Lectures/Lecture_f3dasm/presentation.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "02afc16b",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=./img/Brown_logo.svg width=30%>\n",
    "\n",
    "# F3DASM: Framework for Data-Driven Design & Analysis of Structures & Materials\n",
    "\n",
    "https://github.com/bessagroup/f3dasm\n",
    "\n",
    "### Martin van der Schelling | <a href = \"mailto: martin_van_der_schelling@brown.edu\">martin_van_der_schelling@brown.edu</a>  | PhD candidate"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "44553c83",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Outline of today"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7ea87d52",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "### Presentation\n",
    "\n",
    "* Motivation & Challenges\n",
    "* What's in the box?"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ef058087",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "### Practical session"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6fe9ac3a",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Motivation\n",
    "\n",
    "* Continuous search for the most efficient design based on specific criteria.\n",
    "* High-dimensional solution space due to the **overwhelming number of potential combinations**\n",
    "* Challenge to **accelerate** the optimization process. \n",
    "* Impractical to conduct experimental investigations of each concept. \n",
    "* Data-driven computational analyses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "eeab467d",
   "metadata": {
    "slideshow": {
     "slide_type": "notes"
    }
   },
   "outputs": [],
   "source": [
    "from IPython.display import Video"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7e7804e4",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<img src=./img/schematics/supercompressible_icon.png width=5%, align='right'>\n",
    "\n",
    "## Example 1: Designing Supercompressible materials"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "ada7f960",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<video src=\"./video/fragile_becomes_supercompressible.mp4\" controls  width=\"1024\"  height=\"576\">\n",
       "      Your browser does not support the <code>video</code> element.\n",
       "    </video>"
      ],
      "text/plain": [
       "<IPython.core.display.Video object>"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Video(\"./video/fragile_becomes_supercompressible.mp4\", width=1024, height=576)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cefa5109",
   "metadata": {},
   "source": [
    "*Bessa, M. A., Glowacki, P., & Houlder, M. (2019). Bayesian Machine Learning in Metamaterial Design: Fragile Becomes Supercompressible. Advanced Materials, 31(48), 1–6. https://doi.org/10.1002/adma.201904845*"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "624209aa",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<img src=./img/schematics/supercompressible_icon.png width=5%, align='right'>\n",
    "\n",
    "## Example 1: Designing Supercompressible materials\n",
    "\n",
    "<img src=./img/supercompressible/process.png width=60%, align='left'>\n",
    "\n",
    "<img src=./img/supercompressible/schematic.png width=30%, align='right'>\n",
    "\n",
    "<!-- <br><br><br><br><br><br><br><br><br><br><br><br><br><br><br> -->\n",
    "\n",
    "* Parametrizing your structure from a concept\n",
    "* Sampling designs\n",
    "* Run finite element simulation (ABAQUS)\n",
    "* Learn the objective for the full range of designs\n",
    "* Manufacture your found optimum"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fe55d14b",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<img src=./img/schematics/l2o_icon.png width=5%, align='right'>\n",
    "\n",
    "## Example 2: Learning to Choose Optimizers\n",
    "\n",
    "<img src=./img/l2o/Adam.gif width=40%, align='right'>\n",
    "\n",
    "\n",
    "\n",
    "<img src=./img/functions/Sphere.png width=15%, align='left'><img src=./img/functions/Schwefel.png width=15%, align='left'>\n",
    "\n",
    "<img src=./img/l2o/optimization.png width=30%, align='left'>\n",
    "\n",
    "\n",
    "* Optimizers performance is dependent on the loss-landscape characteristics\n",
    "* No 'one optimizer to rule them all'\n",
    "* Learn a model to determine the training dynamics in a data-driven manner"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9e7727c3",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Data-driven process\n",
    "\n",
    "Closer look at the two examples:\n",
    "\n",
    "<img src=./img/schematics/supercompressible.png width=100%>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8e6a2e08",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "<img src=./img/schematics/l2co.png width=100%>\n",
    "\n",
    "Looks like completely different studies .."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "424ef037",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Data-driven process\n",
    "\n",
    "Common *elements* in the data-driven process:\n",
    "\n",
    "<img src=./img/schematics/blocks.png width=100%>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c83ca292",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "* **Design of experiments**, in which input variables describing the microstructure, structure, properties and external conditions of the system to be evaluated are determined and sampled;"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e7e0b13f",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "* **Data generation**, typically through computational analysis, resulting in the creation of a material response database;"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d86cb4f3",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "* **Machine learning**, in which a surrogate model is trained to fit experimental findings;"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d522df59",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "* **Optimization**, where we try to iteratively improve the model to obtain a superior design."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9423d341",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<img src=./img/schematics/blocks.png width=100%>\n",
    "\n",
    "<br>\n",
    "\n",
    "<img src=./img/schematics/supercompressible_colored.png width=100%>\n",
    "\n",
    "<img src=./img/schematics/l2co_colored.png width=100%>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "879de459",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Challenge\n",
    "\n",
    "* Not all data-driven processes are the same ..\n",
    "* Conservative when it comes to **openly sharing databases and software**. \n",
    "* Numerous machine-learning packages.\n",
    "* ... we are not all software engineers"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4d12e510",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "Unified interface for the data-driven process so that researchers do not have to ‘reinvent the wheel’ for common procedures:\n",
    "\n",
    "* Parallelization\n",
    "* Data management\n",
    "* Comparing and benchmarking state-of-the-art methods "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4e960cc4",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "\n",
    "# ``f3dasm``: Framework for Data-Driven Design & Analysis of Structures & Materials\n",
    "\n",
    "https://github.com/bessagroup/f3dasm"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6ad4a60b",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## What's in the box?\n",
    "\n",
    "* Python open-source package hosted on [PyPI](https://pypi.org/project/f3dasm/):\n",
    "\n",
    "```\n",
    "pip install f3dasm\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "896bf820",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "✅ Compatible with the **three major operating systems** and **Python 3.7+**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "61fd0694",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "⭐ User-friendly **interfaces** for each of the stages in the data-driven process"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "48867a88",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "🏃‍♂️ Built-in **defaults** for getting started right away! "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7f3cacad",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "📈 Easy parallelization on **high-performance computing systems**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e3e1cc8c",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "💾 Automatic **data-management** of your experiments"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c70fccb9",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "### Documentation\n",
    "\n",
    "📖 Extensive [online documentation](https://bessagroup.github.io/f3dasm/)\n",
    "* Installation instructions\n",
    "* Getting started guide\n",
    "* In-depth explanation of each part of the data-driven process\n",
    "* Complete API documentation on every function and class\n",
    "\n",
    "<img src=./img/sphinx.png width=50%>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "357e294d",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Practical session\n",
    "\n",
    "\n",
    "<img src=./img/schematics/practical_session.png width=80%, align='center'>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "4dc859dd",
   "metadata": {
    "slideshow": {
     "slide_type": "notes"
    }
   },
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "904ae93c",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-10-09 13:07:41,396 - f3dasm - INFO - Imported f3dasm (version: 1.4.1)\n"
     ]
    }
   ],
   "source": [
    "import f3dasm"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2fe1ed27",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=./img/schematics/block_domain.png width=10%, align='right'>\n",
    "\n",
    "## 1. Design-of-Experiments\n",
    "\n",
    "\n",
    "* Parameters are singular features of the input search space. They are used to define the search space of the design.\n",
    "* The ``Domain`` is a set of ``Parameter`` instances that make up the feasible search space.\n",
    "\n",
    "<img src=./img/f3dasm-domain.png width=60%>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dd3679a4",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<img src=./img/schematics/block_domain.png width=10%, align='right'>\n",
    "\n",
    "We can import the ``Domain`` object from the `f3dasm.design` submodule"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "bc9c4ec8",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "from f3dasm.design import Domain"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a9d44786",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "We create a new `Domain` object:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "8b0046f9",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "domain = Domain()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9a5b4942",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "You can add various parameters with the ``add_float``, ``add_int``, ``add_category`` and ``add_constant`` methods:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "81b195dc",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Domain(space={'x0': ContinuousParameter(lower_bound=0.0, upper_bound=1.0, log=False), 'x1': ContinuousParameter(lower_bound=-5.0, upper_bound=5.0, log=False), 'x2': ContinuousParameter(lower_bound=2.0, upper_bound=3.0, log=False)})"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "domain.add_float(\"x0\", low=0.0, high=1.0)\n",
    "domain.add_float(\"x1\", low=-5.0, high=5.0)\n",
    "domain.add_float(\"x2\", low=2.0, high=3.0)\n",
    "\n",
    "domain"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6fdcf56f",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### The ExperimentData object\n",
    "\n",
    "Now that we have created a ``Domain``, we can construct the central piece of the data-driven process: the ``ExperimentData`` object"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "ba6f8e6b",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "from f3dasm import ExperimentData"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dc24e047",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "source": [
    "Creating an empty data object is easy:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "9524f350",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "experimentdata = ExperimentData(domain)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9764fdc1",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "# 🌊 Ride the `f3dasm` wave! 🌊\n",
    "\n",
    "* Once you create the `ExperimentData` object, you can invoke **data-driven processes** on your data!\n",
    "* However, once you **export** it to another datatype, you loose the connection!\n",
    "* If so, make sure to **recreate** the `ExperimentData` object\n",
    "\n",
    "<img src=./video/surfing_f3dasm.gif width=60%>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "478f26ed",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<img src=./img/schematics/block_sampling.png width=10%, align='right'>\n",
    "\n",
    "There are two ways of populating your object with input data.\n",
    "\n",
    "**Method 1**: One way of creating data is by sampling it with a sampler!"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fde173e2",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<img src=./img/schematics/block_sampling.png width=10%, align='right'>\n",
    "\n",
    "Implementing your sampling technique into ``f3dasm`` is easy; you have to provide a function that accepts:\n",
    "* a ``Domain`` object\n",
    "* the number of samples (``n_samples``)\n",
    "* a random seed (``seed``):"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "38009c44",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "To show an example; here is an implementation of a **grid-search** sampler with `numpy`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "93ecfb47",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "def my_sampler(domain: Domain, n_samples: int, seed: int) -> pd.DataFrame:\n",
    "    samples_per_parameter = np.round(n_samples**(1/len(domain)))\n",
    "    \n",
    "    # Create the mesh\n",
    "    mesh = np.meshgrid(*[np.linspace(parameter.lower_bound, parameter.upper_bound, int(samples_per_parameter)) \n",
    "                       for parameter in domain.values()])\n",
    "    return pd.DataFrame(np.array([m.flatten() for m in mesh]).T, columns=domain.names)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b2fc5178",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "We can now pass this to the ``sample`` method of the ``ExperimentData`` object:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "28bf0d5c",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "experimentdata.sample(sampler=my_sampler, n_samples=125, seed=2023)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e0eb2001",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<img src=./img/schematics/block_sampling.png width=10%, align='right'>\n",
    "\n",
    "Printing the ``ExperimentData`` object to a cell will show its content in a tabular matter:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "5ac68f29",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>jobs</th>\n",
       "      <th colspan=\"3\" halign=\"left\">input</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>x0</th>\n",
       "      <th>x1</th>\n",
       "      <th>x2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>open</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>2.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>open</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>2.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>open</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>2.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>open</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>2.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>open</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-5.0</td>\n",
       "      <td>3.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>120</th>\n",
       "      <td>open</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>121</th>\n",
       "      <td>open</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>122</th>\n",
       "      <td>open</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>123</th>\n",
       "      <td>open</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>124</th>\n",
       "      <td>open</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>125 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "<f3dasm._src.experimentdata.experimentdata.ExperimentData at 0x7f2bdaeda8b0>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "experimentdata"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "b0a31e0e",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAugAAALlCAYAAACWxBNOAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/OQEPoAAAACXBIWXMAAA9hAAAPYQGoP6dpAABozklEQVR4nO3dfXhU9Z338U8gySRAiBBIQBIBjYIWkIrIBlHxMgpV2YLWpeoK+ECrm3RFWpXgAz4UQys+sK4VtQLu3bJYvRv1FoqlSHDRWCpICxVxoyJETZQUk4GQkJBz/9GLKTEP5GFmzu+bvF/XlavNOWcmn88kv5NvxjNDjOd5ngAAAAA4oZvfAQAAAAD8AwM6AAAA4BAGdAAAAMAhDOgAAACAQxjQAQAAAIcwoAMAAAAOYUAHAAAAHMKADgAAADikyw3onuepsrJS/PtMgP9Yj4AbWIuAW7rcgB4MBpWcnKxgMOh3FKDLYz0CbmAtAm7pcgM6AAAA4DIGdAAAAMAhDOgAAACAQxjQAQAAAIf4OqC/+eabmjJlik488UTFxMTo5ZdfPu5tCgsLddZZZykQCCgzM1MrVqyIeE4AAAAgWmL9/OIHDx7UmWeeqRtuuEFXXHHFcY//5JNPdNlll+nmm2/Wr3/9a61fv1433XSTBg4cqEmTJkUhMYBoKtlfpWB1nSoP1So5MU69EmKV3qeH37HaxHoH6/kl+x1cyO9Cho6y3sF6fsl+h2jm93VA/853vqPvfOc7rT5+6dKlGjp0qB555BFJ0umnn65NmzbpscceY0AHOplPyw9qfsF2vVVcHto2ITNFC6eN1OCUnj4maz3rHaznl+x3cCG/Cxk6ynoH6/kl+x2ind/UNehFRUXKzs5usG3SpEkqKiryKRGASCjZX9XoRChJm4rLdVfBdpXsr/IpWetZ72A9v2S/gwv5XcjQUdY7WM8v2e/gR35fn0Fvq9LSUqWlpTXYlpaWpsrKSh06dEiJiYmNblNTU6OamprQ55WVla3+env27NG+ffvaH/g4+vXrp5NOOili9x/p/JL9DtbzS3//GQ8EAhG7/3A+Rq1dj8HqukYnwqM2FZcrWF0XljyRZL2D9fyS/Q6RzM9a/DsLHaznl+x38CO/qQG9PfLz83X//fe3+XZ79uzR8OGn69ChyP1Vl5jYQx98sDMiA2I08kv2O1jPL0mKiZEi+M9zh/Mxau16rDxU2+L+YHXL+11gvYP1/JL9DpHMz1r8B9c7WM8v2e/gR35TA/qAAQNUVlbWYFtZWZl69+7d5LPnkpSXl6e5c+eGPq+srFRGRsZxv9a+fft06FCVxt2wQL0HDulQ7qZUfrFbf1x2v/bt2xeR4TDS+SX7Haznl6Qvthdpx6vPaPQ1d6r/0OFhv/9wP0atXY+9E+NavJ+khJb3u8B6B+v5JfsdIpmftfgPrnewnl+y38GP/KYG9KysLK1Zs6bBtnXr1ikrK6vZ2wQCgQ795//eA4eo70nD2n17v1nPL9nvEMn8lV/sliT1Sj3JxGPU2vWYlBCrCZkp2tTEf1KckJmipAT3T13WO1jPL9nvEMn8rMW/s9DBen7Jfgc/8vv6ItEDBw5o27Zt2rZtm6S/v43itm3btGfPHkl//wt/xowZoeNvvvlmffzxx7rjjjv0wQcf6Be/+IV+85vf6LbbbvMjPoAISe/TQwunjdSEzJQG24++Yt7C23JZ72A9v2S/gwv5XcjQUdY7WM8v2e/gR35f/2R59913deGFF4Y+P/qf22bOnKkVK1boiy++CA3rkjR06FCtXr1at912m5YsWaL09HT98pe/5C0WgU5ocEpPLbpylILVdQpW1yopIU5Jxt4z13oH6/kl+x1cyO9Cho6y3sF6fsl+h2jn93VAnzhxorwWXtzW1L8SOnHiRL333nsRTAXAFVZO3C2x3sF6fsl+Bxfyu5Cho6x3sJ5fst8hmvlNvQ86AAAA0NkxoAMAAAAOYUAHAAAAHMKADgAAADiEAR0AAABwCAM6AAAA4BAGdAAAAMAhDOgAAACAQxjQAQAAAIcwoAMAAAAOYUAHAAAAHMKADgAAADiEAR0AAABwCAM6AAAA4BAGdAAAAMAhDOgAAACAQxjQAQAAAIcwoAMAAAAOYUAHAAAAHMKADgAAADiEAR0AAABwCAM6AAAA4BAGdAAAAMAhDOgAAACAQxjQAQAAAIcwoAMAAAAOYUAHAAAAHMKADgAAADiEAR0AAABwiO8D+pNPPqkhQ4YoISFB48aN0+bNm1s8/vHHH9ewYcOUmJiojIwM3Xbbbaquro5SWgAAACCyfB3QX3jhBc2dO1cLFizQ1q1bdeaZZ2rSpEn68ssvmzx+5cqVmjdvnhYsWKCdO3fqueee0wsvvKD58+dHOTkAAAAQGb4O6I8++qhmz56t66+/XmeccYaWLl2qHj16aNmyZU0e//bbb+vcc8/VNddcoyFDhuiSSy7R1Vdffdxn3QEAAAArfBvQDx8+rC1btig7O/sfYbp1U3Z2toqKipq8zfjx47Vly5bQQP7xxx9rzZo1uvTSS5v9OjU1NaqsrGzwAcAfrEfADaxFwG2+Dej79u3TkSNHlJaW1mB7WlqaSktLm7zNNddcowceeEATJkxQXFycTjnlFE2cOLHFS1zy8/OVnJwc+sjIyAhrDwCtx3oE3MBaBNzm+4tE26KwsFAPPfSQfvGLX2jr1q367W9/q9WrV+vBBx9s9jZ5eXmqqKgIfezduzeKiQEci/UIuIG1CLgt1q8v3K9fP3Xv3l1lZWUNtpeVlWnAgAFN3uaee+7Rddddp5tuukmSNHLkSB08eFA/+MEPdNddd6lbt8Z/bwQCAQUCgfAXANBmrEfADaxFwG2+PYMeHx+vMWPGaP369aFt9fX1Wr9+vbKyspq8TVVVVaMhvHv37pIkz/MiFxYAAACIEt+eQZekuXPnaubMmTr77LN1zjnn6PHHH9fBgwd1/fXXS5JmzJihQYMGKT8/X5I0ZcoUPfroo/r2t7+tcePGqbi4WPfcc4+mTJkSGtQBAAAAy3wd0KdPn66vvvpK9957r0pLSzV69GitXbs29MLRPXv2NHjG/O6771ZMTIzuvvtuffbZZ+rfv7+mTJmihQsX+lUBAAAACCtfB3RJys3NVW5ubpP7CgsLG3weGxurBQsWaMGCBVFIBsBvJfurFKyuU+WhWiUnxqlXQqzS+/TwO1abWO9gPb9kv4ML+V3I0FHWO1jPL9nvEM38vg/oANCUT8sPan7Bdr1VXB7aNiEzRQunjdTglJ4+Jms96x2s55fsd3AhvwsZOsp6B+v5Jfsdop3f1NssAugaSvZXNToRStKm4nLdVbBdJfurfErWetY7WM8v2e/gQn4XMnSU9Q7W80v2O/iRnwEdgHOC1XWNToRHbSouV7C6LsqJ2s56B+v5JfsdXMjvQoaOst7Ben7Jfgc/8jOgA3BO5aHaFvcHq1ve7wLrHaznl+x3cCG/Cxk6ynoH6/kl+x38yM+ADsA5vRPjWtyflNDyfhdY72A9v2S/gwv5XcjQUdY7WM8v2e/gR34GdADOSUqI1YTMlCb3TchMUVKC+69vt97Ben7JfgcX8ruQoaOsd7CeX7LfwY/8DOgAnJPep4cWThvZ6IR49BXzFt6Wy3oH6/kl+x1cyO9Cho6y3sF6fsl+Bz/yu/0nC4Aua3BKTy26cpSC1XUKVtcqKSFOScbeM9d6B+v5JfsdXMjvQoaOst7Ben7Jfodo52dAB+AsKyfulljvYD2/ZL+DC/ldyNBR1jtYzy/Z7xDN/FziAgAAADiEAR0AAABwCAM6AAAA4BAGdAAAAMAhDOgAAACAQxjQAQAAAIcwoAMAAAAOYUAHAAAAHMKADgAAADiEAR0AAABwCAM6AAAA4BAGdAAAAMAhDOgAAACAQxjQAQAAAIcwoAMAAAAOYUAHAAAAHMKADgAAADiEAR0AAABwCAM6AAAA4BAGdAAAAMAhvg/oTz75pIYMGaKEhASNGzdOmzdvbvH4r7/+Wjk5ORo4cKACgYBOO+00rVmzJkppAQAAgMiK9fOLv/DCC5o7d66WLl2qcePG6fHHH9ekSZO0a9cupaamNjr+8OHDuvjii5WamqqXXnpJgwYN0qeffqoTTjgh+uEBAACACPB1QH/00Uc1e/ZsXX/99ZKkpUuXavXq1Vq2bJnmzZvX6Phly5bpb3/7m95++23FxcVJkoYMGRLNyAAAAEBE+XaJy+HDh7VlyxZlZ2f/I0y3bsrOzlZRUVGTt3n11VeVlZWlnJwcpaWlacSIEXrooYd05MiRZr9OTU2NKisrG3wA8AfrEXADaxFwm28D+r59+3TkyBGlpaU12J6WlqbS0tImb/Pxxx/rpZde0pEjR7RmzRrdc889euSRR/TTn/602a+Tn5+v5OTk0EdGRkZYewBoPdYj4AbWIuA2318k2hb19fVKTU3VM888ozFjxmj69Om66667tHTp0mZvk5eXp4qKitDH3r17o5gYwLFYj4AbWIuA23y7Br1fv37q3r27ysrKGmwvKyvTgAEDmrzNwIEDFRcXp+7du4e2nX766SotLdXhw4cVHx/f6DaBQECBQCC84QG0C+sRcANrEXCbb8+gx8fHa8yYMVq/fn1oW319vdavX6+srKwmb3PuueequLhY9fX1oW0ffvihBg4c2ORwDgAAAFjj6yUuc+fO1bPPPqvnn39eO3fu1C233KKDBw+G3tVlxowZysvLCx1/yy236G9/+5tuvfVWffjhh1q9erUeeugh5eTk+FUBAAAACCtf32Zx+vTp+uqrr3TvvfeqtLRUo0eP1tq1a0MvHN2zZ4+6dfvH3xAZGRl6/fXXddttt2nUqFEaNGiQbr31Vt15551+VQAAAADCytcBXZJyc3OVm5vb5L7CwsJG27KysvTOO+9EOBUAAADgD1Pv4gIAAAB0dgzoAAAAgEMY0AEAAACHMKADAAAADmFABwAAABzCgA4AAAA4hAEdAAAAcEi7B/SamhrV1NSEMwsAAADQ5bVpQF+3bp0uvfRS9enTRz169FCPHj3Up08fXXrppfrDH/4QqYwAAABAl9Hqf0n0+eef10033aTvfe97euyxx5SWliZJKisr0+9//3tdeumleu6553TddddFLCyArqVkf5WC1XWqPFSr5MQ49UqIVXqfHn7HahPrHaznl+x3cCG/Cxk6ynoH6/kl+x2imb/VA/rChQv1+OOPKycnp9G+WbNmacKECXrggQcY0AGExaflBzW/YLveKi4PbZuQmaKF00ZqcEpPH5O1nvUO1vNL9ju4kN+FDB1lvYP1/JL9DtHO3+pLXPbs2aPs7Oxm91900UUqKSkJSygAXVvJ/qpGJ0JJ2lRcrrsKtqtkf5VPyVrPegfr+SX7HVzI70KGjrLewXp+yX4HP/K3ekD/1re+peeee67Z/cuWLdMZZ5wRllAAurZgdV2jE+FRm4rLFayui3KitrPewXp+yX4HF/K7kKGjrHewnl+y38GP/K2+xOWRRx7R5ZdfrrVr1yo7O7vBNejr16/Xxx9/rNWrV4c9IICup/JQbYv7g9Ut73eB9Q7W80v2O7iQ34UMHWW9g/X8kv0OfuRv9YA+ceJE7dixQ0uXLlVRUZFKS0slSQMGDNB3vvMd3XzzzRoyZEjYAwLoenonxrW4Pymh5f0usN7Ben7JfgcX8ruQoaOsd7CeX7LfwY/8bXqbxSFDhmjSpEnauHGjdu3apV27dmnjxo1atGiRhgwZoqeffjrsAQF0PUkJsZqQmdLkvgmZKUpKaPVzC76x3sF6fsl+Bxfyu5Cho6x3sJ5fst/Bj/xt/oeKJk+erNtvv121tf94On/fvn2aMmWK5s2bF9ZwALqm9D49tHDayEYnxKOvmLfwtlzWO1jPL9nv4EJ+FzJ0lPUO1vNL9jv4kb/NI/+GDRs0Y8YMrVu3TitXrtQnn3yiG2+8UcOGDdO2bdvCHhBA1zQ4pacWXTlKweo6BatrlZQQpyRj75lrvYP1/JL9Di7kdyFDR1nvYD2/ZL9DtPO3eUAfP368tm3bpptvvllnnXWW6uvr9eCDD+qOO+5QTExMJDIC6KKsnLhbYr2D9fyS/Q4u5HchQ0dZ72A9v2S/QzTzt/kSF0n68MMP9e677yo9PV2xsbHatWuXqqrcfg9LAAAAwII2D+iLFi1SVlaWLr74Yu3YsUObN2/We++9p1GjRqmoqCgSGQEAAIAuo80D+pIlS/Tyyy/riSeeUEJCgkaMGKHNmzfriiuu0MSJEyMQEQAAAOg62nwN+vbt29WvX78G2+Li4vTwww/r8ssvD1swAAAAoCtq8zPo3xzOj3XBBRd0KAwAAADQ1bXrRaIAAAAAIoMBHQAAAHAIAzoAAADgEAZ0AAAAwCFODOhPPvmkhgwZooSEBI0bN06bN29u1e1WrVqlmJgYTZ06NbIBAQAAgCjxfUB/4YUXNHfuXC1YsEBbt27VmWeeqUmTJunLL79s8Xa7d+/WT37yE5133nlRSgoAAABEnu8D+qOPPqrZs2fr+uuv1xlnnKGlS5eqR48eWrZsWbO3OXLkiK699lrdf//9Ovnkk6OYFgAAAIgsXwf0w4cPa8uWLcrOzg5t69atm7Kzs1VUVNTs7R544AGlpqbqxhtvjEZMAAAAIGra/C+JhtO+fft05MgRpaWlNdielpamDz74oMnbbNq0Sc8995y2bdvWqq9RU1Ojmpqa0OeVlZXtzgugY1iPgBtYi4DbfL/EpS2CwaCuu+46Pfvssy3+i6bHys/PV3JycugjIyMjwikBNIf1CLiBtQi4zdcBvV+/furevbvKysoabC8rK9OAAQMaHf/RRx9p9+7dmjJlimJjYxUbG6v/+q//0quvvqrY2Fh99NFHjW6Tl5enioqK0MfevXsj1gdAy1iPgBtYi4DbfL3EJT4+XmPGjNH69etDb5VYX1+v9evXKzc3t9Hxw4cP1/bt2xtsu/vuuxUMBrVkyZImnwEIBAIKBAIRyQ+gbViPgBtYi4DbfB3QJWnu3LmaOXOmzj77bJ1zzjl6/PHHdfDgQV1//fWSpBkzZmjQoEHKz89XQkKCRowY0eD2J5xwgiQ12g4AAABY5PuAPn36dH311Ve69957VVpaqtGjR2vt2rWhF47u2bNH3bqZulQeAAAAaDffB3RJys3NbfKSFkkqLCxs8bYrVqwIfyAAAADAJzw1DQAAADiEAR0AAABwCAM6AAAA4BAGdAAAAMAhDOgAAACAQxjQAQAAAIcwoAMAAAAOYUAHAAAAHMKADgAAADiEAR0AAABwCAM6AAAA4BAGdAAAAMAhDOgAAACAQxjQAQAAAIcwoAMAAAAOYUAHAAAAHMKADgAAADiEAR0AAABwCAM6AAAA4JBYvwMAQHNK9lcpWF2nykO1Sk6MU6+EWKX36eF3rDax3sF6fsl+Bxfyu5Cho6x3sJ5fst8hmvkZ0AE46dPyg5pfsF1vFZeHtk3ITNHCaSM1OKWnj8laz3oH6/kl+x1cyO9Cho6y3sF6fsl+h2jn5xIXAM4p2V/V6EQoSZuKy3VXwXaV7K/yKVnrWe9gPb9kv4ML+V3I0FHWO1jPL9nv4Ed+BnQAzglW1zU6ER61qbhcweq6KCdqO+sdrOeX7HdwIb8LGTrKegfr+SX7HfzIz4AOwDmVh2pb3B+sbnm/C6x3sJ5fst/BhfwuZOgo6x2s55fsd/AjPwM6AOf0ToxrcX9SQsv7XWC9g/X8kv0OLuR3IUNHWe9gPb9kv4Mf+RnQATgnKSFWEzJTmtw3ITNFSQnuv77degfr+SX7HVzI70KGjrLewXp+yX4HP/IzoANwTnqfHlo4bWSjE+LRV8xbeFsu6x2s55fsd3AhvwsZOsp6B+v5Jfsd/Mjv9p8sALqswSk9tejKUQpW1ylYXaukhDglGXvPXOsdrOeX7HdwIb8LGTrKegfr+SX7HaKdnwEdgLOsnLhbYr2D9fyS/Q4u5HchQ0dZ72A9v2S/QzTzO3GJy5NPPqkhQ4YoISFB48aN0+bNm5s99tlnn9V5552nPn36qE+fPsrOzm7xeAAAAMAS3wf0F154QXPnztWCBQu0detWnXnmmZo0aZK+/PLLJo8vLCzU1VdfrQ0bNqioqEgZGRm65JJL9Nlnn0U5OQAAABB+vg/ojz76qGbPnq3rr79eZ5xxhpYuXaoePXpo2bJlTR7/61//Wv/2b/+m0aNHa/jw4frlL3+p+vp6rV+/PsrJAQAAgPDz9Rr0w4cPa8uWLcrLywtt69atm7Kzs1VUVNSq+6iqqlJtba369u3b5P6amhrV1NSEPq+srOxYaADtxnoE3MBaBNzm6zPo+/bt05EjR5SWltZge1pamkpLS1t1H3feeadOPPFEZWdnN7k/Pz9fycnJoY+MjIwO5wbQPqxHwA2sRcBtvl/i0hGLFi3SqlWrVFBQoISEhCaPycvLU0VFRehj7969UU4J4CjWI+AG1iLgNl8vcenXr5+6d++usrKyBtvLyso0YMCAFm+7ePFiLVq0SH/4wx80atSoZo8LBAIKBAJhyQugY1iPgBtYi4DbfH0GPT4+XmPGjGnwAs+jL/jMyspq9nY///nP9eCDD2rt2rU6++yzoxEVAAAAiArf/6GiuXPnaubMmTr77LN1zjnn6PHHH9fBgwd1/fXXS5JmzJihQYMGKT8/X5L0s5/9TPfee69WrlypIUOGhK5V79Wrl3r16uVbDwAAACAcfB/Qp0+frq+++kr33nuvSktLNXr0aK1duzb0wtE9e/aoW7d/PNH/1FNP6fDhw/re977X4H4WLFig++67L5rRAQAAgLDzfUCXpNzcXOXm5ja5r7CwsMHnu3fvjnwgAAAAwCem38UFAAAA6GwY0AEAAACHMKADAAAADmFABwAAABzCgA4AAAA4hAEdAAAAcAgDOgAAAOAQBnQAAADAIQzoAAAAgEMY0AEAAACHMKADAAAADmFABwAAABzCgA4AAAA4hAEdAAAAcAgDOgAAAOAQBnQAAADAIQzoAAAAgEMY0AEAAACHMKADAAAADmFABwAAABzCgA4AAAA4hAEdAAAAcAgDOgAAAOAQBnQAAADAIQzoAAAAgEMY0AEAAACHMKADAAAADmFABwAAABzCgA4AAAA4JNbvAJL05JNP6uGHH1ZpaanOPPNMPfHEEzrnnHOaPf7FF1/UPffco927d+vUU0/Vz372M1166aVRTAwgGkr2VylYXafKQ7VKToxTr4RYpffp4XesNrHewXp+yX4HF/K7kKGjrHewnl+y3yGa+X0f0F944QXNnTtXS5cu1bhx4/T4449r0qRJ2rVrl1JTUxsd//bbb+vqq69Wfn6+Lr/8cq1cuVJTp07V1q1bNWLECB8aAIiET8sPan7Bdr1VXB7aNiEzRQunjdTglJ4+Jms96x2s55fsd3AhvwsZOsp6B+v5Jfsdop3f90tcHn30Uc2ePVvXX3+9zjjjDC1dulQ9evTQsmXLmjx+yZIlmjx5sm6//XadfvrpevDBB3XWWWfpP//zP6OcHECklOyvanQilKRNxeW6q2C7SvZX+ZSs9ax3sJ5fst/BhfwuZOgo6x2s55fsd/Ajv68D+uHDh7VlyxZlZ2eHtnXr1k3Z2dkqKipq8jZFRUUNjpekSZMmNXt8TU2NKisrG3wA8Edr12Owuq7RifCoTcXlClbXRTJmWFjvYD2/ZL9DJPOzFv/OQgfr+SX7HfzI7+uAvm/fPh05ckRpaWkNtqelpam0tLTJ25SWlrbp+Pz8fCUnJ4c+MjIywhMeQJu1dj1WHqpt8X6C1S3vd4H1DtbzS/Y7RDI/a/EfXO9gPb9kv4Mf+X2/xCXS8vLyVFFREfrYu3ev35GALqu167F3YlyL95OU0PJ+F1jvYD2/ZL9DJPOzFv/B9Q7W80v2O/iR39cBvV+/furevbvKysoabC8rK9OAAQOavM2AAQPadHwgEFDv3r0bfADwR2vXY1JCrCZkpjS5b0JmipISfH99+3FZ72A9v2S/QyTzsxb/zkIH6/kl+x38yO/rgB4fH68xY8Zo/fr1oW319fVav369srKymrxNVlZWg+Mlad26dc0eD8Ce9D49tHDayEYnxKOvmLfwtlzWO1jPL9nv4EJ+FzJ0lPUO1vNL9jv4kd/3P1nmzp2rmTNn6uyzz9Y555yjxx9/XAcPHtT1118vSZoxY4YGDRqk/Px8SdKtt96qCy64QI888oguu+wyrVq1Su+++66eeeYZP2sACLPBKT216MpRClbXKVhdq6SEOCUZe89c6x2s55fsd3AhvwsZOsp6B+v5Jfsdop3f9wF9+vTp+uqrr3TvvfeqtLRUo0eP1tq1a0MvBN2zZ4+6dfvHE/3jx4/XypUrdffdd2v+/Pk69dRT9fLLL/Me6EAnZOXE3RLrHaznl+x3cCG/Cxk6ynoH6/kl+x2imd/3AV2ScnNzlZub2+S+wsLCRtuuuuoqXXXVVRFOBQAAAERfp38XFwAAAMASBnQAAADAIU5c4hJNnudJ0nH/RdEDBw5IkipKPlZ93ZGw5wiWfipJ2rp1a+hrhdOHH34oKXL5JfsdrOeXpANffiZJqizdrUAgEPb7P/oYHThw4LhrJikpSTExMW26/9auRwBt09b1yFoEIqM9vxslKcY7uiq7iJKSEv41USACKioq2vzvDLAegcho63pkLQKR0Z7fjVIXHNDr6+v1+eefH/cvmsrKSmVkZGjv3r0m/3Ej6/kl+x2s55fa1qE9zxK0Zj12tcfRRdbzS/Y7tDV/W9cja9EG6/kl+x0ivRaP6nKXuHTr1k3p6emtPt76vz5qPb9kv4P1/FLkOrRlPfI4+s96fsl+B9ZieFjvYD2/ZL9DpPPzIlEAAADAIQzoAAAAgEMY0JsRCAS0YMGCiLwzRjRYzy/Z72A9v+RGBxcydJT1DtbzS/Y7uJDfhQwdZb2D9fyS/Q7Ryt/lXiQKAAAAuIxn0AEAAACHMKADAAAADmFABwAAABzCgA4AAAA4pMsN6J7nqbKyUrw2FvAf6xFwA2sRcEuXG9CDwaCSk5MVDAb9jgJ0eaxHwA2sRcAtXW5ABwAAAFzGgA4AAAA4hAEdAAAAcAgDOgAAAOAQBnQAAADAIbF+B3BRyf4qBavrVHmoVsmJceqVEKv0Pj38jtVq1vNL9jtYzy91jg4AwqMznA+sd7CeX7LfIZr5GdC/4dPyg5pfsF1vFZeHtk3ITNHCaSM1OKWnj8lax3p+yX4H6/mlztEBQHh0hvOB9Q7W80v2O0Q7P5e4HKNkf1WjB1+SNhWX666C7SrZX+VTstaxnl+y38F6fqlzdAAQHp3hfGC9g/X8kv0OfuTnGfRjBKvrGj34R20qLlewui7KidrGen7Jfgfr+SU3O+zZs0f79u2L2P3369dPJ510UsTuX7LfwXp+KfIdampqFAgEInb/0XiMvsnF80FbWe9gPb9kv4Mf+RnQj1F5qLbF/cHqlvf7zXp+yX4H6/kl9zrs2bNHw4efrkOHIvcMS2JiD33wwc6IDT/WO1jPL0Wng2JiJM+L2N1H+jFqimvng/aw3sF6fsl+Bz/yM6Afo3diXIv7kxJa3u836/kl+x2s55fc67Bv3z4dOlSlcTcsUO+BQ8J+/5Vf7NYfl92vffv2RWzwsd7Ben4p8h2+2F6kHa8+o9HX3Kn+Q4eH/f6j8Rg1xbXzQXtY72A9v2S/gx/5GdCPkZQQqwmZKdrUxH/GmJCZoqQEtx8u6/kl+x2s55fc7dB74BD1PWmYL187XKx3sJ5filyHyi92S5J6pZ5k/jE6lqvng7aw3sF6fsl+Bz/y8yLRY6T36aGF00ZqQmZKg+1HX6Xr+lsBWc8v2e9gPb/UOToACI/OcD6w3sF6fsl+Bz/yu/0niw8Gp/TUoitHKVhdp2B1rZIS4pRk6H06reeX7Hewnl/qHB0AhEdnOB9Y72A9v2S/Q7TzM6A3wcoPS3Os55fsd7CeX+ocHQCER2c4H1jvYD2/ZL9DNPNziQsAAADgEAZ0AAAAwCEM6AAAAIBDGNABAAAAhzCgAwAAAA5hQAcAAAAcwoAOAAAAOIQBHQAAAHAIAzoAAADgENMD+qJFixQTE6M5c+b4HQUAAAAIC7MD+p/+9Cc9/fTTGjVqlN9RAAAAgLAxOaAfOHBA1157rZ599ln16dPH7zgAAABA2Jgc0HNycnTZZZcpOzvb7ygAAABAWMX6HaCtVq1apa1bt+pPf/pTq46vqalRTU1N6PPKyspIRQNwHKxHwA2sRcBtpp5B37t3r2699Vb9+te/VkJCQqtuk5+fr+Tk5NBHRkZGhFMCaA7rEXADaxFwm6kBfcuWLfryyy911llnKTY2VrGxsdq4caP+4z/+Q7GxsTpy5Eij2+Tl5amioiL0sXfvXh+SA5BYj4ArWIuA20xd4nLRRRdp+/btDbZdf/31Gj58uO68805179690W0CgYACgUC0IgJoAesRcANrEXCbqQE9KSlJI0aMaLCtZ8+eSklJabQdAAAAsMjUJS4AAABAZ2fqGfSmFBYW+h0BAAAACBueQQcAAAAcwoAOAAAAOIQBHQAAAHAIAzoAAADgEAZ0AAAAwCEM6AAAAIBDGNABAAAAhzCgAwAAAA5hQAcAAAAcwoAOAAAAOIQBHQAAAHAIAzoAAADgEAZ0AAAAwCEM6AAAAIBDGNABAAAAhzCgAwAAAA5hQAcAAAAcwoAOAAAAOIQBHQAAAHBIrN8BXFSyv0rB6jpVHqpVcmKceiXEKr1PD79jtZr1/JL9DtbzS52jA4Dw6AznA+sdrOeX7HeIZn4G9G/4tPyg5hds11vF5aFtEzJTtHDaSA1O6eljstaxnl+y38F6fqlzdAAQHp3hfGC9g/X8kv0O0c7PJS7HKNlf1ejBl6RNxeW6q2C7SvZX+ZSsdaznl+x3sJ5f6hwdAIRHZzgfWO9gPb9kv4Mf+RnQjxGsrmv04B+1qbhcweq6KCdqG+v5JfsdrOeXOkcHAOHRGc4H1jtYzy/Z7+BHfgb0Y1Qeqm1xf7C65f1+s55fst/Ben6pc3QAEB6d4XxgvYP1/JL9Dn7kZ0A/Ru/EuBb3JyW0vN9v1vNL9jtYzy91jg4AwqMznA+sd7CeX7LfwY/8DOjHSEqI1YTMlCb3TchMUVKC26+ptZ5fst/Ben6pc3QAEB6d4XxgvYP1/JL9Dn7kZ0A/RnqfHlo4bWSjb8LRV+m6/lZA1vNL9jtYzy91jg4AwqMznA+sd7CeX7LfwY/8bv/J4oPBKT216MpRClbXKVhdq6SEOCUZep9O6/kl+x2s55c6RwcA4dEZzgfWO1jPL9nvEO38DOhNsPLD0hzr+SX7HaznlzpHBwDh0RnOB9Y7WM8v2e8QzfzmLnHJz8/X2LFjlZSUpNTUVE2dOlW7du3yOxYAAAAQFuYG9I0bNyonJ0fvvPOO1q1bp9raWl1yySU6ePCg39EAAACADjN3icvatWsbfL5ixQqlpqZqy5YtOv/8831KBQAAAISHuQH9myoqKiRJffv2bXJ/TU2NampqQp9XVlZGJReAxliPgBtYi4DbzF3icqz6+nrNmTNH5557rkaMGNHkMfn5+UpOTg59ZGRkRDklgKNYj4AbWIuA20wP6Dk5OdqxY4dWrVrV7DF5eXmqqKgIfezduzeKCQEci/UIuIG1CLjN7CUuubm5eu211/Tmm28qPT292eMCgYACgUAUkwFoDusRcANrEXCbuQHd8zz96Ec/UkFBgQoLCzV06FC/IwEAAABhY25Az8nJ0cqVK/XKK68oKSlJpaWlkqTk5GQlJib6nA4AAADoGHPXoD/11FOqqKjQxIkTNXDgwNDHCy+84Hc0AAAAoMPMPYPueZ7fEQAAAICIMfcMOgAAANCZMaADAAAADmFABwAAABzCgA4AAAA4hAEdAAAAcAgDOgAAAOAQBnQAAADAIQzoAAAAgEMY0AEAAACHMKADAAAADmFABwAAABzCgA4AAAA4hAEdAAAAcAgDOgAAAOAQBnQAAADAIQzoAAAAgEMY0AEAAACHMKADAAAADmFABwAAABzCgA4AAAA4hAEdAAAAcAgDOgAAAOAQBnQAAADAIQzoAAAAgEMY0AEAAACHMKADAAAADmFABwAAABwS9gF9586dOvnkk8N9twAAAECXEPYB/fDhw/r000/DfbeNPPnkkxoyZIgSEhI0btw4bd68OeJfEwAAAIi02LbeYO7cuS3u/+qrr9odprVeeOEFzZ07V0uXLtW4ceP0+OOPa9KkSdq1a5dSU1M7fP8l+6sUrK5T5aFaJSfGqVdCrNL79AhD8uiwnl+y38F6fqlzdAAQHp3hfGC9g/X8kv0O0czf5gF9yZIlGj16tHr37t3k/gMHDnQ41PE8+uijmj17tq6//npJ0tKlS7V69WotW7ZM8+bN69B9f1p+UPMLtuut4vLQtgmZKVo4baQGp/Ts0H1Hg/X8kv0O1vNLnaMDgPDoDOcD6x2s55fsd4h2/jZf4pKZmanbbrtNGzZsaPLj2WefDXvIYx0+fFhbtmxRdnZ2aFu3bt2UnZ2toqKiDt13yf6qRg++JG0qLtddBdtVsr+qQ/cfadbzS/Y7WM8vdY4OAMKjM5wPrHewnl+y38GP/G0e0M8++2xt2bKl2f0xMTHyPK9DoVqyb98+HTlyRGlpaQ22p6WlqbS0tNHxNTU1qqysbPDRnGB1XaMH/6hNxeUKVtd1LHyEWc8v2e9gPb8U2Q5tWY8AIqe1a5Fzmv+s55fsd/Ajf5sH9EceeURz5sxpdv+ZZ56p+vr6jmQKq/z8fCUnJ4c+MjIymj228lBti/cVrG55v9+s55fsd7CeX4psh7asRwCR09q1yDnNf9bzS/Y7+JG/zQP6gAEDNHjwYG3YsKHZY55++ukOhWpJv3791L17d5WVlTXYXlZWpgEDBjQ6Pi8vTxUVFaGPvXv3NnvfvRPjWvzaSQkt7/eb9fyS/Q7W80uR7dCW9Qggclq7Fjmn+c96fsl+Bz/yt/ttFidPnqzbb79dtbX/+Kth3759mjJlSodfqNmS+Ph4jRkzRuvXrw9tq6+v1/r165WVldXo+EAgoN69ezf4aE5SQqwmZKY0uW9CZoqSEtr8mtqosp5fst/Ben4psh3ash4BRE5r1yLnNP9Zzy/Z7+BH/nYP6Bs2bFBBQYHGjh2r999/X6tXr9aIESNUWVmpbdu2hTFiY3PnztWzzz6r559/Xjt37tQtt9yigwcPht7Vpb3S+/TQwmkjG30Tjr5K1/W3ArKeX7LfwXp+qXN0ABAeneF8YL2D9fyS/Q5+5G/3yD9+/Hht27ZNN998s8466yzV19frwQcf1B133KGYmJhwZmxk+vTp+uqrr3TvvfeqtLRUo0eP1tq1axu9cLQ9Bqf01KIrRylYXadgda2SEuKUZOh9Oq3nl+x3sJ5f6hwdAIRHZzgfWO9gPb9kv0O083foOfkPP/xQ7777rtLT0/X5559r165dqqqqUs+ekX8/y9zcXOXm5kbkvq38sDTHen7Jfgfr+aXO0QFAeHSG84H1DtbzS/Y7RDN/uy9xWbRokbKysnTxxRdrx44d2rx5s9577z2NGjWqw+9HDgAAAHRV7R7QlyxZopdffllPPPGEEhISNGLECG3evFlXXHGFJk6cGMaIAAAAQNfR7ktctm/frn79+jXYFhcXp4cffliXX355h4MBAAAAXVG7n0H/5nB+rAsuuKC9dwsAAAB0ae0e0AEAAACEHwM6AAAA4BAGdAAAAMAhDOgAAACAQxjQAQAAAIcwoAMAAAAOYUAHAAAAHMKADgAAADiEAR0AAABwCAM6AAAA4BAGdAAAAMAhDOgAAACAQxjQAQAAAIcwoAMAAAAOYUAHAAAAHMKADgAAADiEAR0AAABwCAM6AAAA4BAGdAAAAMAhDOgAAACAQxjQAQAAAIcwoAMAAAAOYUAHAAAAHMKADgAAADiEAR0AAABwiKkBfffu3brxxhs1dOhQJSYm6pRTTtGCBQt0+PBhv6MBAAAAYRHrd4C2+OCDD1RfX6+nn35amZmZ2rFjh2bPnq2DBw9q8eLFfscDAAAAOszUgD558mRNnjw59PnJJ5+sXbt26amnnmJABwAAQKdgakBvSkVFhfr27dvs/pqaGtXU1IQ+r6ysjEYsAE1gPQJuYC0CbjN1Dfo3FRcX64knntAPf/jDZo/Jz89XcnJy6CMjIyOKCQEci/UIuIG1CLjNiQF93rx5iomJafHjgw8+aHCbzz77TJMnT9ZVV12l2bNnN3vfeXl5qqioCH3s3bs30nUANIP1CLiBtQi4zYlLXH784x9r1qxZLR5z8sknh/7/559/rgsvvFDjx4/XM8880+LtAoGAAoFAOGIC6CDWI+AG1iLgNicG9P79+6t///6tOvazzz7ThRdeqDFjxmj58uXq1s2J/wgAAAAAhIUTA3prffbZZ5o4caIGDx6sxYsX66uvvgrtGzBggI/JAAAAgPAwNaCvW7dOxcXFKi4uVnp6eoN9nuf5lAoAAAAIH1MD+qxZs457rXo4lOyvUrC6TpWHapWcGKdeCbFK79Mj4l83XKznl+x3sJ5f6hwdAIRHZzgfWO9gPb9kv0M085sa0KPh0/KDml+wXW8Vl4e2TchM0cJpIzU4paePyVrHen7Jfgfr+aXO0QFAeHSG84H1DtbzS/Y7RDs/r7A8Rsn+qkYPviRtKi7XXQXbVbK/yqdkrWM9v2S/g/X8UufoACA8OsP5wHoH6/kl+x38yM+AfoxgdV2jB/+oTcXlClbXRTlR21jPL9nvYD2/1Dk6AAiPznA+sN7Ben7Jfgc/8jOgH6PyUG2L+4PVLe/3m/X8kv0O1vNLnaMDgPDoDOcD6x2s55fsd/AjPwP6MXonxrW4Pymh5f1+s55fst/Ben6pc3QAEB6d4XxgvYP1/JL9Dn7kZ0A/RlJCrCZkpjS5b0JmipIS3H5NrfX8kv0O1vNLnaMDgPDoDOcD6x2s55fsd/AjPwP6MdL79NDCaSMbfROOvkrX9bcCsp5fst/Ben6pc3QAEB6d4XxgvYP1/JL9Dn7kd/tPFh8MTumpRVeOUrC6TsHqWiUlxCnJ0Pt0Ws8v2e9gPb/UOToACI/OcD6w3sF6fsl+h2jnZ0BvgpUfluZYzy/Z72A9v9Q5OgAIj85wPrDewXp+yX6HaObnEhcAAADAIQzoAAAAgEMY0AEAAACHMKADAAAADmFABwAAABzCgA4AAAA4hAEdAAAAcAgDOgAAAOAQBnQAAADAIQzoAAAAgEMY0AEAAACHMKADAAAADmFABwAAABzCgA4AAAA4hAEdAAAAcAgDOgAAAOAQBnQAAADAIQzoAAAAgEMY0AEAAACHMKADAAAADjE7oNfU1Gj06NGKiYnRtm3b/I4DAAAAhIXZAf2OO+7QiSee6HcMAAAAIKxMDui/+93v9Pvf/16LFy/2OwoAAAAQVrF+B2irsrIyzZ49Wy+//LJ69Ohx3ONrampUU1MT+ryysjKS8QC0gPUIuIG1CLjN1DPonudp1qxZuvnmm3X22We36jb5+flKTk4OfWRkZEQ4JYDmsB4BN7AWAbc5MaDPmzdPMTExLX588MEHeuKJJxQMBpWXl9fq+87Ly1NFRUXoY+/evRFsAqAlrEfADaxFwG1OXOLy4x//WLNmzWrxmJNPPllvvPGGioqKFAgEGuw7++yzde211+r5559vdLtAINDoeAD+YD0CbmAtAm5zYkDv37+/+vfvf9zj/uM//kM//elPQ59//vnnmjRpkl544QWNGzcukhEBAACAqHBiQG+tk046qcHnvXr1kiSdcsopSk9P9yMSAAAAEFZOXIMOAAAA4O9MPYP+TUOGDJHneX7HAAAAAMKGZ9ABAAAAhzCgAwAAAA5hQAcAAAAcwoAOAAAAOIQBHQAAAHAIAzoAAADgEAZ0AAAAwCEM6AAAAIBDGNABAAAAh5j+l0QjpWR/lYLVdao8VKvkxDj1SohVep8efsdqNev5JfsdrOeXOkcHAOHRGc4H1jtYzy/Z7xDN/Azo3/Bp+UHNL9iut4rLQ9smZKZo4bSRGpzS08dkrWM9v2S/g/X8UufoACA8OsP5wHoH6/kl+x2inZ9LXI5Rsr+q0YMvSZuKy3VXwXaV7K/yKVnrWM8v2e9gPb/UOToACI/OcD6w3sF6fsl+Bz/yM6AfI1hd1+jBP2pTcbmC1XVRTtQ21vNL9jtYzy91jg4AwqMznA+sd7CeX7LfwY/8DOjHqDxU2+L+YHXL+/1mPb9kv4P1/FLn6AAgPDrD+cB6B+v5Jfsd/MjPgH6M3olxLe5PSmh5v9+s55fsd7CeX+ocHQCER2c4H1jvYD2/ZL+DH/kZ0I+RlBCrCZkpTe6bkJmipAS3X1NrPb9kv4P1/FLn6AAgPDrD+cB6B+v5Jfsd/MjPgH6M9D49tHDayEbfhKOv0nX9rYCs55fsd7CeX+ocHQCER2c4H1jvYD2/ZL+DH/nd/pPFB4NTemrRlaMUrK5TsLpWSQlxSjL0Pp3W80v2O1jPL3WODgDCozOcD6x3sJ5fst8h2vkZ0Jtg5YelOdbzS/Y7WM8vdY4OAMKjM5wPrHewnl+y3yGa+bnEBQAAAHAIAzoAAADgEAZ0AAAAwCFd7hp0z/MkSZWVlT4nATqXpKQkxcTEtOk2rVmPBw4ckCRVlHys+roj7Q/YjGDpp5KkrVu3hr5WuH344YeS7Hawnl+KfIcDX34mSaos3a1AIBD2+z/6GB04cKBVv7/auh753QhERnt+N0pSjHd0VXYRJSUlysjI8DsG0OlUVFSod+/ebboN6xGIjLauR9YiEBnt+d0odcEBvb6+Xp9//vlx/6KprKxURkaG9u7d264H1m/W80v2O1jPL7WtQ3ueJWjNeuxqj6OLrOeX7Hdoa/62rkfWog3W80v2O0R6LR7V5S5x6datm9LT01t9fO/evU3+AB1lPb9kv4P1/FLkOrRlPfI4+s96fsl+B9ZieFjvYD2/ZL9DpPPzIlEAAADAIQzoAAAAgEMY0JsRCAS0YMGCiLwaPxqs55fsd7CeX3KjgwsZOsp6B+v5JfsdXMjvQoaOst7Ben7Jfodo5e9yLxIFAAAAXMYz6AAAAIBDGNABAAAAhzCgAwAAAA5hQAcAAAAc0uUGdM/zVFlZKV4bC/iP9Qi4gbUIuKXLDejBYFDJyckKBoN+RwG6PNYj4AbWIuCWLjegAwAAAC5jQAcAAAAcwoAOAAAAOIQBHQAAAHCIrwP6U089pVGjRql3797q3bu3srKy9Lvf/a7F27z44osaPny4EhISNHLkSK1ZsyZKaQEAAIDIi/Xzi6enp2vRokU69dRT5Xmenn/+eX33u9/Ve++9p29961uNjn/77bd19dVXKz8/X5dffrlWrlypqVOnauvWrRoxYkTYcpXsr1Kwuk6Vh2qVnBinXgmxSu/TI2z3H2nW80v2O1jPL7nRwYUMHWW9g/X8kv0O1vMDaLsYz7E3Pe3bt68efvhh3XjjjY32TZ8+XQcPHtRrr70W2vZP//RPGj16tJYuXdqq+6+srFRycrIqKirUu3fvRvs/LT+o+QXb9VZxeWjbhMwULZw2UoNTerajUXRZzy/Z72A9vxS9Di2tRx5H/1nPL9nv4MJaBBB9zlyDfuTIEa1atUoHDx5UVlZWk8cUFRUpOzu7wbZJkyapqKgoLBlK9lc1OhFK0qbict1VsF0l+6vC8nUixXp+yX4H6/klNzq4kKGjrHewnl+y38F6fgDt5+slLpK0fft2ZWVlqbq6Wr169VJBQYHOOOOMJo8tLS1VWlpag21paWkqLS1t9v5rampUU1MT+ryysrLZY4PVdY1OhEdtKi5XsLqupSq+s55fst/Ben4psh1aux55HP1nPb9kv4MLa/Gb9uzZo3379rX767ZGv379dNJJJ0Xs/iPdwXp+yX6HmpoaBQKBiN1/pB8fyYEBfdiwYdq2bZsqKir00ksvaebMmdq4cWOzQ3pb5efn6/7772/VsZWHalvcH6xueb/frOeX7Hewnl+KbIfWrkceR/9Zzy/Z7+DCWjzWnj17NHz46Tp0KLLP3Ccm9tAHH+yMyAAUjQ7W80udoENMjBTBK7gj+fgc5fuAHh8fr8zMTEnSmDFj9Kc//UlLlizR008/3ejYAQMGqKysrMG2srIyDRgwoNn7z8vL09y5c0OfV1ZWKiMjo8ljeyfGtZg1KaHl/X6znl+y38F6fimyHVq7Hnkc/Wc9v2S/gwtr8Vj79u3ToUNVGnfDAvUeOKTdX7sllV/s1h+X3a99+/ZFZPiJdAfr+SX7Hb7YXqQdrz6j0dfcqf5Dh4f9/iP9+Bzl+4D+TfX19Q3+s9uxsrKytH79es2ZMye0bd26dc1esy5JgUCg1f+ZIykhVhMyU7Spif+kOCEzRUkJzj1cDVjPL9nvYD2/FNkOrV2PPI7+s55fst/BhbXYlN4Dh6jvScPa/bVdYL2D9fxS5DpUfrFbktQr9STTj5GvLxLNy8vTm2++qd27d2v79u3Ky8tTYWGhrr32WknSjBkzlJeXFzr+1ltv1dq1a/XII4/ogw8+0H333ad3331Xubm5YcmT3qeHFk4bqQmZKQ22H33FvOtva2U9v2S/g/X8khsdXMjQUdY7WM8v2e9gPT+A9vP16YMvv/xSM2bM0BdffKHk5GSNGjVKr7/+ui6++GJJf79OqVu3f/wNMX78eK1cuVJ333235s+fr1NPPVUvv/xyWN8DfXBKTy26cpSC1XUKVtcqKSFOSYbec9Z6fsl+B+v5JTc6uJCho6x3sJ5fst/Ben4A7ePrgP7cc8+1uL+wsLDRtquuukpXXXVVhBL9nfUTn/X8kv0O1vNLbnRwIUNHWe9gPb9kv4P1/ADazpn3QQcAAADAgA4AAAA4hQEdAAAAcAgDOgAAAOAQBnQAAADAIQzoAAAAgEMY0AEAAACHMKADAAAADmFABwAAABzCgA4AAAA4hAEdAAAAcAgDOgAAAOAQBnQAAADAIQzoAAAAgEMY0AEAAACHMKADAAAADmFABwAAABzCgA4AAAA4hAEdAAAAcAgDOgAAAOAQBnQAAADAIQzoAAAAgEMY0AEAAACHMKADAAAADmFABwAAABzCgA4AAAA4hAEdAAAAcAgDOgAAAOAQBnQAAADAIb4O6Pn5+Ro7dqySkpKUmpqqqVOnateuXce93eOPP65hw4YpMTFRGRkZuu2221RdXR2FxAAAAEBk+Tqgb9y4UTk5OXrnnXe0bt061dbW6pJLLtHBgwebvc3KlSs1b948LViwQDt37tRzzz2nF154QfPnz49icgAAACAyYv384mvXrm3w+YoVK5SamqotW7bo/PPPb/I2b7/9ts4991xdc801kqQhQ4bo6quv1h//+MeI5wUAAAAizalr0CsqKiRJffv2bfaY8ePHa8uWLdq8ebMk6eOPP9aaNWt06aWXNnl8TU2NKisrG3wA8AfrEXADaxFwmzMDen19vebMmaNzzz1XI0aMaPa4a665Rg888IAmTJiguLg4nXLKKZo4cWKzl7jk5+crOTk59JGRkRGpCgCOg/UIuIG1CLjNmQE9JydHO3bs0KpVq1o8rrCwUA899JB+8YtfaOvWrfrtb3+r1atX68EHH2zy+Ly8PFVUVIQ+9u7dG4n4AFqB9Qi4gbUIuM3Xa9CPys3N1WuvvaY333xT6enpLR57zz336LrrrtNNN90kSRo5cqQOHjyoH/zgB7rrrrvUrVvDvzkCgYACgUDEsgNoPdYj4AbWIuA2Xwd0z/P0ox/9SAUFBSosLNTQoUOPe5uqqqpGQ3j37t1D9wcAAABY5uuAnpOTo5UrV+qVV15RUlKSSktLJUnJyclKTEyUJM2YMUODBg1Sfn6+JGnKlCl69NFH9e1vf1vjxo1TcXGx7rnnHk2ZMiU0qAMAAABW+TqgP/XUU5KkiRMnNti+fPlyzZo1S5K0Z8+eBs+Y33333YqJidHdd9+tzz77TP3799eUKVO0cOHCaMUGAAAAIsb3S1yOp7CwsMHnsbGxWrBggRYsWBChVFLJ/ioFq+tUeahWyYlx6pUQq/Q+PSL29cLNen7Jfgfr+SU3OriQoaOsd7CeX7LfwXp+AG3nxItEXfJp+UHNL9iut4rLQ9smZKZo4bSRGpzS08dkrWM9v2S/g/X8khsdXMjQUdY7WM8v2e9gPT+A9nHmbRZdULK/qtGJUJI2FZfrroLtKtlf5VOy1rGeX7LfwXp+yY0OLmToKOsdrOeX7Hewnh9A+zGgHyNYXdfoRHjUpuJyBavropyobaznl+x3sJ5fcqODCxk6ynoH6/kl+x2s5wfQfgzox6g8VNvi/mB1y/v9Zj2/ZL+D9fySGx1cyNBR1jtYzy/Z72A9P4D2Y0A/Ru/EuBb3JyW0vN9v1vNL9jtYzy+50cGFDB1lvYP1/JL9DtbzA2g/BvRjJCXEakJmSpP7JmSmKCnB7dfUWs8v2e9gPb/kRgcXMnSU9Q7W80v2O1jPD6D9GNCPkd6nhxZOG9nohHj0FfOuv62V9fyS/Q7W80tudHAhQ0dZ72A9v2S/g/X8ANqPP7+/YXBKTy26cpSC1XUKVtcqKSFOSYbec9Z6fsl+B+v5JTc6uJCho6x3sJ5fst/Ben4A7cOA3gTrJz7r+SX7Haznl9zo4EKGjrLewXp+yX4H6/kBtB2XuAAAAAAOYUAHAAAAHMKADgAAADiEAR0AAABwCAM6AAAA4BAGdAAAAMAhDOgAAACAQxjQAQAAAIcwoAMAAAAOYUAHAAAAHMKADgAAADiEAR0AAABwCAM6AAAA4BAGdAAAAMAhDOgAAACAQxjQAQAAAIcwoAMAAAAOYUAHAAAAHMKADgAAADjE1wE9Pz9fY8eOVVJSklJTUzV16lTt2rXruLf7+uuvlZOTo4EDByoQCOi0007TmjVropAYAAAAiKxYP7/4xo0blZOTo7Fjx6qurk7z58/XJZdcovfff189e/Zs8jaHDx/WxRdfrNTUVL300ksaNGiQPv30U51wwgnRDQ8AAABEgK8D+tq1axt8vmLFCqWmpmrLli06//zzm7zNsmXL9Le//U1vv/224uLiJElDhgyJdFQAAAAgKpy6Br2iokKS1Ldv32aPefXVV5WVlaWcnBylpaVpxIgReuihh3TkyJFoxQQAAAAixtdn0I9VX1+vOXPm6Nxzz9WIESOaPe7jjz/WG2+8oWuvvVZr1qxRcXGx/u3f/k21tbVasGBBo+NrampUU1MT+ryysjIi+QEcH+sRcANrEXCbM8+g5+TkaMeOHVq1alWLx9XX1ys1NVXPPPOMxowZo+nTp+uuu+7S0qVLmzw+Pz9fycnJoY+MjIxIxAfQCqxHwA2sRcBtTgzoubm5eu2117Rhwwalp6e3eOzAgQN12mmnqXv37qFtp59+ukpLS3X48OFGx+fl5amioiL0sXfv3rDnB9A6rEfADaxFwG2+XuLieZ5+9KMfqaCgQIWFhRo6dOhxb3Puuedq5cqVqq+vV7duf//74sMPP9TAgQMVHx/f6PhAIKBAIBD27ADajvUIuIG1CLjN12fQc3Jy9Ktf/UorV65UUlKSSktLVVpaqkOHDoWOmTFjhvLy8kKf33LLLfrb3/6mW2+9VR9++KFWr16thx56SDk5OX5UAAAAAMLK12fQn3rqKUnSxIkTG2xfvny5Zs2aJUnas2dP6JlyScrIyNDrr7+u2267TaNGjdKgQYN066236s4774xWbAAAACBifL/E5XgKCwsbbcvKytI777wTgUQAAACAv5x4kSgAAACAv2NABwAAABzCgA4AAAA4hAEdAAAAcAgDOgAAAOAQBnQAAADAIQzoAAAAgEMY0AEAAACHtGlA//Of/6yf/vSn+sUvfqF9+/Y12FdZWakbbrghrOEAAACArqbVA/rvf/97nXPOOVq1apV+9rOfafjw4dqwYUNo/6FDh/T8889HJCQAAADQVcS29sD77rtPP/nJT7Rw4UJ5nqeHH35Y//zP/6wXX3xRkydPjmTGqCvZX6VgdZ0qD9UqOTFOvRJild6nh9+xWs16fsl+B+v5JTc6uJCho6x3sJ5fst/Ben4AbdfqAf2vf/2r/s//+T+SpJiYGN1xxx1KT0/X9773Pa1atUpjx46NWMho+rT8oOYXbNdbxeWhbRMyU7Rw2kgNTunpY7LWsZ5fst/Ben7JjQ4uZOgo6x2s55fsd7CeH0D7tPoSl0AgoK+//rrBtmuuuUa//OUvNX36dBUUFIQ7W9SV7K9qdCKUpE3F5bqrYLtK9lf5lKx1rOeX7Hewnl9yo4MLGTrKegfr+SX7HaznB9B+rX4GffTo0dqwYYPGjBnTYPv3v/99eZ6nmTNnhj1ctAWr6xqdCI/aVFyuYHVdlBO1jfX8kv0O1vNLbnRwIUNHWe9gPb9kv4P1/ADar9UD+i233KI333yzyX1XX321PM/Ts88+G7Zgfqg8VNvi/mB1y/v9Zj2/ZL+D9fySGx1cyNBR1jtYzy/Z72A9P4D2a/WAPm3aNE2bNk0bNmzQhRde2Gj/Nddco2AwGNZw0dY7Ma7F/UkJLe/3m/X8kv0O1vNLbnRwIUNHWe9gPb9kv4P1/ADar83/UNHkyZN1++23q7b2H3+579u3T1OmTNG8efPCGi7akhJiNSEzpcl9EzJTlJTQ6r9nfGE9v2S/g/X8khsdXMjQUdY7WM8v2e9gPT+A9mvzgL5hwwYVFBRo7Nixev/997V69WqNGDFCFRUV2rZtWwQiRk96nx5aOG1koxPi0VfMu/62VtbzS/Y7WM8vudHBhQwdZb2D9fyS/Q7W8wNovzb/+T1+/Hht27ZNN998s8466yzV19frwQcf1B133KGYmJhIZIyqwSk9tejKUQpW1ylYXaukhDglGXrPWev5JfsdrOeX3OjgQoaOst7Ben7Jfgfr+QG0T7v++9iHH36od999V+np6fr888+1a9cuVVVVqWfPzvGerNZPfNbzS/Y7WM8vudHBhQwdZb2D9fyS/Q7W8wNouzZf4rJo0SJlZWXp4osv1o4dO7R582a99957GjVqlIqKiiKREQAAAOgy2jygL1myRC+//LKeeOIJJSQkaMSIEdq8ebOuuOIKTZw4MQIRAQAAgK6jzZe4bN++Xf369WuwLS4uTg8//LAuv/zysAUDAAAAuqI2P4P+zeH8WBdccEGHwgAAAABdXZsHdAAAAACRw4AOAAAAOIQBHQAAAHAIAzoAAADgEAZ0AAAAwCG+Duj5+fkaO3askpKSlJqaqqlTp2rXrl2tvv2qVasUExOjqVOnRi4kAAAAEEW+DugbN25UTk6O3nnnHa1bt061tbW65JJLdPDgwePedvfu3frJT36i8847LwpJAQAAgOho8z9UFE5r165t8PmKFSuUmpqqLVu26Pzzz2/2dkeOHNG1116r+++/X//zP/+jr7/+OsJJAQAAgOhw6hr0iooKSVLfvn1bPO6BBx5QamqqbrzxxmjEAgAAAKLG12fQj1VfX685c+bo3HPP1YgRI5o9btOmTXruuee0bdu2Vt1vTU2NampqQp9XVlZ2NCqAdmI9Am5gLQJuc+YZ9JycHO3YsUOrVq1q9phgMKjrrrtOzz77rPr169eq+83Pz1dycnLoIyMjI1yRAbQR6xFwA2sRcJsTA3pubq5ee+01bdiwQenp6c0e99FHH2n37t2aMmWKYmNjFRsbq//6r//Sq6++qtjYWH300UeNbpOXl6eKiorQx969eyNZBUALWI+AG1iLgNt8vcTF8zz96Ec/UkFBgQoLCzV06NAWjx8+fLi2b9/eYNvdd9+tYDCoJUuWNPkMQCAQUCAQCGtuAO3DegTcwFoE3ObrgJ6Tk6OVK1fqlVdeUVJSkkpLSyVJycnJSkxMlCTNmDFDgwYNUn5+vhISEhpdn37CCSdIUovXrQMAAABW+DqgP/XUU5KkiRMnNti+fPlyzZo1S5K0Z88edevmxJU4AAAAQMT5fonL8RQWFra4f8WKFeEJAwAAADiAp6YBAAAAhzCgAwAAAA5hQAcAAAAcwoAOAAAAOIQBHQAAAHAIAzoAAADgEAZ0AAAAwCEM6AAAAIBDGNABAAAAhzCgAwAAAA5hQAcAAAAcwoAOAAAAOIQBHQAAAHAIAzoAAADgEAZ0AAAAwCEM6AAAAIBDGNABAAAAhzCgAwAAAA5hQAcAAAAcEut3ABeV7K9SsLpOlYdqlZwYp14JsUrv08PvWK1mPb9kv4P1/JIbHVzI0FHWO1jPL9nvYD0/gLZjQP+GT8sPan7Bdr1VXB7aNiEzRQunjdTglJ4+Jmsd6/kl+x2s55fc6OBCho6y3sF6fsl+B+v5AbQPl7gco2R/VaMToSRtKi7XXQXbVbK/yqdkrWM9v2S/g/X8khsdXMjQUdY7WM8v2e9gPT+A9mNAP0awuq7RifCoTcXlClbXRTlR21jPL9nvYD2/5EYHFzJ0lPUO1vNL9jtYzw+g/RjQj1F5qLbF/cHqlvf7zXp+yX4H6/klNzq4kKGjrHewnl+y38F6fgDtx4B+jN6JcS3uT0poeb/frOeX7Hewnl9yo4MLGTrKegfr+SX7HaznB9B+DOjHSEqI1YTMlCb3TchMUVKC26+ptZ5fst/Ben7JjQ4uZOgo6x2s55fsd7CeH0D7MaAfI71PDy2cNrLRCfHoK+Zdf1sr6/kl+x2s55fc6OBCho6y3sF6fsl+B+v5AbQff35/w+CUnlp05SgFq+sUrK5VUkKckgy956z1/JL9DtbzS250cCFDR1nvYD2/ZL+D9fwA2ocBvQnWT3zW80v2O1jPL7nRwYUMHWW9g/X8kv0O1vMDaDtfL3HJz8/X2LFjlZSUpNTUVE2dOlW7du1q8TbPPvuszjvvPPXp00d9+vRRdna2Nm/eHKXEAAAAQGT5OqBv3LhROTk5euedd7Ru3TrV1tbqkksu0cGDB5u9TWFhoa6++mpt2LBBRUVFysjI0CWXXKLPPvssiskBAACAyPD1Epe1a9c2+HzFihVKTU3Vli1bdP755zd5m1//+tcNPv/lL3+p//t//6/Wr1+vGTNmRCwrAAAAEA1OXYNeUVEhSerbt2+rb1NVVaXa2tpmb1NTU6OamprQ55WVlR0LCaDdWI+AG1iLgNuceZvF+vp6zZkzR+eee65GjBjR6tvdeeedOvHEE5Wdnd3k/vz8fCUnJ4c+MjIywhUZQBuxHgE3sBYBtzkzoOfk5GjHjh1atWpVq2+zaNEirVq1SgUFBUpISGjymLy8PFVUVIQ+9u7dG67IANqI9Qi4gbUIuM2JS1xyc3P12muv6c0331R6enqrbrN48WItWrRIf/jDHzRq1KhmjwsEAgoEAuGKCqADWI+AG1iLgNt8HdA9z9OPfvQjFRQUqLCwUEOHDm3V7X7+859r4cKFev3113X22WdHOCUAAAAQPb4O6Dk5OVq5cqVeeeUVJSUlqbS0VJKUnJysxMRESdKMGTM0aNAg5efnS5J+9rOf6d5779XKlSs1ZMiQ0G169eqlXr16+VMEAAAACBNfr0F/6qmnVFFRoYkTJ2rgwIGhjxdeeCF0zJ49e/TFF180uM3hw4f1ve99r8FtFi9e7EcFAAAAIKx8v8TleAoLCxt8vnv37siEAQAAABzgzLu4AAAAAGBABwAAAJzCgA4AAAA4hAEdAAAAcAgDOgAAAOAQBnQAAADAIQzoAAAAgEMY0AEAAACHMKADAAAADmFABwAAABzCgA4AAAA4hAEdAAAAcAgDOgAAAOAQBnQAAADAIQzoAAAAgEMY0AEAAACHMKADAAAADmFABwAAABzCgA4AAAA4hAEdAAAAcAgDOgAAAOAQBnQAAADAIQzoAAAAgEMY0AEAAACHMKADAAAADmFABwAAABzCgA4AAAA4hAEdAAAAcIivA3p+fr7Gjh2rpKQkpaamaurUqdq1a9dxb/fiiy9q+PDhSkhI0MiRI7VmzZoopAUAAAAiL9bPL75x40bl5ORo7Nixqqur0/z583XJJZfo/fffV8+ePZu8zdtvv62rr75a+fn5uvzyy7Vy5UpNnTpVW7du1YgRI8KSq2R/lYLVdao8VKvkxDj1SohVep8eYbnvaLCeX7LfwXp+yY0OLmToKOsdrOeX7Hewnh9A2/k6oK9du7bB5ytWrFBqaqq2bNmi888/v8nbLFmyRJMnT9btt98uSXrwwQe1bt06/ed//qeWLl3a4Uyflh/U/ILtequ4PLRtQmaKFk4bqcEpTf/R4BLr+SX7Haznl9zo4EKGjrLewXp+yX4H6/kBtI9T16BXVFRIkvr27dvsMUVFRcrOzm6wbdKkSSoqKurw1y/ZX9XoRChJm4rLdVfBdpXsr+rw14gk6/kl+x2s55fc6OBCho6y3sF6fsl+B+v5AbSfr8+gH6u+vl5z5szRueee2+KlKqWlpUpLS2uwLS0tTaWlpU0eX1NTo5qamtDnlZWVzd53sLqu0YnwqE3F5QpW17VUwXfW80v2O1jPL0W2Q2vXI4+j/6znl+x3cGEtAvCHM8+g5+TkaMeOHVq1alVY7zc/P1/Jycmhj4yMjGaPrTxU2+J9Batb3u836/kl+x2s55ci26G165HH0X/W80v2O7iwFgH4w4kBPTc3V6+99po2bNig9PT0Fo8dMGCAysrKGmwrKyvTgAEDmjw+Ly9PFRUVoY+9e/c2e9+9E+Na/NpJCS3v95v1/JL9DtbzS5Ht0Nr1yOPoP+v5JfsdXFiLAPzh64DueZ5yc3NVUFCgN954Q0OHDj3ubbKysrR+/foG29atW6esrKwmjw8EAurdu3eDj+YkJcRqQmZKk/smZKYoKcGZK4KaZD2/ZL+D9fxSZDu0dj3yOPrPen7JfgcX1iIAf/g6oOfk5OhXv/qVVq5cqaSkJJWWlqq0tFSHDh0KHTNjxgzl5eWFPr/11lu1du1aPfLII/rggw9033336d1331Vubm6H86T36aGF00Y2OiEefcW8629rZT2/ZL+D9fySGx1cyNBR1jtYzy/Z72A9P4D28/Xpg6eeekqSNHHixAbbly9frlmzZkmS9uzZo27d/vF3xPjx47Vy5Urdfffdmj9/vk499VS9/PLLYXsP9MEpPbXoylEKVtcpWF2rpIQ4JRl6z1nr+SX7Haznl9zo4EKGjrLewXp+yX4H6/kBtI+vA7rnecc9prCwsNG2q666SldddVUEEv2d9ROf9fyS/Q7W80tudHAhQ0dZ72A9v2S/g/X8ANrOiReJAgAAAPg7BnQAAADAIQzoAAAAgEPcfo+pCDh63Tv/ahoQXklJSYqJiWnTbViPQGS0dT22Zi0eOHBAklRR8rHq6450LGAzgqWfSpK2bt0a+nrh9OGHH0qKXAfr+SX7HQ58+ZkkqbJ0twKBQNjv/+jjc+DAgVb97mrP70ZJivFa80rNTqSkpIR/MQ2IgIqKija/lzLrEYiMtq5H1iIQGe353Sh1wQG9vr5en3/++XH/oqmsrFRGRob27t1r8h9wsJ5fst/Ben6pbR3a8yxBa9ZjV3scXWQ9v2S/Q1vzt3U9shZtsJ5fst8h0mvxqC53iUu3bt2Unp7e6uOt/wtr1vNL9jtYzy9FrkNb1iOPo/+s55fsd2Athof1DtbzS/Y7RDo/LxIFAAAAHMKADgAAADiEAb0ZgUBACxYsiMgrgKPBen7Jfgfr+SU3OriQoaOsd7CeX7LfwYX8LmToKOsdrOeX7HeIVv4u9yJRAAAAwGU8gw4AAAA4hAEdAAAAcAgDOgAAAOAQBnQAAADAIV1qQH/yySc1ZMgQJSQkaNy4cdq8eXOLx7/44osaPny4EhISNHLkSK1Zs6bBfs/zdO+992rgwIFKTExUdna2/vd//9eJ/M8++6zOO+889enTR3369FF2dnaj42fNmqWYmJgGH5MnT3Yi/4oVKxplS0hIaHBMtB//tnaYOHFiow4xMTG67LLLQsdE83vw5ptvasqUKTrxxBMVExOjl19++bi3KSws1FlnnaVAIKDMzEytWLGi0TFtXVetMWTIkEaPy6JFi1q8TXV1tXJycpSSkqJevXrpyiuvVFlZWYeztNXu3bt14403aujQoUpMTNQpp5yiBQsW6PDhwy3erqmfl5tvvjlKqcN/foym/Px8jR07VklJSUpNTdXUqVO1a9euFm/TmnNMtNx3332NsgwfPrzF23T08W/PY9aarxvN83J7Orj0uzFSP7fR+h60J79rvxefeuopjRo1KvSPDmVlZel3v/tdi7eJ2hrwuohVq1Z58fHx3rJly7y//vWv3uzZs70TTjjBKysra/L4t956y+vevbv385//3Hv//fe9u+++24uLi/O2b98eOmbRokVecnKy9/LLL3t//vOfvX/+53/2hg4d6h06dMj3/Ndcc4335JNPeu+99563c+dOb9asWV5ycrJXUlISOmbmzJne5MmTvS+++CL08be//S3s2duTf/ny5V7v3r0bZCstLW1wTDQf//Z0KC8vb5B/x44dXvfu3b3ly5eHjonm92DNmjXeXXfd5f32t7/1JHkFBQUtHv/xxx97PXr08ObOneu9//773hNPPOF1797dW7t2beiYtj4mrTV48GDvgQceaPC4HDhwoMXb3HzzzV5GRoa3fv1679133/X+6Z/+yRs/fnyHcrTH7373O2/WrFne66+/7n300UfeK6+84qWmpno//vGPW7zdBRdc4M2ePbtB54qKiqhkjsT5MZomTZrkLV++3NuxY4e3bds279JLL/VOOumkFn9mWnOOiZYFCxZ43/rWtxpk+eqrr5o9PhyPf3seM9d+L7ang0u/GyP1cxut70F78rv2e/HVV1/1Vq9e7X344Yferl27vPnz53txcXHejh07mjw+mmugywzo55xzjpeTkxP6/MiRI96JJ57o5efnN3n8v/zLv3iXXXZZg23jxo3zfvjDH3qe53n19fXegAEDvIcffji0/+uvv/YCgYD33//9377n/6a6ujovKSnJe/7550PbZs6c6X33u98Nd9QmtTX/8uXLveTk5GbvL9qPv+d1/Hvw2GOPeUlJSQ1OXtH8HhyrNQP6HXfc4X3rW99qsG369OnepEmTQp939DFpzuDBg73HHnus1cd//fXXXlxcnPfiiy+Gtu3cudOT5BUVFXUoSzj8/Oc/94YOHdriMRdccIF36623RifQN4T7/Oi3L7/80pPkbdy4sdljjneOiaYFCxZ4Z555ZquPj8Tj35rHzLXfi9/Umg7f5PfvxmOF4+fWz+9Bex5/l34vHtWnTx/vl7/8ZZP7orkGusQlLocPH9aWLVuUnZ0d2tatWzdlZ2erqKioydsUFRU1OF6SJk2aFDr+k08+UWlpaYNjkpOTNW7cuGbvM5r5v6mqqkq1tbXq27dvg+2FhYVKTU3VsGHDdMstt6i8vDys2aX25z9w4IAGDx6sjIwMffe739Vf//rX0L5oPv4d6XCs5557Tt///vfVs2fPBtuj8T1oj+OtgXA8Ji1ZtGiRUlJS9O1vf1sPP/yw6urqmj12y5Ytqq2tbZBl+PDhOumkkyLy89BWFRUVjdZeU37961+rX79+GjFihPLy8lRVVRXxbJE4P/qtoqJCko77mLd0jom2//3f/9WJJ56ok08+Wddee6327NnT7LGRePxb85i59HuxKa39vh/Lz9+N3xSOn1s/vwftefxd+r145MgRrVq1SgcPHlRWVlaTx0RzDcS2Mb9J+/bt05EjR5SWltZge1pamj744IMmb1NaWtrk8aWlpaH9R7c1d0y4tCf/N91555068cQTG/zQTJ48WVdccYWGDh2qjz76SPPnz9d3vvMdFRUVqXv37r7mHzZsmJYtW6ZRo0apoqJCixcv1vjx4/XXv/5V6enpUX3829vhWJs3b9aOHTv03HPPNdgere9BezS3BiorK3Xo0CHt37+/wz+Xzfn3f/93nXXWWerbt6/efvtt5eXl6YsvvtCjjz7abNb4+HidcMIJjbJE4uehLYqLi/XEE09o8eLFLR53zTXXaPDgwTrxxBP1l7/8RXfeead27dql3/72txHNF4nzo5/q6+s1Z84cnXvuuRoxYkSzxx3vHBNN48aN04oVKzRs2DB98cUXuv/++3Xeeedpx44dSkpKanR8uB//1j5mLv1e/KbWdvgmP383HitcP7d+fQ/a8/i78ntx+/btysrKUnV1tXr16qWCggKdccYZTR4bzTXQJQb0rm7RokVatWqVCgsLG7yY5Pvf/37o/48cOVKjRo3SKaecosLCQl100UV+RA3Jyspq8Bfs+PHjdfrpp+vpp5/Wgw8+6GOy9nnuuec0cuRInXPOOQ22u/w9CLd58+bpZz/7WYvH7Ny5U8OHD9fcuXND20aNGqX4+Hj98Ic/VH5+vm//PHRb8h/12WefafLkybrqqqs0e/bsFm/7gx/8IPT/R44cqYEDB+qiiy7SRx99pFNOOaVj4buQnJwc7dixQ5s2bWrxOJfOMd/5zndC/3/UqFEaN26cBg8erN/85je68cYbI/71W/uYuaw9HVz63Wjx5/ZY7Xn8Xfm9OGzYMG3btk0VFRV66aWXNHPmTG3cuLHZIT1ausQlLv369VP37t0bvaNDWVmZBgwY0ORtBgwY0OLxR/+3LffZXu3Jf9TixYu1aNEi/f73v9eoUaNaPPbkk09Wv379VFxc3OHMx+pI/qPi4uL07W9/O5Qtmo+/1LEOBw8e1KpVq1r1izZS34P2aG4N9O7dW4mJiW1+TH784x9r586dLX6cfPLJTWYZN26c6urqtHv37mazHj58WF9//XWrsrRHW/N//vnnuvDCCzV+/Hg988wzbf5648aNk6SI/yxE4vzol9zcXL322mvasGFDm58F/+Y5xk8nnHCCTjvttGazhPPxb8tj5tLvxWO15/vuwu/Go8L5c+vH96A9+V36vRgfH6/MzEyNGTNG+fn5OvPMM7VkyZImj43mGugSA3p8fLzGjBmj9evXh7bV19dr/fr1zV5nlJWV1eB4SVq3bl3o+KFDh2rAgAENjqmsrNQf//jHZu8zmvkl6ec//7kefPBBrV27VmefffZxv05JSYnKy8s1cODAsOQ+qr35j3XkyBFt3749lC2aj7/UsQ4vvviiampq9K//+q/H/TqR+h60x/HWQFsfk/79+2v48OEtfsTHxzeZZdu2berWrZtSU1Ob3D9mzBjFxcU1yLJr1y7t2bMnbD8Pbcn/2WefaeLEiRozZoyWL1+ubt3afqrdtm2bJEX8ZyES58do8zxPubm5Kigo0BtvvKGhQ4e2+T6+eY7x04EDB/TRRx81myUcj397HjOXfi9K7f++u/K7MRI/t9H8HnQkv8u/F+vr61VTU9PkvqiugTa9pNSwVatWeYFAwFuxYoX3/vvvez/4wQ+8E044IfT2RNddd503b9680PFvvfWWFxsb6y1evNjbuXOnt2DBgibfSueEE07wXnnlFe8vf/mL993vfjeib7PYlvyLFi3y4uPjvZdeeqnBWxUFg0HP8zwvGAx6P/nJT7yioiLvk08+8f7whz94Z511lnfqqad61dXVvue///77Q29Tt2XLFu/73/++l5CQ4P31r39t0DFaj397Ohw1YcIEb/r06Y22R/t7EAwGvffee8977733PEneo48+6r333nvep59+6nme582bN8+77rrrQscffZvF22+/3du5c6f35JNPNvk2iy09Ju3x9ttve4899pi3bds276OPPvJ+9atfef379/dmzJgROqakpMQbNmyY98c//jG07eabb/ZOOukk74033vDeffddLysry8vKymp3jvYqKSnxMjMzvYsuusgrKSlpsP6ay19cXOw98MAD3rvvvut98skn3iuvvOKdfPLJ3vnnnx+VzJE4P0bTLbfc4iUnJ3uFhYUNHu+qqqrQMe05x0TLj3/8Y6+wsND75JNPvLfeesvLzs72+vXr53355ZdNZg/H49+ex8y134vt6eDS78ZI/dxG63vQnvxHufJ7cd68ed7GjRu9Tz75xPvLX/7izZs3z4uJifF+//vfN5k/mmugywzonud5TzzxhHfSSSd58fHx3jnnnOO98847oX0XXHCBN3PmzAbH/+Y3v/FOO+00Lz4+3vvWt77lrV69usH++vp675577vHS0tK8QCDgXXTRRd6uXbucyD948GBPUqOPBQsWeJ7neVVVVd4ll1zi9e/f34uLi/MGDx7szZ49O6LvA9yW/HPmzAkdm5aW5l166aXe1q1bG9xftB//tnbwPM/74IMPPEmhxX6saH8PNmzY0OTPxNHMM2fO9C644IJGtxk9erQXHx/vnXzyyQ3eq/aolh6T9tiyZYs3btw4Lzk52UtISPBOP/1076GHHmpwcv7kk088Sd6GDRtC2w4dOuT927/9m9enTx+vR48e3rRp0xoMxdGyfPnyJh/nY58P+Wb+PXv2eOeff77Xt29fLxAIeJmZmd7tt98etfdB97zwnx+jqbnH+9if1/acY6Jl+vTp3sCBA734+Hhv0KBB3vTp073i4uLQ/kg8/u15zFrzdaN5Xm5PB5d+N0bq5zZa34P2/gy59Hvxhhtu8AYPHuzFx8d7/fv39y666KIGufxcAzGe53lte84dAAAAQKR0iWvQAQAAACsY0AEAAACHMKADAAAADmFABwAAABzCgA4AAAA4hAEdAAAAcAgDOgAAAOAQBnQAAADAIQzocF5hYaHOOussBQIBZWZmasWKFX5HArqcL774Qtdcc41OO+00devWTXPmzPE7EtAl/fa3v9XFF1+s/v37q3fv3srKytLrr7/udyyEGQM6nPbJJ5/osssu04UXXqht27Zpzpw5uummmzgZAVFWU1Oj/v376+6779aZZ57pdxygy3rzzTd18cUXa82aNdqyZYsuvPBCTZkyRe+9957f0RBGDOjw1VdffaUBAwbooYceCm17++23FR8fr/Xr12vp0qUaOnSoHnnkEZ1++unKzc3V9773PT322GM+pgY6n+OtxSFDhmjJkiWaMWOGkpOTfUwKdG7HW4uPP/647rjjDo0dO1annnqqHnroIZ166qn6f//v//mYGuHGgA5f9e/fX8uWLdN9992nd999V8FgUNddd51yc3N10UUXqaioSNnZ2Q1uM2nSJBUVFfmUGOicjrcWAURHW9difX29gsGg+vbt60NaREqs3wGASy+9VLNnz9a1116rs88+Wz179lR+fr4kqbS0VGlpaQ2OT0tLU2VlpQ4dOqTExEQ/IgOdUktrEUD0tGUtLl68WAcOHNC//Mu/RDklIokBHU5YvHixRowYoRdffFFbtmxRIBDwOxLQJbEWATe0Zi2uXLlS999/v1555RWlpqb6kBKRwiUucMJHH32kzz//XPX19dq9e3do+4ABA1RWVtbg2LKyMvXu3Ztnz4EIaG4tAoiu463FVatW6aabbtJvfvObRpeCwj6eQYfvDh8+rH/913/V9OnTNWzYMN10003avn27UlNTlZWVpTVr1jQ4ft26dcrKyvIpLdB5tbQWAUTP8dbif//3f+uGG27QqlWrdNlll/mcFpEQ43me53cIdG233367XnrpJf35z39Wr169dMEFFyg5OVmvvfaaPvnkE40YMUI5OTm64YYb9MYbb+jf//3ftXr1ak2aNMnv6ECn0tJalKRt27ZJkm666SYNGzZMt99+u+Lj43XGGWf4mBrofFpaiytXrtTMmTO1ZMkSXXHFFaHbJCYm8g5LnYkH+GjDhg1ebGys9z//8z+hbZ988onXu3dv7xe/+EXomNGjR3vx8fHeySef7C1fvtyntEDn1Zq1KKnRx+DBg31KDHROx1uLF1xwQZNrcebMmf6FRtjxDDoAAADgEF4kCgAAADiEAR0AAABwCAM6AAAA4BAGdAAAAMAhDOgAAACAQxjQAQAAAIcwoAMAAAAOYUAHAAAAHMKADgAAADiEAR0AAABwCAM6AAAA4BAGdAAAAMAh/x9e/x12goF5JwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 750x750 with 12 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "\n",
    "def plot(experimentdata: ExperimentData):\n",
    "    input_data, output_data = experimentdata.to_pandas()\n",
    "    sns.pairplot(input_data)   \n",
    "    \n",
    "plot(experimentdata)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a4c402ac",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Fantastic! You can use any packages or software to sample from your ``Domain`` as long as your sampling function complies to the interface that ``f3dasm`` describes!\n",
    "\n",
    "*The column 'jobs' keeps track of input_data that needs to be evaluated. In the case of our data, they are all open since we haven't evaluated our samples yet!*"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "70702f51",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<img src=./img/schematics/block_sampling.png width=10%, align='right'>\n",
    "\n",
    "``f3dasm`` comes with some built-in samplers that you can use right away!\n",
    "\n",
    "| Name            | Method                             | Reference                                                                                              |\n",
    "|-----------------|-----------------------------------|--------------------------------------------------------------------------------------------------------|\n",
    "| `\"random\"`      | Random Uniform sampling           | [numpy.random.uniform](https://numpy.org/doc/stable/reference/random/generated/numpy.random.uniform.html) |\n",
    "| `\"latin\"`       | Latin Hypercube sampling          | [SALib.latin](https://salib.readthedocs.io/en/latest/api/SALib.sample.html?highlight=latin%20hypercube#SALib.sample.latin.sample)         |\n",
    "| `\"sobol\"`       | Sobol Sequence sampling           | [SALib.sobol_sequence](https://salib.readthedocs.io/en/latest/api/SALib.sample.html?highlight=sobol%20sequence#SALib.sample.sobol_sequence.sample) |\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "9b890d97",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "experimentdata = ExperimentData(domain)\n",
    "experimentdata.sample(sampler='random', n_samples=100, seed=2023)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "4756298e",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>jobs</th>\n",
       "      <th colspan=\"3\" halign=\"left\">input</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>x0</th>\n",
       "      <th>x1</th>\n",
       "      <th>x2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>open</td>\n",
       "      <td>0.321988</td>\n",
       "      <td>3.904225</td>\n",
       "      <td>2.588052</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>open</td>\n",
       "      <td>0.126596</td>\n",
       "      <td>-3.586588</td>\n",
       "      <td>2.467896</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>open</td>\n",
       "      <td>0.022090</td>\n",
       "      <td>2.272747</td>\n",
       "      <td>2.524387</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>open</td>\n",
       "      <td>0.544935</td>\n",
       "      <td>-0.436267</td>\n",
       "      <td>2.501382</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>open</td>\n",
       "      <td>0.394469</td>\n",
       "      <td>-3.488277</td>\n",
       "      <td>2.360875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95</th>\n",
       "      <td>open</td>\n",
       "      <td>0.323738</td>\n",
       "      <td>2.853825</td>\n",
       "      <td>2.381702</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>open</td>\n",
       "      <td>0.850285</td>\n",
       "      <td>4.655796</td>\n",
       "      <td>2.049404</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>open</td>\n",
       "      <td>0.626839</td>\n",
       "      <td>-2.189198</td>\n",
       "      <td>2.505979</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>open</td>\n",
       "      <td>0.532520</td>\n",
       "      <td>-0.532099</td>\n",
       "      <td>2.971766</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>open</td>\n",
       "      <td>0.433291</td>\n",
       "      <td>1.099487</td>\n",
       "      <td>2.996351</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "<f3dasm._src.experimentdata.experimentdata.ExperimentData at 0x7f2bd7ebbca0>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "experimentdata"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "138b8c91",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 750x750 with 12 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot(experimentdata)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "577ae081",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<img src=./img/schematics/block_sampling.png width=10%, align='right'>\n",
    "\n",
    "**Method 2**: If you already have input data, you can provide it as a ``pandas`` Dataframe, ``numpy`` array or .csv file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "6c36ac99",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>jobs</th>\n",
       "      <th colspan=\"3\" halign=\"left\">input</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>x0</th>\n",
       "      <th>x1</th>\n",
       "      <th>x2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>open</td>\n",
       "      <td>0.321988</td>\n",
       "      <td>3.904225</td>\n",
       "      <td>2.588052</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>open</td>\n",
       "      <td>0.126596</td>\n",
       "      <td>-3.586588</td>\n",
       "      <td>2.467896</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>open</td>\n",
       "      <td>0.022090</td>\n",
       "      <td>2.272747</td>\n",
       "      <td>2.524387</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>open</td>\n",
       "      <td>0.544935</td>\n",
       "      <td>-0.436267</td>\n",
       "      <td>2.501382</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>open</td>\n",
       "      <td>0.394469</td>\n",
       "      <td>-3.488277</td>\n",
       "      <td>2.360875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95</th>\n",
       "      <td>open</td>\n",
       "      <td>0.323738</td>\n",
       "      <td>2.853825</td>\n",
       "      <td>2.381702</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>open</td>\n",
       "      <td>0.850285</td>\n",
       "      <td>4.655796</td>\n",
       "      <td>2.049404</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>open</td>\n",
       "      <td>0.626839</td>\n",
       "      <td>-2.189198</td>\n",
       "      <td>2.505979</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>open</td>\n",
       "      <td>0.532520</td>\n",
       "      <td>-0.532099</td>\n",
       "      <td>2.971766</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>open</td>\n",
       "      <td>0.433291</td>\n",
       "      <td>1.099487</td>\n",
       "      <td>2.996351</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "<f3dasm._src.experimentdata.experimentdata.ExperimentData at 0x7f2bd7964460>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "experimentdata_from_df = ExperimentData(domain=domain, input_data='./input_data.csv')\n",
    "experimentdata_from_df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cbfb1552",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## ExperimentSample\n",
    "\n",
    "Each sample of the ``ExperimentData`` can be accessed with the ``._get_experiment_sample`` method with the index of the sample you want to access:\n",
    "\n",
    "<img src=\"./img/f3dasm-design.png\" width=\"80%\" title=\"f3dasm experimentsample\">"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "a73371bf",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ExperimentSample(0 : {'x0': 0.3219883039919677, 'x1': 3.904224517894807, 'x2': 2.588052255450474} - {})"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "experimentdata_from_df._get_experiment_sample(0)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e082d2ed",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<img src=./img/schematics/block_datageneration.png width=10%, align='right'>\n",
    "\n",
    "\n",
    "This will create a new object of type `ExperimentSample` that is useful in the next section: Data generation!"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0d69de89",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=./img/schematics/block_datageneration.png width=10%, align='right'>\n",
    "\n",
    "## 2. Data generation\n",
    "\n",
    "Now that we have sampled input-data, we can evaluate the designs with a ``DataGenerator``!\n",
    "\n",
    "To implement your datagenerator, you have to create a new class that inherits from the ``DataGenerator`` base class:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eaf7dff8",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "Let's create a simple data generator that computes $f(x) = x \\; \\mathrm{sin}(x)$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "f9e30301",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "from f3dasm.datageneration import DataGenerator\n",
    "import numpy as np\n",
    "\n",
    "class MyDataGenerator(DataGenerator):\n",
    "    def execute(self):\n",
    "        \n",
    "        # Get the input_data and output_data of the ExperimentSample in numpy format\n",
    "        X, y = self.experiment_sample.to_numpy()\n",
    "        \n",
    "        # f(x) = x*sin(x)\n",
    "        y = np.sum(X * np.sin(X))\n",
    "        \n",
    "        # store the objective value under the name 'y'\n",
    "        self.experiment_sample.store(y, 'y')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6830a50a",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<img src=./img/schematics/block_datageneration.png width=10%, align='right'>\n",
    "\n",
    "Now we instantiate the data generator:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "0bd970e1",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "my_data_generator = MyDataGenerator()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5e1faa76",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "And we pass it to the ``evaluate`` method of your data object!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "ee390c43",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>jobs</th>\n",
       "      <th colspan=\"3\" halign=\"left\">input</th>\n",
       "      <th>output</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>x0</th>\n",
       "      <th>x1</th>\n",
       "      <th>x2</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.321988</td>\n",
       "      <td>3.904225</td>\n",
       "      <td>2.588052</td>\n",
       "      <td>-1.234704</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.126596</td>\n",
       "      <td>-3.586588</td>\n",
       "      <td>2.467896</td>\n",
       "      <td>0.011794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.022090</td>\n",
       "      <td>2.272747</td>\n",
       "      <td>2.524387</td>\n",
       "      <td>3.196932</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.544935</td>\n",
       "      <td>-0.436267</td>\n",
       "      <td>2.501382</td>\n",
       "      <td>1.961059</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.394469</td>\n",
       "      <td>-3.488277</td>\n",
       "      <td>2.360875</td>\n",
       "      <td>0.627909</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>95</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.323738</td>\n",
       "      <td>2.853825</td>\n",
       "      <td>2.381702</td>\n",
       "      <td>2.553553</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.850285</td>\n",
       "      <td>4.655796</td>\n",
       "      <td>2.049404</td>\n",
       "      <td>-2.190253</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.626839</td>\n",
       "      <td>-2.189198</td>\n",
       "      <td>2.505979</td>\n",
       "      <td>3.639196</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.532520</td>\n",
       "      <td>-0.532099</td>\n",
       "      <td>2.971766</td>\n",
       "      <td>1.042584</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.433291</td>\n",
       "      <td>1.099487</td>\n",
       "      <td>2.996351</td>\n",
       "      <td>1.595202</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "<f3dasm._src.experimentdata.experimentdata.ExperimentData at 0x7f2bd7ebbca0>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "experimentdata.evaluate(my_data_generator, mode='sequential')\n",
    "experimentdata"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "74bdf87c",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "The values in the 'jobs' column have been changed from ``open`` to ``finished`` !"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "401aee4e",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<img src=./img/schematics/block_datageneration.png width=10%, align='right'>\n",
    "\n",
    "There are three modes to evaluate your samples:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b51ffac7",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "* ``sequential``: each ``ExperimentSample`` object is processed one at a time"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f444fd96",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "* ``parallel``: utilizing the multiprocessing capabilities, each ``ExperimentSample`` object is evaluate on a **separate core**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a716655f",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "* ``cluster``: Each ``ExperimentSample`` object is run in a **separate node**. After finishing of an evaluation, the node will automatically pick the next available ``ExperimentSample``. This is especially useful in a high-performance computation cluster setting"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f98736bb",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<img src=./img/schematics/block_datageneration.png width=10%, align='right'>\n",
    "\n",
    "<img src=\"./img/Levy.png\" width=\"30%\" title=\"Levy Image\" align=\"right\">\n",
    "\n",
    "A datagenerator can be a call to a **third-party simulation software** (e.g. ABAQUS).\n",
    "In this example, we are replacing an expensive simulation with any of the built-in benchmark functions: The ``\"Levy\"`` function:\n",
    "\n",
    "*Note: the bounds have been normalized to $[0,1]$ in this figure*\n",
    "\n",
    "``f3dasm`` comes with **73** built-in benchmark function that you can use right away!\n",
    "\n",
    "*Ports to third-party software will be available in the [f3dasm_simulate](https://bessagroup.github.io/f3dasm_simulate/) extension package*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "3e5ecd60",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "experimentdata = ExperimentData(domain)\n",
    "experimentdata.sample(sampler='random', n_samples=100, seed=2023)\n",
    "experimentdata.evaluate(data_generator=\"Levy\", mode=\"sequential\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "25aa730d",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<img src=./img/schematics/block_datageneration.png width=10%, align='right'>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "662ef83e",
   "metadata": {
    "slideshow": {
     "slide_type": "-"
    }
   },
   "outputs": [
    {
     "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>jobs</th>\n",
       "      <th colspan=\"3\" halign=\"left\">input</th>\n",
       "      <th>output</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>x0</th>\n",
       "      <th>x1</th>\n",
       "      <th>x2</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.321988</td>\n",
       "      <td>3.904225</td>\n",
       "      <td>2.588052</td>\n",
       "      <td>1.188647</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.126596</td>\n",
       "      <td>-3.586588</td>\n",
       "      <td>2.467896</td>\n",
       "      <td>5.485291</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.022090</td>\n",
       "      <td>2.272747</td>\n",
       "      <td>2.524387</td>\n",
       "      <td>1.724939</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.544935</td>\n",
       "      <td>-0.436267</td>\n",
       "      <td>2.501382</td>\n",
       "      <td>0.542788</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.394469</td>\n",
       "      <td>-3.488277</td>\n",
       "      <td>2.360875</td>\n",
       "      <td>5.956175</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>95</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.323738</td>\n",
       "      <td>2.853825</td>\n",
       "      <td>2.381702</td>\n",
       "      <td>1.619962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.850285</td>\n",
       "      <td>4.655796</td>\n",
       "      <td>2.049404</td>\n",
       "      <td>4.707982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.626839</td>\n",
       "      <td>-2.189198</td>\n",
       "      <td>2.505979</td>\n",
       "      <td>7.305066</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.532520</td>\n",
       "      <td>-0.532099</td>\n",
       "      <td>2.971766</td>\n",
       "      <td>0.640278</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.433291</td>\n",
       "      <td>1.099487</td>\n",
       "      <td>2.996351</td>\n",
       "      <td>0.515674</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>100 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "<f3dasm._src.experimentdata.experimentdata.ExperimentData at 0x7f2bd797ab80>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "experimentdata"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fb478404",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=./img/schematics/block_optimization.png width=10%, align='right'>\n",
    "\n",
    "## 3. Optimization\n",
    "\n",
    "With ``f3dasm``, you can optimize the input-parameters of your ``ExperimentData`` object by striving to minimize the output_data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "ef32eabb",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/martin/Documents/GitHub/F3DASM/src/f3dasm/_src/optimization/adapters/scipy_implementations.py:54: OptimizeWarning: Unknown solver options: population, force_bounds\n",
      "  minimize(\n"
     ]
    }
   ],
   "source": [
    "experimentdata.optimize(data_generator=\"Levy\", optimizer=\"L-BFGS-B\", iterations=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "11e37263",
   "metadata": {
    "scrolled": true,
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>jobs</th>\n",
       "      <th colspan=\"3\" halign=\"left\">input</th>\n",
       "      <th>output</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>x0</th>\n",
       "      <th>x1</th>\n",
       "      <th>x2</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.321988</td>\n",
       "      <td>3.904225</td>\n",
       "      <td>2.588052</td>\n",
       "      <td>1.188647</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.126596</td>\n",
       "      <td>-3.586588</td>\n",
       "      <td>2.467896</td>\n",
       "      <td>5.485291</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.022090</td>\n",
       "      <td>2.272747</td>\n",
       "      <td>2.524387</td>\n",
       "      <td>1.724939</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.544935</td>\n",
       "      <td>-0.436267</td>\n",
       "      <td>2.501382</td>\n",
       "      <td>0.542788</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>finished</td>\n",
       "      <td>0.394469</td>\n",
       "      <td>-3.488277</td>\n",
       "      <td>2.360875</td>\n",
       "      <td>5.956175</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>195</th>\n",
       "      <td>finished</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000001</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.125000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>196</th>\n",
       "      <td>finished</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000001</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.125000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>finished</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000001</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.125000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198</th>\n",
       "      <td>finished</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000001</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.125000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>199</th>\n",
       "      <td>finished</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000001</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.125000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>200 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "<f3dasm._src.experimentdata.experimentdata.ExperimentData at 0x7f2bd797ab80>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "experimentdata"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "872e2723",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<img src=./img/schematics/block_optimization.png width=10%, align='right'>\n",
    "\n",
    "``f3dasm`` comes with some built-in optimizers that you can use right away!\n",
    "\n",
    "| Name           | Reference                                                      |\n",
    "| ---------------| --------------------------------------------------------------- |\n",
    "| `\"CG\"`             | [scipy.minimize CG](https://docs.scipy.org/doc/scipy/reference/optimize.minimize-cg.html) |\n",
    "| `\"L-BFGS-B\"`         | [scipy.minimize L-BFGS-B](https://docs.scipy.org/doc/scipy/reference/optimize.minimize-lbfgsb.html>) |\n",
    "| `\"NelderMead\"`     | [scipy.minimize NelderMead](https://docs.scipy.org/doc/scipy/reference/optimize.minimize-neldermead.html) |\n",
    "| `\"RandomSearch\"`   | [numpy.random](https://numpy.org/doc/>)       |\n",
    "\n",
    "*More optimizers are available in the [f3dasm_optimize](https://bessagroup.github.io/f3dasm_optimize/) extension package*\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bcac432f",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Storing\n",
    "\n",
    "After your data-driven process, you can store the data with the ``store`` method:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b89fee85",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "```\n",
    "experimentdata.store(filename='my_data')\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7ee50760",
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "source": [
    "This will store the `input_data` (.csv), `output_data` (.csv) and `domain` (.pkl) to disk "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "56c3a6bb",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## Current state of the project\n",
    "\n",
    "* Researchers of the 'Bessa Research Group' are implementing their research into the ``f3dasm`` framework\n",
    "* Students are working on projects with `f3dasm` at both Delft University of Technology and Brown University\n",
    "* Expect to submit the framework to the [Journal of Open Source Software](https://joss.theoj.org/)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ff0ba7ac",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Summary\n",
    "\n",
    "⭐ ``f3dasm`` provides a **unified interface** for data-driven processes for structures and materials\n",
    "\n",
    "🫵 You handle the **implementation**, we handle the **process**!\n",
    "\n",
    "🔨 Built for **scaling** your experiments\n",
    "\n",
    "💡 Defaults for **benchmarking** and **comparison**\n",
    "\n",
    "📘 Extensive **documentation** and **testing**\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "15218d47",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "<img src=./img/Brown_logo.svg width=30%>\n",
    "\n",
    "# F3DASM: Framework for Data-Driven Design & Analysis of Structures & Materials\n",
    "\n",
    "https://github.com/bessagroup/f3dasm\n",
    "\n",
    "### Martin van der Schelling | <a href = \"mailto: martin_van_der_schelling@brown.edu\">martin_van_der_schelling@brown.edu</a>  | PhD candidate"
   ]
  }
 ],
 "metadata": {
  "celltoolbar": "Slideshow",
  "kernelspec": {
   "display_name": "Python [conda env:presentation] *",
   "language": "python",
   "name": "conda-env-presentation-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.17"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}