTjxuXcMen/99WPePPjBr5z1RTGDSn3u5yMoUAXyWUFBXDLLd6RLqtX+11NSkTjCX7w6kdcMK6Ka8/O7ePKj5UCXSTXzZsHoZDXHM0BO/Z2EIkl+OKU4X6XknEU6CK5bsgQ+NM/9Zqj7e1+V3Pctu3x3sOw/sU+V5J5FOgi+SAchoaGrG+OxuIJvve7tQAMq1CgH0rHoYvkg4svhpNO8pqj11/vdzXHZM22Zl74YBtL1+/mw+17aWqNctnEoUwYpmbooRToIvmgszl6xx2wZg1MyNxJj9sicVZtbWbp+t3UfdLAa+t2EU84Jo+oYMbEYVx48mBmTBqKmQ5VPJQCXSRfzJsHd9/tNUe/+12/q9nPOcdbHzfw+/pdrNm2l5fX7CCe8K4xNX5IGXPOHsWtF49jxIASnyvNfAp0kXwxdOiB5ug//RMU+zcG3dQa4allm3h3YxOL1+6gI5bADIaWFzPvvGpqxlRy7thBOXVRrb6gQBfJJ+EwPPMMPPccXHddn2++IxbnzfUNPPjyOpZvaGRweRFXTx3J6SMHcNnEoVQqwI+LAl0kn1xyCYwd6zVH0xTokViCdzc1sXprMw0tEXa3dLClsY0tTW1sbGilPZogUGA8OOcMZp0xIi015CsFukg+6WyO3nknrF0Lp5ySkpdtbo/y+oe7eOvj3by4cjvbmg8c7z6gNMSIASWMGdSP806qYtrYgdRUD6SqrCgl25YDFOgi+WbePPj7v/eao//yL71+mT1tUR5542Ne/XAnqz9tJhJPEAoYnxs/mDuvnMD0sYMYVFZEQBfO6jMKdJF8M2wYzJoFP/mJ1xwt6n5P2TnHht2tvLe5ifU7W3h7YyOrPm2moTXC2WMG8pXzq7l4whBOHzmAksJA+t+DHJECXSQfzZ/vnTX63HMwd+7+xYmEY3dLhOUbGtjS1M4nu1pYtbWZTQ2t7NjbsX+9sVX9mHbSIL56fjVTxwz04x3IESjQRfJELJ6goTVCQ0uEdVWncuHwkTR+6wEeLDiV3S0Rdu3t4ONdLbRF4/t/pyQUYPKICi4YV8WZYyo5u7qSEwaUUF4U1Ik9GUiBLpKlEgnH+l0tNLREaGyNsKc1SlNbhMbWKE2tUZpaI95tW5Qdze3sbokc9Pu3jr+E2197lG3L3qO1+iSG9i9i2thBDKsoYuqYgYyt6seA0pCCO4so0EUyiHOOjliClo4YLR1xGlojbNvTRmNrlIaWCE2tEXbs9Q4D7AzzQwULjAGlhVSWhvYfYXLGqAEMKS+iqqyQgf2KGD2wlHG3ngEnPcGTtgJu+7IP71ZSTYEukkKtkRg793awtz3mhXIkxr6OeDKgvZDe1xFle3MHe9qi7G2Psq8jxt72GPvavfWTZ70fUXGogKqyIkZWlnDphCGcc+JAhleUMCAZ3gNKC+lXGOjhXnUFzJzpNUfvu69HzVHJbAp0yUvxhCMaTxCNJ+iIJWiPxmnpiNMaidEWjdMWie+/bY9691s64jQmx6D3tsdoi8Rp6bJ+S0eM5vZYt9suChYwtH8xlaUhyotDDCkvprw4SFlxkLKiIKWFQfoVBSgtDFJREmJ4RTGDygqpLC2kOJTiI0jmz/cao7/8JcyendrXlj6nQJc+k0g4ookEsbgjFndE4gliycdeuHq3kbg35LC3PUZrMlA7A3dve5T2WJyOqLde520k5oVzLBnUkdjBy6Nxt39ZNJ7AfcZe8GepKAkxsF8h/YuDlBQGGNq/mJLCACWhAGVFQarKChlWUUJZUZDy4iD9ioKUFQXo1xnUhQGCgQyahuALX4Dqanj4YQV6DlCg57BYPMGetuiBPchIfH+jbG9HjFj8QNjF4o5Ywgs8L3ATRJK3sS57s7G4I5pwB/9uMpS7/m40+TudYR1LuP1X0DsepYUBikMBioIFFAYL9t8WBgoIBgooDhXQvzhIKFBAKFhAUaCAUMBbx1tmFCaXBQPe/aJgAUXBAKVFAfoVBikOBSgtDOwP6uKQd784WJBZYZwKnWeO3nUXfPghnHyy3xXJcVCgZ4hoPLH/T/v2SMIL4ag3BNAejdMWSdCSHJ/d0dzOvo4DQwEdsTjt0QQdh+y57trXQewYQrTAIBjoDEcjWFBAKGD7wy9UkLwNeMuDBQUUh+zA+oECQgXJ2y7rHPQagQKCBV1e46DHB0K2rDi5h1sYpChU4IVsKMP2bnPFV74C99wDP/oRfOc7flcjx0GBfgyiyT3ePW1RmtuiyQZXjNZIjJZInLZIjLZkGLd3+fHCOXH4skiCjuT9Ywne/sVByotD3l5jqIDioLcnWVESOmiPdXB5EUPKiygt9IYHSgsDDCgNUVlaSHlx6LCwLtAp2vlp+HCvOfrjH8M3vqHmaBbL+UB3ziX3dOO0dsRpjcYO3I8k70cO3G+JxGhu8/aEm5Ph3fnT9YSLz1IYLKA4WJAMXC9si0IBSkIFDOxXSMlBy7xQ7vzTvjhUQElh0FunMLln2vk4FGBweZFOrZbUC4fhP/8Tnn8err3W72qkl7Iu0Pd1xPh/733KjuaO5J5xjNYOL4hbI/Euh4bF2NMWpSUSO6YGWGHQG4OtKitiQGmI6qpSKkpC9C8OUVESoqLUuy0vDlJWFKJfcty1tOhA8OpiRJJ1amthzBjvsroK9KyVdYH+8Ksf8e+v1APe4V/e0QOB5OFe3tEEg8uL6FcUTAZviNLkcENpYfCI9/sVJYckNEYr+aqzOXr33VBfD+PG+V2R9IK5Huy+mtnlwINAAPiRc+7bhzxfBDwKTAV2A7Odc5981mvW1NS4urq6Yy64oSXCx7taOH1khcJXJJU+/RRGj4YFC+D++/2uRo7CzJY752qO9Fy3iWhmAeAh4ApgIjDXzCYestrNQKNzbhzwAJC2b8PAfoVMHVOpMBdJtRNOgD/+Y685Gjn8kgKS+Xoy5HIOUO+cWw9gZk8Bs4BVXdaZBfxD8v6zwPfNzFxPdv9FJHOEw95Zo0OGwAhND5c299yTlhO5ehLoI4BNXR5vBs492jrOuZiZ7QEGAbu6rmRmYSAMMHr06F6WLCJpU1vrTU9XX0+vT6eV7lVWpuVl+7Qp6pxbCCwEbwy9L7ctIj0QCMA3v+l3FdJLPRmI3gKM6vJ4ZHLZEdcxsyBQgdccFRGRPtKTQF8GjDezE82sEJgDLDpknUVA5wWVrwZe0fi5iEjf6nbIJTkmfhvwIt5hi//hnFtpZvcCdc65RcAjwGNmVg804IW+iIj0oR6NoTvnXgBeOGTZPV3utwPXpLY0ERE5FjqYW0QkRyjQRURyhAJdRCRHKNBFRHJEjy7OlZYNm+0ENvTy16s45CxU0WdyCH0eh9NncrBs/TzGOOcGH+kJ3wL9eJhZ3dGuNpav9JkcTJ/H4fSZHCwXPw8NuYiI5AgFuohIjsjWQF/odwEZSJ/JwfR5HE6fycFy7vPIyjF0ERE5XLbuoYuIyCEU6CIiOSLrAt3MLjeztWZWb2Z3+F1PXzCzUWa22MxWmdlKM/ur5PKBZvY7M1uXvK1MLjcz+7fkZ/S+mZ3l7ztIDzMLmNk7Zvar5OMTzezN5Pv+efJyz5hZUfJxffL5aj/rThczG2Bmz5rZGjNbbWbT8/k7Ymb/M/nfywoze9LMinP9O5JVgd7DCatzUQxY4JybCEwD/jL5vu8AXnbOjQdeTj4G7/MZn/wJAz/o+5L7xF8Bq7s8vh94IDlZeSPe5OXQh5OY++xB4DfOuQnA6XifTV5+R8xsBPA/gBrn3Gl4l/6eQ65/R5xzWfMDTN9OHtAAAAJmSURBVAde7PL4TuBOv+vy4XN4HrgMWAsMTy4bDqxN3n8YmNtl/f3r5coP3sxZLwOXAL8CDO+sv+Ch3xW8a/lPT94PJtczv99Dij+PCuDjQ99Xvn5HODDP8cDkv/mvgBm5/h3Jqj10jjxhdV5NTZ78U/BM4E1gqHNua/KpbcDQ5P18+Jz+FbgdSCQfDwKanHOx5OOu7/mgScyBzknMc8mJwE7gx8lhqB+ZWT/y9DvinNsC/AuwEdiK92++nBz/jmRboOc1MysDfgH8tXOuuetzztu1yItjUM3sj4AdzrnlfteSQYLAWcAPnHNnAi0cGF4B8u47UgnMwvsf3QlAP+ByX4vqA9kW6D2ZsDonmVkIL8wfd849l1y83cyGJ58fDuxILs/1z+l8YKaZfQI8hTfs8iAwIDlJORz8nvNhEvPNwGbn3JvJx8/iBXy+fke+AHzsnNvpnIsCz+F9b3L6O5Jtgd6TCatzjpkZ3rytq51z3+vyVNfJub+MN7beufym5JEM04A9Xf7sznrOuTudcyOdc9V434FXnHPXA4vxJimHwz+PnJ7E3Dm3DdhkZqckF10KrCJPvyN4Qy3TzKw0+d9P5+eR298Rvwfxe9HsuBL4EPgIuMvvevroPV+A96fy+8C7yZ8r8cb4XgbWAS8BA5PrG97RQB8BH+B1+n1/H2n6bC4CfpW8PxZ4C6gHngGKksuLk4/rk8+P9bvuNH0WZwB1ye/JL4HKfP6OAP8IrAFWAI8BRbn+HdGp/yIiOSLbhlxEROQoFOgiIjlCgS4ikiMU6CIiOUKBLiKSIxToIiI5QoEuIpIj/j+VIgjad1kLeQAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BNl_PinY-aMU",
        "colab_type": "text"
      },
      "source": [
        "## The new untested optimisation"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "judJsZLL7jji",
        "colab_type": "text"
      },
      "source": [
        "15 Minutes before the deadline, we knew that we overfitted the optimisation.\n",
        "We figured out a method to decreases the overfitting but unfortunately we couldn't test it."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "P731j7R8_HDw",
        "colab_type": "text"
      },
      "source": [
        "The main idea is making a step that depends on the difference in the index between each two lines (we could use likely Increasing Index too) ."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WDexcSrGw1cA",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 202
        },
        "outputId": "745df206-ccb0-477a-8b2e-c065c68ebe0c"
      },
      "source": [
        "#We aren't so sure that allocating 100 Zillion DSML to the 340 first rows is optimal  \n",
        "ss.opt_pred=0\n",
        "n = 340\n",
        "ss.opt_pred[:n]=100\n",
        "ss.opt_pred[n:] = 0\n",
        "c=100\n",
        "#This is a simple test we could add more steps or maybe changing these one\n",
        "for i in range(n,890):\n",
        "  if ss.pred[i]-ss.pred[i-1]<4*10**(-6):\n",
        "    alpha=0.13\n",
        "  if ss.pred[i]-ss.pred[i-1]>=4*10**(-6) and ss.pred[i]-ss.pred[i-1]<4*10**(-5):\n",
        "    alpha=0.21222895\n",
        "  if ss.pred[i]-ss.pred[i-1]>=4*10**(-5) and ss.pred[i]-ss.pred[i-1]<10**(-4):\n",
        "    alpha=0.33\n",
        "  if ss.pred[i]-ss.pred[i-1]>=10**(-4) and ss.pred[i]-ss.pred[i-1]<5*10**(-3) :\n",
        "    alpha=0.48\n",
        "  if ss.pred[i]-ss.pred[i-1]>=5*10**(-3) and ss.pred[i]-ss.pred[i-1]<10**(-3):\n",
        "    alpha=0.58\n",
        "  if ss.pred[i]-ss.pred[i-1]>=10**(-3):\n",
        "    alpha=0.73\n",
        "  c-=alpha\n",
        "  if c-alpha>0:\n",
        "    ss.loc[i,'opt_pred'] =c\n",
        "  else:\n",
        "    ss.loc[i,'opt_pred'] =0\n",
        "\n",
        "print(ss.opt_pred.sum())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:4: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  after removing the cwd from sys.path.\n",
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:5: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  \"\"\"\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "49999.999927200086\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jszq6xGT_0F3",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 67
        },
        "outputId": "dbf92b5b-4f64-40ae-c640-e9025cbf9575"
      },
      "source": [
        "#Checking if the sum of opt_pred in rows having eei<0.7 is >5000\n",
        "print(\"sum\",ss.opt_pred.sum())\n",
        "print(\"left\", (50000-ss.opt_pred.sum()))\n",
        "print(\"eei\",ss[ss.eei<0.7]['opt_pred'].sum())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "sum 49999.999927200086\n",
            "left 7.279991405084729e-05\n",
            "eei 6542.4094493500015\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kbtD897Fj3tc",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "outputId": "448eceb1-92af-47f3-b161-118caefdd9a1"
      },
      "source": [
        "plt.plot(ss.pred)\n",
        "plt.plot(ss.opt_pred/100,color=\"red\")\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAD4CAYAAAD8Zh1EAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nO3deZRU1bn+8e9bQ89Nd0MzySCiKIIT2nGIRo2JBjSCK3oVjOIAIorTdSYafzdqjEMiEiMqEmMk/kJAvcpVonEg3piIAsaBWVRGZWzGnqqrat8/ToHNIDRQ1aeG57NWr65z6nSdt8vicfc+++xtzjlERCTzBfwuQEREkkOBLiKSJRToIiJZQoEuIpIlFOgiIlki5NeJKysrXbdu3fw6vYhIRpo5c+Ya51zbnT3nW6B369aNGTNm+HV6EZGMZGaLv+05dbmIiGQJBbqISJZQoIuIZAkFuohIllCgi4hkid0Gupk9bWarzGzWtzxvZvZbM1toZp+Y2dHJL1NERHanOS30Z4C+u3i+H9Aj8TUMeHzfyxIRkT2123Hozrn/NbNuuzhkAPCs8+bhnWZm5WbW0Tn3dZJq3Na778Lf/paSlxaftGsHI0aAmd+ViGS0ZNxY1AlY2mR7WWLfDoFuZsPwWvF07dp178723ntw771797OSfrbMx9+7N3z/+/7WIpLhWvROUefcWGAsQFVV1d6trHHLLd6XZIeaGjjwQBg2DD74ACoq/K5IJGMlY5TLcqBLk+3OiX0iu1dcDC+8AIsXw8CBEI36XZFIxkpGoE8GBidGuxwPbEhZ/7lkpxNPhDFjvGsjd97pdzUiGas5wxb/DLwHHGJmy8xsiJkNN7PhiUOmAF8AC4GngKtTVq1kr6FDYfBgeOQRWL3a72pEMpL5tUh0VVWV02yLso3Zs+Hoo+G002DKFI16EdkJM5vpnKva2XO6U1TSR+/e8MAD8Npr8K9/+V2NSMZRoEt6GToUSkrgjjt0gVRkDynQJb2UlMCjj8I778DEiX5XI5JRFOiSfgYPhu7dYdw4vysRySgKdEk/gQAMGQJTp8LChX5XI5IxFOiSni691Av2n//c70pEMoYCXdLTfvvBTTfBhAleS11EdkuBLunr7ruhUye47z6/KxHJCAp0SV8FBXDllfDmm+pLF2kGBbqktyFDIBj0hjKKyC4p0CW97bcfXH65F+ivveZ3NSJpTYEu6e+RR+DQQ+GGGyAe97sakbSlQJf0V1QEt98O8+fDSy/5XY1I2lKgS2a44AKvlX7nnd8sWyci21CgS2bIy4PbboO5c+E3v/G7GpG0pECXzDF4MPTr502x29jodzUiaUeBLpnDDK66CtasUV+6yE4o0CWz9O0LvXp50wI0NPhdjUhaUaBLZgmH4cEHYelSePFFv6sRSSsKdMk8P/oRHH44jBihKQFEmlCgS+YJhbw+dDMYMAA2bfK7IpG0oECXzNS9O0ya5N1sdPHFuoNUBAW6ZLLTToNRo+Dll+EXv/C7GhHfKdAls11zDQwa5I1NX7vW72pEfKVAl8xmBiNHQizmBbtuOJIcpkCXzHf44TB2LLzxBlx7reZ6kZwV8rsAkaS47DJYsADuvx969PBuPBLJMWqhS/b45S/hvPPglltg2jS/qxFpcQp0yR6BAPzhD1BcrIWlJScp0CW7lJR4i2H8z//A9Ol+VyPSohTokn2uvdYL9lGj/K5EpEU1K9DNrK+ZzTezhWZ2+06e72pmU83s32b2iZmdmfxSRZqpVStvmt0JE2DZMr+rEWkxuw10MwsCjwH9gF7AIDPrtd1hdwITnXN9gIHAmGQXKrJHrrzS61MfMULDGCVnNKeFfiyw0Dn3hXMuAkwABmx3jANaJR6XAV8lr0SRvXDggfDQQzB5Mkyc6Hc1Ii2iOYHeCVjaZHtZYl9T/wVcZGbLgCnAtTt7ITMbZmYzzGzG6tWr96JckT1w3XVw5JFw661QV+d3NSIpl6yLooOAZ5xznYEzgfFmtsNrO+fGOueqnHNVbdu2TdKpRb5FMAijR8OSJXDzzX5XI5JyzQn05UCXJtudE/uaGgJMBHDOvQcUAJXJKFBkn5xyinfX6Jgx3lBGkSzWnECfDvQwswPMLA/voufk7Y5ZAvwAwMwOxQt09alIerjvPjjsMBg61Fu6TiRL7TbQnXNR4BrgdWAu3miW2WZ2t5n1Txx2E3CFmX0M/Bm41DkNLZA0kZfnXRitrYUhQ7QYhmQt8yt3q6qq3IwZM3w5t+SoMWO8YYwjR2pqAMlYZjbTOVe1s+d0p6jkjquuggsvhF/9Cp57zu9qRJJOgS65w8ybvOuUU7yul1mz/K5IJKkU6JJbtvSnl5TA5ZdDJOJ3RSJJo0CX3NOuHTz5pDcb4w03+F2NSNIo0CU3nXuutxDG44973TAiWUCBLrnrvvvgBz/wRr58/bXf1YjsMwW65K5QyGuh19XBU0/5XY3IPlOgS27r0QPOPtsbyrh8+xktRDKLAl3kkUegvh6eeMLvSkT2iQJdpHt3+MlP4IEHYOZMv6sR2WsKdBGAceOgvBxuu83vSkT2mgJdBKCiAu64A956C555xu9qRPaKAl1ki+HDvWkBhg2DTz/1uxqRPaZAF9kiPx+efx7KyuCcc2DlSr8rEtkjCnSRpior4eWXvSGMt97qdzUie0SBLrK9737Xm+Nl/Hj48EO/qxFpNgW6yM6MHAmtW3uLS2vxLckQCnSRnSkrg1/8AqZOhUcf9bsakWZRoIt8m6uuggED4MYbYd48v6sR2S0Fusi3CQS8edO3fBdJcwp0kV1p394bwvjHP3qzMoqkMQW6yO6MGAHr1sHvf+93JSK7pEAX2Z1TToHjjvPmTteIF0ljCnSR5rjiCpgzB6ZN87sSkW+lQBdpjgsugJISrWwkaU2BLtIcJSVw0UXw7LMwa5bf1YjslAJdpLnuvRfCYRgzxu9KRHZKgS7SXG3awPnnw5/+BJs3+12NyA4U6CJ7YsQI2LQJrr/e70pEdqBAF9kTxx7rTdz19NMwfbrf1Yhso1mBbmZ9zWy+mS00s9u/5ZjzzWyOmc02s/+f3DJF0sjIkVBYCA895HclItvYbaCbWRB4DOgH9AIGmVmv7Y7pAYwETnTO9QZuSEGtIumhtBRuuQUmTYL33vO7GpGtmtNCPxZY6Jz7wjkXASYAA7Y75grgMefcOgDn3KrklimSZm691RvK+OCDuntU0kZzAr0TsLTJ9rLEvqYOBg42s3+a2TQz67uzFzKzYWY2w8xmrF69eu8qFkkHxcXws5/BSy/BhAl+VyMCJO+iaAjoAZwKDAKeMrPy7Q9yzo11zlU556ratm2bpFOL+OS22+CQQ2DUKLXSJS00J9CXA12abHdO7GtqGTDZOdfonPsSWIAX8CLZKxCA//xPb7SLWumSBpoT6NOBHmZ2gJnlAQOBydsd8xJe6xwzq8TrgvkiiXWKpKehQ72ZGIcPh88+87sayXG7DXTnXBS4BngdmAtMdM7NNrO7zax/4rDXgbVmNgeYCtzinFubqqJF0kYwCBMnQjTqXSAV8ZE5n/r+qqqq3IwZM3w5t0jSDRkCf/kLfPUVtGrldzWSxcxspnOuamfP6U5RkWQYPhxqarwLpCI+UaCLJENVFQwY4AW61h4VnyjQRZLBDK67DjZs8Mami/hAgS6SLKeeCvvvD3/4g9+VSI5SoIskSyAAl1wCb74JS5b4XY3kIAW6SDJdeql31+izz/pdieQgBbpIMh1wgNf18swzmg5AWpwCXSTZLr8cPv8c/vu//a5EcowCXSTZBg2Cww7z5nnR2qPSghToIskWCsGYMbB0Kdxzj9/VSA5RoIukwve+B/37exdHYzG/q5EcoUAXSZXBg2HFCnjiCb8rkRyhQBdJlXPOgdNPhzvuAK3QJS1AgS6SKoEAjB7tXRj9+c/9rkZygAJdJJUOPdSb4+XJJ2HqVL+rkSynQBdJtfvug86d4e67/a5EfBaNxXlt1gqWr0/NjJwKdJFUKyiAa66Bv//dW39UctbamgjD/zSTv89flZLXV6CLtIRhw6BjR7jxRr8rER+t2dwAQJvivJS8vgJdpCVUVMDNN8O778KHH/pdjfikuiYCQJuS/JS8vgJdpKVcfjmUl8OvfuV3JeKTLYHeWi10kQxXXg5XXOFN2vX1135XIz5YsznRQlegi2SBK67wpgJ46im/KxEfVNc0EAoYrQrCKXl9BbpIS+rRA378Y6/bZc0av6uRFlZdE6GiOI9AwFLy+gp0kZZ2zz1QXw8TJ/pdibSwNZsjKetuAQW6SMs78kg44gh46CEv2CUnLFtXy7+XrKcyRSNcQIEu0vLM4Je/hEWL4NVX/a5GWshNEz9mzeYGBp+wf8rOoUAX8UO/ft6NRo8+qrVHc8SS6lp+0qcTZ/TukLJzKNBF/BAMwm23wTvveFMCSNZbVxuhsjR13S2gQBfxz7Bh3tj0ceP8rkRSrC4So74xTkVR6i6IggJdxD+FhXDRRfDCC1Bd7Xc1kkLVtd4NRRVFqRl/voUCXcRPQ4dCQ4M3X7pkrcVragDoXFGU0vM0K9DNrK+ZzTezhWZ2+y6OO9fMnJlVJa9EkSx25JHeYtL33ANz5vhdjaTI3BWbAOjZsTSl59ltoJtZEHgM6Af0AgaZWa+dHFcKXA+8n+wiRbLa449Dq1YwcKBGvGSpeV9vpLIkP6Vj0KF5LfRjgYXOuS+ccxFgAjBgJ8fdAzwA6E4JkT2x337w4IPw6afw5pt+VyMpMG/FJnp2SG3rHJoX6J2ApU22lyX2bWVmRwNdnHO7vEvCzIaZ2Qwzm7Faq6CLfOOCC6BLF7jkEvjyS7+rkSRbvLaGAyqLU36efb4oamYB4GHgpt0d65wb65yrcs5VtW3bdl9PLZI98vNhyhSoqYGbdvtPSTLIxvpGNtZH6VxRmPJzNSfQlwNdmmx3TuzbohQ4DPi7mS0Cjgcm68KoyB467DC4+mp4+WWYN8/vaiRJlq/zFoRO9QgXaF6gTwd6mNkBZpYHDAQmb3nSObfBOVfpnOvmnOsGTAP6O+dmpKRikWx23XVQVga33OJ3JZIkyxKB3ikdWujOuShwDfA6MBeY6JybbWZ3m1n/VBcoklM6doRrrvEm7Zo92+9qJAmWrasFSJsuF5xzU5xzBzvnDnTO/TKx7y7n3OSdHHuqWuci+2DECKishJ/8BDZs8Lsa2UfL19VREA6kdB70LXSnqEi6ad8enn8ePvsMfv1rv6uRfbRsXR2dK4owS80qRU0p0EXS0cknw1lnwdNPe2uQSsZauq62RbpbQIEukr4uvhi++krT62aw6poI81ZsolfHVi1yPgW6SLo6+2xvSoCHH9aUABloQ10j5z7+L2JxR/+j9muRcyrQRdJVYSHcead3w5GmBMg4N0/6mC/X1HD3gN707KAWuohcfz106ACPPOJ3JbIHPviymjfmrOT8qs4MPqFbi51XgS6SzvLyYPhwr5WuvvSMcf9f51KaH+KmMw5p0fMq0EXS3XXXQffu3vS6Wtko7c1fsYkPl6znhtMPpn2rghY9twJdJN1VVHjj0teu9bpgJK396/M1APQ9rEOLn1uBLpIJ+vSBa6+FCRNg5Uq/q5FvsXx9Hb/66zzat8pnv7KWbZ2DAl0kcwwbBtEoPPus35XITjjnuPeVOUSicR678OgWuTN0ewp0kUzRsyecdBKMG6dx6Wnms5Wb+N6DU/nrrBUcd0Brqrq19qUOBbpIJhk2DBYsgFde8bsSSRg/bTH9Rv+DDXWN/KJ/b54bepxvtSjQRTLJoEHeGqSjRkE87nc1Oe+tuSv5+UuzOPngtky9+VQu+W43QkH/YlWBLpJJQiEYORKmToWxY/2uJudNX7SOcNB48uJjqCzJ97scBbpIxhkxAo45Bp54wu9Kct6S6ho6VxQR9rFV3lR6VCEizWcGl10GH38MH33kdzU568UPl/HGnJUc0bnM71K2UqCLZKKBA71pAZ55xu9Kctbv3l5IY8zx/87u7XcpWynQRTJRmzbQvz889xxEIn5Xk3OisThLqmu5+tQDad0CS8s1lwJdJFNddhmsWeMtKC0t6s25q4jGHQe1K/G7lG0o0EUy1RlneFPrqtulxf1t9goqisKcdURHv0vZhgJdJFOFQt4yda++CitW+F1NzqhvjPHOgtWccGAb8kNBv8vZhgJdJJNdcYU3DcDo0X5XkjOefW8Ra2siXHx8N79L2UHI7wJEZB/06AFnnQXjx8O990IwvVqM2aIxFufeV+bwxpyVfLWhnh8e2o7ju/szX8uuqIUukukuugiWL9csjCninOOnT73PH99bzCEdSrnjzEMZPbCPL7Mp7o5a6CKZ7txz4bvf9aYEOP98KC72u6KsMmv5Rj5YVM2Npx/MdT/o4Xc5u6QWukimCwbhgQe8hS/GjPG7mqxSF4nx6NufEQ4ag0/Y3+9ydkuBLpINTjoJTjnFC/S6Or+ryQrxuOOn46bxtzkrGXJSd8qL0ucGom+jQBfJFrfeCosWwR//6HclGa0hGuPjpeu59YVP+HDJeu4e0Jvb+/X0u6xmUR+6SLbo1w969fIC/corvUm8pNkaY3H+/MESHnptPpsaouSFAlx0fFcGfqer36U1W7MC3cz6AqOBIDDOOXf/ds/fCAwFosBq4HLn3OIk1yoiu2LmrWh0ww3w8stwzjl+V5Qx1tdG6P+7f7KkupaeHUoZfsqBfK9HJW3SYI7zPWFuN2sTmlkQWACcDiwDpgODnHNzmhzzfeB951ytmV0FnOqcu2BXr1tVVeVmzJixr/WLSFOxGBx0kDclwLvvalz6bixaU8PT//ySdz9bwxdravjdhX3od1hHgoH0/evGzGY656p29lxz+tCPBRY6575wzkWACcCApgc456Y652oTm9OAzvtSsIjspWDQ60ufNg3uusvvatJSNBbn9dkruG/KXAaOncaz7y2mVWGYB849nB8fsV9ah/nuNKfLpROwtMn2MmBXq6AOAf66syfMbBgwDKBr18zplxLJKFdd5S1R99hjcMcdUFTkd0VpoTYS5Wcvfsp7X6xl5cYGwkHj8E5lPH7R0fTpWuF3eUmR1IuiZnYRUAWcsrPnnXNjgbHgdbkk89wi0sSIETBpErzwgjeBV46rjUS5cvxM3l24hh8e2p7TerbjrCM60qog7HdpSdWcQF8OdGmy3Tmxbxtm9kPgDuAU51xDcsoTkb1y8sleX/pTT+V8oC9ZW8sFY9/j6w31PHTeEfxHVZfd/1CGak4f+nSgh5kdYGZ5wEBgctMDzKwP8CTQ3zm3Kvllisge2TLi5R//gL/utAc0J4x+8zNOfmgqX2+oZ+zFx2R1mEMzAt05FwWuAV4H5gITnXOzzexuM+ufOOwhoASYZGYfmdnkb3k5EWkp118P3bp50wLkmLWbG3j5o+WMenMBPdqVMPmaEzmjdwe/y0q5ZvWhO+emAFO223dXk8c/THJdIrKv8vK8vvRbboH334fjdjWWIbMtW1fLm3NW8vb81Xy6bD3rahsBqCgK85crT0irdT9TSXeKimSz4cPhnnvg8cezMtCdc9w06WP++9/LcQ4ObFvMj3p34KB2JfRoX8pRXcopK8yuC5+7okAXyWYlJXDeeTBxIvzmN9Cmjd8VJdWoNxbw4ofLOb+qM1efehDdKnN76mBNziWS7a69FhoavPnSs8hHS9fz27cXUrV/Bff/5IicD3NQoItkv6OO8rpenn7am40xC9RFYvxqylwA7j/3cAIZfHdnMinQRXLBzTd7i0k/9ZTflSTFE+98zvtfVvPguUdwULtSv8tJGwp0kVzQtSuceabXSm9s9LuafdIYi/P4O59z0kGVnP+d7B5XvqcU6CK54uqrYcUK+NOf/K5kn6za1EAkGuesIzr6XUraUaCL5Iq+feGww2DsWL8r2ScrNtQD0KFVgc+VpB8FukiuMINLL/Wm1p03z+9q9ko0FufhN+YD0KFMgb49BbpILvnpT7050zNo3dF5Kzby8BsLOP/J96j65Zv8c+FaTu/Vnp4ddDF0e7qxSCSXdOjgrT06fjzce29armhUF4kx5+uNTPtiLTMWVfO/n60hFncc3qmMH/XqwMkHt+VHvdtjWjN1Bwp0kVxzySXwyivw1ltwxhl+V4Nzjg++rOafC9cwb8Um3pq3iljcWy6hR7sSBn6nC1d//yA6lRf6XGn6U6CL5Jqzz4aKCnjmGd8CfX1thAnTl/LRkvVMnb+KhmgcM2hfWsCl3+1G1f4VHNe9Tc5MqpUsCnSRXJOfD4MGeWPSN2yAsrIWO3VDNMb7X1Qz+q3PmLl4HW1L8znvmM4c2bmc03u1p0IBvk8U6CK56NJLYcwY+MtfvIUwkigSjfPR0vXM/Xoj1TUR1tY0sHxdHcvX17Gkupb6xjjBgDF64FEMOKpTUs+d68w5f5b2rKqqcjNmzPDl3CI5zzno0wfq62HWLAjtW9tuY30j/1iwhg++XMvrs1eyYmP91ufKi8J0Ki9kv/JCOpUXcnz31lR1a01lSf6+/hY5ycxmOueqdvacWugiucgM7roLzj0X/vznvVp3dENdI79/90veWbCauV9tJBKLEw4a3+vRlpFn9uSE7m1oU5JPUBNntRi10EVyVTzuzcQYDMKHH3oh/y2ccyxeW8vHy9bzxeoaPlyyjjlfbaS6NsJ39m9Nn67lfL9nO47sXE5hXvoNhcwmaqGLyI4CAbjqKm+Ol+nT4dhjiccda2sizFxczfL19SxaU8OcrzeytLqWVZsatv5o98pijj+wDZef2I1j9m/t4y8hTSnQRXJENBanujZCdU2Ez1ZuZkl1LZvbHMMN+YXMH3I9d1x2HwvXNVDXGNv6M4XhIId3KuOkgyrps38F3+lWwX7lhZTmh3RjTxpSoItkqHjc8cWaGqprIqyrjbChtpH1dRHW1TayvraR9bUR73tdI6s21rO2JrLDaxSGgwT7DePml0Yz4m/j+GDEHXQoy+eY/VvTvbKY8qKwgjuDKNBF0ohzjoZonJqGKDUNMaprI6zYUMe62kaqayKsr42wapM3DHBLmG8vFDDKi/KoKApvHWFyVJdy2pXmU1mSR+vifLq2LuKgdiWJ/u6+cOEq+k55lb4v/R4KNOlVplKgiyRRbSTK6k0NbKqPeqEcibK5IZYIaC+kNzc0snJjAxvqGtlU38jmhiib6qNsrveOj+9inEJBOEBlST6dKwr5Qc92HHtAazqWFVKeCO/yojyK84J73qq+7DJvtMurr3ojXyQjKdAlJ8XijsZYnMZYnIZonPrGGDUNMWojUeoaY9RFYlu/1zd6j2saYqxL9EFvqo9SF4lR0+T4moYoG+ujuz13fihA+1YFVBSFKS0I0660gNKCECUFIUryQxTlhSjOD1KUF6KsMEzHsgLalORRUZRHQThFI0hOO82buGv8eAV6BlOgS4uJxx2N8TjRmCMac0RicaKJbS9cve+RmNflsKk+Sm0iULcE7qb6RuqjMRoaveO2fI9EvXCOJoI6Et12f2PMbd3XGIuzt6N1ywrDtC7Oo1VBiMK8IO1bFVCYF6QwHKQkP0RlSR4dygopyQ9RWhCiOD9ESX6Q4i1BnRckFEzDWauDQbjwQnj0UVi7Ftq08bsi2QsK9CwWjcXZUNf4TQsyEtt6oWxTQ5Ro7Juwi8Yc0bgXeF7gxokkvkebtGajMUdj3G37s4lQbvqzjYmf2RLW0bjbOoPevijKC1IQDpIfCpAXCmz9nhcMEAoGKAgHaFUQIhwMEA4FyA8GCAe9Y7x9Rl5iXyjoPc4PBcgPBSnKD1KcF6IgHKQoL7g1qAvC3uOCUCA9wzhZLr4YHn4YJk2C4cP9rkb2ggI9TTTG4lv/tK+PxL0QbvS6AOobY9RF4tQk+mdXbaxnc8M3XQEN0Rj1jXEatmu5rtncQHQPQjRgEApuCUcjFAgQDtrW8AsHEt+D3v5QIEBB2L45PhggHEh8b3LMNq8RDBAKNHmNbba/CdmSgkQLNy9EfjjghWw4TVu32eLII70l6h58EAYPhqIivyuSPaRA3wONiRbvhrpGNtY1Ji5wRamNRKmJxKiLRKlLhHF9ky8vnOM77ovEaUg83pPgbVUQorQg7LUawwEKQl5LsqwwvE2LtW1pPu1K8ynK87oHivKClBeFqSjKo7QgvENYB3SLdm4zg9/+1utPHzcOrrvO74pkD2X9rf/OuURLN0ZtQ4zaxug3jyOJx5FvHtdEomys81rCGxPhveWr6Q0Xu5IXClAQCiQC1wvb/HCQwkRLs3CbfV4ob/nTviAcoDAv5B2Tl2iZbtkOB2lbmq9bqyW1jj0W6urgk092OR2A+COrbv3f3BDlfz7+ilUbGxIt4yi1DV4Q10ZiTYaGRdlQ10hNJLpHF8DyQl4fbGVJPuVFYbpVFlFWGKZVQZiywjBlRd730oIQJflhihP9rkX53wSvJiOSjDZ0KFx5JXzwARx3nN/VyB7IuEB/8p3PefTthYA3/MsbPRBMDPfyRhO0Lc2nOD+UCN4wRYnuhqK80E4fF+cnuiTURyviLX5x443efOkK9IzSrEA3s77AaCAIjHPO3b/d8/nAs8AxwFrgAufcouSW6rnsxAM49ZB2HNm5TOErkgqlpV4rffRouPVW6N3b74qkmXabiGYWBB4D+gG9gEFm1mu7w4YA65xzBwGjgAeSXegWrYvzOGb/CoW5SCr97GfebIyTJvldieyB5rTQjwUWOue+ADCzCcAAYE6TYwYA/5V4/DzwOzMz59cVVxHZN+3aecMYR41SqKfCXXfBBRck/WWbE+idgKVNtpcB23esbT3GORc1sw1AG2BN04PMbBgwDKBr1657WbKItIi77oLnnvO7iuxUUZGSl23Ri6LOubHAWPCGLbbkuUVkD51zjvclGaM5HdHLgS5Ntjsn9u30GDMLAWV4F0dFRKSFNCfQpwM9zOwAM8sDBgKTtztmMnBJ4vF5wNvqPxcRaVm77XJJ9IlfA7yON2zxaefcbDO7G5jhnJsM/B4Yb2YLgWq80BcRkRbUrD5059wUYMp2++5q8rge+I/kliYiIntCg7lFRLKEAl1EJEso0EVEsoQCXUQkS/g2H7qZrQYW7+WPV2E5TF8AAAO0SURBVLLdXaii92Q7ej92pPdkW5n6fuzvnGu7syd8C/R9YWYzvm2C91yl92Rbej92pPdkW9n4fqjLRUQkSyjQRUSyRKYG+li/C0hDek+2pfdjR3pPtpV170dG9qGLiMiOMrWFLiIi21Ggi4hkiYwLdDPra2bzzWyhmd3udz0twcy6mNlUM5tjZrPN7PrE/tZm9oaZfZb4XpHYb2b228R79ImZHe3vb5AaZhY0s3+b2SuJ7QPM7P3E7/2XxHTPmFl+Ynth4vluftadKmZWbmbPm9k8M5trZifk8mfEzP4z8e9llpn92cwKsv0zklGB3swFq7NRFLjJOdcLOB4Ykfi9bwfecs71AN5KbIP3/vRIfA0DHm/5klvE9cDcJtsPAKMSi5Wvw1u8HFpwEXOfjQZec871BI7Ee29y8jNiZp2A64Aq59xheFN/DyTbPyPOuYz5Ak4AXm+yPRIY6XddPrwPLwOnA/OBjol9HYH5icdPAoOaHL/1uGz5wls56y3gNOAVwPDu+gtt/1nBm8v/hMTjUOI48/t3SPL7UQZ8uf3vlaufEb5Z57h14r/5K8CPsv0zklEtdHa+YHUnn2rxReJPwT7A+0B759zXiadWAO0Tj3PhfXoEuBWIJ7bbAOudc9HEdtPfeZtFzIEti5hnkwOA1cAfEt1Q48ysmBz9jDjnlgO/BpYAX+P9N59Jln9GMi3Qc5qZlQAvADc45zY2fc55TYucGINqZj8GVjnnZvpdSxoJAUcDjzvn+gA1fNO9AuTcZ6QCGID3P7r9gGKgr69FtYBMC/TmLFidlcwsjBfmzznnXkzsXmlmHRPPdwRWJfZn+/t0ItDfzBYBE/C6XUYD5YlFymHb3zkXFjFfBixzzr2f2H4eL+Bz9TPyQ+BL59xq51wj8CLe5yarPyOZFujNWbA665iZ4a3bOtc593CTp5ouzn0JXt/6lv2DEyMZjgc2NPmzO+M550Y65zo757rhfQbeds79FJiKt0g57Ph+ZPUi5s65FcBSMzsksesHwBxy9DOC19VyvJkVJf79bHk/svsz4ncn/l5c7DgTWAB8Dtzhdz0t9DufhPen8ifAR4mvM/H6+N4CPgPeBFonjje80UCfA5/iXen3/fdI0XtzKvBK4nF34ANgITAJyE/sL0hsL0w8393vulP0XhwFzEh8Tl4CKnL5MwL8ApgHzALGA/nZ/hnRrf8iIlki07pcRETkWyjQRUSyhAJdRCRLKNBFRLKEAl1EJEso0EVEsoQCXUQkS/wfMQwyksnKs+gAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mSIBQCGGIykD",
        "colab_type": "text"
      },
      "source": [
        "#Submiting the result"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T68NiOFJ9t1b",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#Reordering the list using the real index\n",
        "ss=ss.sort_values('Index',ascending=True)\n",
        "ss=ss.reset_index(drop=True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NVdgjh37A5Zx",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "ss[['Index', 'pred', 'opt_pred']].to_csv('sub.csv', index=False)"
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}