diff --git a/.DS_Store b/.DS_Store
new file mode 100644
index 0000000..deb57d8
Binary files /dev/null and b/.DS_Store differ
diff --git a/.gitignore b/.gitignore
index 93b1242..c7768a1 100644
--- a/.gitignore
+++ b/.gitignore
@@ -1,4 +1,4 @@
-*.cfg
+
!*template*
@@ -7,13 +7,17 @@
*.gexf.gz
*.h5
*.mat
-*.mp4
*.log
*.png
*.pyc
*.sh
*.swp
*.zip
+.ipynb_checkpoints
+*/.ipynb_checkpoints/*
+
+*.DS_Store
+**/.DS_Store
!examples/image1.mat
diff --git a/examples/photoreceptor_demo/run_photoreceptor_demo.sh b/examples/photoreceptor_demo/run_photoreceptor_demo.sh
deleted file mode 100755
index 0ef08a1..0000000
--- a/examples/photoreceptor_demo/run_photoreceptor_demo.sh
+++ /dev/null
@@ -1,58 +0,0 @@
-#!/bin/bash
-
-usage() { echo "Usage: $0 [-c CONF_FILE] [-r REPEAT] [-h]
- -c configuration file
- -r repeat script execution this number of times
- -h prints this help message
-" 1>&2;}
-
-conf=photoreceptor_default.cfg
-image=image1.mat
-rep=1
-
-while getopts "hc:r:" arg; do
- case "$arg" in
- c)
- conf=${OPTARG}
- ;;
- r)
- rep=${OPTARG}
- ;;
- h)
- usage
- exit
- ;;
- *)
- usage
- exit 1
- ;;
- esac
-done
-
-# add extension if not exists
-if [[ $conf != *.cfg ]]; then
- file=$conf.cfg
-fi
-
-# create default(empty) configuration if file not exists
-if [ ! -f $conf ]; then
- echo 'Configuration file' $conf 'does not exist. Generating ' \
- 'default configuration.'
- touch $conf
-fi
-
-# if image does not exist in current folder make a copy
-if [ ! -f $image ]; then
- echo 'Making a local copy of ' $image
- cp ../$image $image
-fi
-
-echo 'Using' $conf 'configuration file, executing' $rep 'times'
-if (( $rep > 1 )) ; then
- for (( i=0; i<$rep; i++ ))
- do
- python photoreceptor_demo.py -c $conf -v $i
- done
-else
- python photoreceptor_demo.py -c $conf -v -1
-fi
diff --git a/examples/retina_demo/run_retina_demo.sh b/examples/retina_demo/run_retina_demo.sh
deleted file mode 100755
index 68ed50d..0000000
--- a/examples/retina_demo/run_retina_demo.sh
+++ /dev/null
@@ -1,60 +0,0 @@
-#!/bin/bash
-
-usage() { echo "Usage: $0 [-c CONF_FILE] [-r REPEAT] [-h]
- -c configuration file
- -r repeat script execution this number of times
- -h prints this help message
-" 1>&2;}
-
-conf=retina_default.cfg
-image=image1.mat
-rep=1
-
-while getopts "hc:r:" arg; do
- case "$arg" in
- c)
- conf=${OPTARG}
- ;;
- r)
- rep=${OPTARG}
- ;;
- h)
- usage
- exit
- ;;
- *)
- usage
- exit 1
- ;;
- esac
-done
-
-# add extension if not exists
-if [[ $conf != *.cfg ]]; then
- file=$conf.cfg
-fi
-
-# create default(empty) configuration if file not exists
-if [ ! -f $conf ]; then
- echo 'Configuration file' $conf 'does not exist. Generating ' \
- 'default configuration.'
- touch $conf
-fi
-
-# if image does not exist in current folder make a copy
-if [ ! -f $image ]; then
- echo 'Making a local copy of ' $image
- cp ../$image $image
-fi
-
-echo 'Using' $conf 'configuration file, executing' $rep 'times'
-if (( $rep > 1 )) ; then
- for (( i=0; i<$rep; i++ ))
- do
- echo 'Execution:' $((i + 1))
- python retina_demo.py -c $conf -v $i
- done
-else
- python retina_demo.py -c $conf -v -1
-fi
-
diff --git a/examples/retina_demo_RPM/larger_squido.mp4 b/examples/retina_demo_RPM/larger_squido.mp4
new file mode 100644
index 0000000..0d424ca
Binary files /dev/null and b/examples/retina_demo_RPM/larger_squido.mp4 differ
diff --git a/examples/retina_demo_RPM/retina_data_visualization.ipynb b/examples/retina_demo_RPM/retina_data_visualization.ipynb
new file mode 100644
index 0000000..d16505c
--- /dev/null
+++ b/examples/retina_demo_RPM/retina_data_visualization.ipynb
@@ -0,0 +1,802 @@
+{
+ "cells": [
+ {
+ "cell_type": "code",
+ "execution_count": 1,
+ "id": "e135e581-8b3f-4a4b-82b5-0c6a9171e679",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "%matplotlib widget\n",
+ "import numpy as np\n",
+ "import matplotlib.pyplot as plt\n",
+ "from matplotlib import cm\n",
+ "from matplotlib.colors import Normalize\n",
+ "import h5py\n",
+ "import os\n",
+ "import cv2 as cv\n",
+ "import pandas as pd\n",
+ "from mpl_toolkits.mplot3d import Axes3D\n",
+ "import matplotlib.animation\n",
+ "import matplotlib.animation as animation\n",
+ "from matplotlib.animation import FuncAnimation, PillowWriter, MovieWriter"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "3a95127c-f6c6-43f1-8e1d-781a84bf18db",
+ "metadata": {},
+ "source": [
+ "### This file is for the full model, DO NOT MESS UP!!!"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "4fa8a036-499c-4b23-a57b-fbb9e51bd030",
+ "metadata": {},
+ "source": [
+ "#### the H5 file are discribed in the README file of retina_lamina_demo (https://github.com/neurokernel/retina-lamina/tree/master/examples/retlam_demo)\\\n",
+ "#### This is basically the rewriting of visualize_result.m "
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 2,
+ "id": "ef57d3d0-c082-42c9-b2d6-daa1021a9c37",
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "x1 is of the shape(1459,)\n",
+ "x2 is of the shape(800, 50)\n"
+ ]
+ }
+ ],
+ "source": [
+ "# read ommatidia coordinates\n",
+ "elevr1 = h5py.File('retina_elev.h5', 'r')['array']\n",
+ "azim1 = h5py.File('retina_azim.h5', 'r')['array']\n",
+ "r1 = 1\n",
+ "y1 = -r1 * np.cos(elevr1) * np.sin(azim1)\n",
+ "x1 = -r1 * np.cos(elevr1) * np.cos(azim1)\n",
+ "z1 = r1 * np.sin(elevr1)\n",
+ "print(f'x1 is of the shape{np.shape(x1)}') # x1, y1, z1 is of the same dimension of ommatidiates\n",
+ "\n",
+ "# read screen coordinates(or grids more precisely)\n",
+ "elevr2 = h5py.File('grid_dima.h5', 'r')['array']\n",
+ "azim2 = h5py.File('grid_dimb.h5', 'r')['array']\n",
+ "r2 = 10\n",
+ "y2 = -r2 * np.cos(elevr2) * np.sin(azim2)\n",
+ "x2 = -r2 * np.cos(elevr2) * np.cos(azim2)\n",
+ "z2 = r2 * np.sin(elevr2)\n",
+ "print(f'x2 is of the shape{np.shape(x2)}') # x2, y2, z2 is of the same size of input matrix, see below"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 3,
+ "id": "2c97d1ea-6823-4952-b351-8f0cf8805f02",
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "87777.52241897064\n",
+ "0.0\n",
+ "the cutted R1 in/output has shape of (200, 1459)\n",
+ "the screen is of the shape (200, 800, 50)\n"
+ ]
+ }
+ ],
+ "source": [
+ "inputs = h5py.File('retina_input.h5', 'r')['photon']['data']\n",
+ "inputs = np.array(inputs)\n",
+ "print(np.max(inputs))\n",
+ "print(np.min(inputs))\n",
+ "\n",
+ "inputs = inputs[::10, :] # just drop off some frames to make life easier\n",
+ "outputs = h5py.File('retina_output0.h5', 'r')['V']['data']\n",
+ "outputs = np.array(outputs)\n",
+ "outputs = outputs[::10, :]\n",
+ "\n",
+ "R1_inputs = inputs[:, 0:-1:6] # get R1 inputs, since here there is no R7&R8\n",
+ "R1_outputs = outputs[:, 0:-1:6]\n",
+ "print('the cutted R1 in/output has shape of {}'.format(np.shape(R1_inputs)))\n",
+ "\n",
+ "\n",
+ "# intensities is the the screen input, each frame is a 800 * 50 matrix, \n",
+ "# and it has steps/10 frames, due to the performance reason\n",
+ "screen = h5py.File('intensities.h5', 'r')['array']\n",
+ "print(f'the screen is of the shape {np.shape(screen)}')\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 4,
+ "id": "2097a310-25b9-4942-af08-5157cf553bd3",
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "87777.52241897064\n",
+ "0.0\n"
+ ]
+ }
+ ],
+ "source": [
+ "print(np.max(inputs))\n",
+ "print(np.min(inputs))"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 5,
+ "id": "aa4294e2-9ffe-4b5e-93bb-18291233d599",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "# albers projection, see RFC3\n",
+ "def cartisian_to_sphere(x, y, z):\n",
+ " r = np.sqrt(x**2 + y**2 + z**2)\n",
+ " theta = np.arctan2(y,x)\n",
+ " phi = np.arccos(z/r)\n",
+ " return r, theta, phi\n",
+ "\n",
+ "def albers(theta, phi):\n",
+ " \n",
+ " x = np.sign(np.cos(theta)) * np.sqrt((1-np.cos(phi))/(1 + np.tan(theta)**2))\n",
+ " y = x * np.tan(theta)\n",
+ " return x, y\n",
+ "\n",
+ "def albers_projection_cartisian(x, y, z):\n",
+ " r, theta, phi = cartisian_to_sphere(x, z,-y)\n",
+ " x_albers, y_albers = albers(theta, phi)\n",
+ " return x_albers, y_albers\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 6,
+ "id": "c8565922-845c-4eaf-9fef-fadcd679b97c",
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "the min is 5964.667865102264\n",
+ "the max is 45200.387142702064\n"
+ ]
+ }
+ ],
+ "source": [
+ "frame_index = 2\n",
+ "print(f'the min is {np.min(R1_inputs[frame_index])}')\n",
+ "print(f'the max is {np.max(R1_inputs[frame_index])}')\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 7,
+ "id": "af47ffdc-81db-4aaa-9a78-854f19632724",
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "application/vnd.jupyter.widget-view+json": {
+ "model_id": "066b72dd83bf42059797c72abdd68f88",
+ "version_major": 2,
+ "version_minor": 0
+ },
+ "image/png": "iVBORw0KGgoAAAANSUhEUgAAAoAAAAHgCAYAAAA10dzkAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAyeElEQVR4nO3de3RU9b3//9cQkglQMlwiSagxoAe5BS0EycXDTTCABaT1QCgY0UWj9FQFo6uaWhU4q0Zsq/WGiguNeIFUQ6ouLjVYbp4EEAgogoieaBAzRpDMQCtJhM/vD37M12FyI0Bmwuf5WGuv5f7Me3/yea/NmFd2Zu84jDFGAAAAsEabYC8AAAAALYsACAAAYBkCIAAAgGUIgAAAAJYhAAIAAFiGAAgAAGAZAiAAAIBlCIAAAACWIQACAABYhgAIAABgGQIgAACAZQiAAAAAliEAAgAAWIYACAAAYBkCIAAAgGUIgAAAAJYhAAIAAFiGAAgAAGAZAiAAAIBlCIAAAACWIQACAABYhgAIAABgGQIgAACAZQiAAAAAliEAAgAAWIYACAAAYBkCIAAAgGUIgAAAAJYhAAIAAFiGAAgAAGAZAiAAAIBlCIAAAACWIQACAABYhgAIAABgGQIgAACAZQiAAAAAliEAAgAAWIYACAAAYBkCIAAAgGUIgAAAAJYhAAIAAFiGAAgAAGAZAiAAAIBlCIAAAACWIQACAABYhgAIAABgGQIgAACAZQiAAAAAliEAAgAAWIYACAAAYBkCIAAAgGUIgAAAAJZpFQFww4YNmjBhgrp37y6Hw6G///3vjR6zfv16JSUlKTIyUpdeeqmee+65gJqCggL169dPTqdT/fr1U2Fh4XlYPQAAQGhpFQHwX//6l6688ko9/fTTTaovKyvTddddp6FDh6q0tFS///3vdeedd6qgoMBXU1JSooyMDGVmZmrnzp3KzMzUlClTtHnz5vPVBgAAQEhwGGNMsBdxJhwOhwoLCzVp0qR6a+699169/fbb2rNnj29s1qxZ2rlzp0pKSiRJGRkZ8nq9WrVqla9m7Nix6ty5s5YuXXre1g8AABBsreIK4JkqKSlRenq639iYMWO0detW1dbWNlhTXFzcYusEAAAIhrbBXsD54Ha7FRMT4zcWExOjH374QQcPHlRcXFy9NW63u955q6urVV1d7ds/ceKEvvvuO3Xt2lUOh+PcNgEAAM4LY4yOHDmi7t27q02bC/JaWKMuyAAoKSCQnfpN94/H66ppKMjl5uZq3rx553CVAAAgWPbv36+LL7442MsIigsyAMbGxgZcyausrFTbtm3VtWvXBmtOvyr4Yzk5OcrOzvbtezweXXLJJdq/f7+ioqLOYQcAAOB88Xq9io+PV8eOHYO9lKC5IANgamqq3nnnHb+xd999V4MHD1Z4eLivpqioSHfddZdfTVpaWr3zOp1OOZ3OgPGoqCgCIAAArYzNH99qFQHw6NGj+uyzz3z7ZWVl2rFjh7p06aJLLrlEOTk5OnDggJYsWSLp5B2/Tz/9tLKzs5WVlaWSkhItXrzY7+7e2bNna9iwYVqwYIGuv/56vfXWW1qzZo3ef//9Fu8PAACgJbWKTz5u3bpVAwcO1MCBAyVJ2dnZGjhwoB588EFJUkVFhcrLy331PXv21MqVK7Vu3Tr97Gc/0//8z//oySef1A033OCrSUtL07Jly/TSSy/piiuuUF5envLz85WcnNyyzQEAALSwVvccwFDi9Xrlcrnk8Xj4FTAAAK0E379byRVAAAAAnDsEQAAAAMsQAAEAACxDAAQAALAMARAAAMAyBEAAAADLEAABAAAsQwAEAACwDAEQAADAMgRAAAAAyxAAAQAALEMABAAAsAwBEAAAwDIEQAAAAMsQAAEAACxDAAQAALAMARAAAMAyBEAAAADLEAABAAAsQwAEAACwDAEQAADAMgRAAAAAyxAAAQAALEMABAAAsAwBEAAAwDIEQAAAAMsQAAEAACxDAAQAALAMARAAAMAyBEAAAADLEAABAAAsQwAEAACwDAEQAADAMgRAAAAAyxAAAQAALNNqAuDChQvVs2dPRUZGKikpSRs3bqy39uabb5bD4QjY+vfv76vJy8urs+bYsWMt0Q4AAEDQtIoAmJ+frzlz5uj+++9XaWmphg4dqnHjxqm8vLzO+ieeeEIVFRW+bf/+/erSpYsmT57sVxcVFeVXV1FRocjIyJZoCQAAIGhaRQB87LHHNHPmTP36179W37599de//lXx8fF69tln66x3uVyKjY31bVu3btXhw4d1yy23+NU5HA6/utjY2JZoBwAAIKhCPgDW1NRo27ZtSk9P9xtPT09XcXFxk+ZYvHixRo8erYSEBL/xo0ePKiEhQRdffLHGjx+v0tLSc7ZuAACAUNU22AtozMGDB3X8+HHFxMT4jcfExMjtdjd6fEVFhVatWqXXX3/db7xPnz7Ky8vTgAED5PV69cQTT+jqq6/Wzp071atXrzrnqq6uVnV1tW/f6/U2oyMAAIDgCvkrgKc4HA6/fWNMwFhd8vLy1KlTJ02aNMlvPCUlRTfeeKOuvPJKDR06VH/72990+eWX66mnnqp3rtzcXLlcLt8WHx/frF4AAACCKeQDYHR0tMLCwgKu9lVWVgZcFTydMUYvvviiMjMzFRER0WBtmzZtdNVVV2nfvn311uTk5Mjj8fi2/fv3N70RAACAEBHyATAiIkJJSUkqKiryGy8qKlJaWlqDx65fv16fffaZZs6c2ejXMcZox44diouLq7fG6XQqKirKbwMAAGhtQv4zgJKUnZ2tzMxMDR48WKmpqVq0aJHKy8s1a9YsSSevzB04cEBLlizxO27x4sVKTk5WYmJiwJzz5s1TSkqKevXqJa/XqyeffFI7duzQM8880yI9AQAABEurCIAZGRk6dOiQ5s+fr4qKCiUmJmrlypW+u3orKioCngno8XhUUFCgJ554os45q6qqdOutt8rtdsvlcmngwIHasGGDhgwZct77AQAACCaHMcYEexGtldfrlcvlksfj4dfBAAC0Enz/bgWfAQQAAMC5RQAEAACwDAEQAADAMgRAAAAAyxAAAQAALEMABAAAsAwBEAAAwDIEQAAAAMsQAAEAACxDAAQAALAMARAAAMAyBEAAAADLEAABAAAsQwAEAACwDAEQAADAMgRAAAAAyxAAAQAALEMABAAAsAwBEAAAwDIEQAAAAMsQAAEAACxDAAQAALAMARAAAMAyBEAAAADLEAABAAAsQwAEAACwDAEQAADAMgRAAAAAyxAAAQAALEMABAAAsAwBEAAAwDIEQAAAAMsQAAEAACxDAAQAALAMARAAAMAyBEAAAADLtJoAuHDhQvXs2VORkZFKSkrSxo0b661dt26dHA5HwPbJJ5/41RUUFKhfv35yOp3q16+fCgsLz3cbAAAAQdcqAmB+fr7mzJmj+++/X6WlpRo6dKjGjRun8vLyBo/bu3evKioqfFuvXr18r5WUlCgjI0OZmZnauXOnMjMzNWXKFG3evPl8twMAABBUDmOMCfYiGpOcnKxBgwbp2Wef9Y317dtXkyZNUm5ubkD9unXrNHLkSB0+fFidOnWqc86MjAx5vV6tWrXKNzZ27Fh17txZS5cubdK6vF6vXC6XPB6PoqKizqwpAAAQFHz/bgVXAGtqarRt2zalp6f7jaenp6u4uLjBYwcOHKi4uDiNGjVKa9eu9XutpKQkYM4xY8Y0OGd1dbW8Xq/fBgAA0NqEfAA8ePCgjh8/rpiYGL/xmJgYud3uOo+Ji4vTokWLVFBQoOXLl6t3794aNWqUNmzY4Ktxu91nNKck5ebmyuVy+bb4+Piz6AwAACA42gZ7AU3lcDj89o0xAWOn9O7dW7179/btp6amav/+/frzn/+sYcOGNWtOScrJyVF2drZv3+v1EgIBAECrE/JXAKOjoxUWFhZwZa6ysjLgCl5DUlJStG/fPt9+bGzsGc/pdDoVFRXltwEAALQ2IR8AIyIilJSUpKKiIr/xoqIipaWlNXme0tJSxcXF+fZTU1MD5nz33XfPaE4AAIDWqFX8Cjg7O1uZmZkaPHiwUlNTtWjRIpWXl2vWrFmSTv5q9sCBA1qyZIkk6a9//at69Oih/v37q6amRq+++qoKCgpUUFDgm3P27NkaNmyYFixYoOuvv15vvfWW1qxZo/fffz8oPQIAALSUVhEAMzIydOjQIc2fP18VFRVKTEzUypUrlZCQIEmqqKjweyZgTU2N7rnnHh04cEDt2rVT//79tWLFCl133XW+mrS0NC1btkx/+MMf9MADD+iyyy5Tfn6+kpOTW7w/AACAltQqngMYqniOEAAArQ/fv1vBZwABAABwbhEAAQAALEMABAAAsAwBEAAAwDIEQAAAAMsQAAEAACxDAAQAALAMARAAAMAyBEAAAADLEAABAAAsQwAEAACwDAEQAADAMgRAAAAAyxAAAQAALEMABAAAsAwBEAAAwDIEQAAAAMsQAAEAACxDAAQAALAMARAAAMAyBEAAAADLEAABAAAsQwAEAACwDAEQAADAMgRAAAAAyxAAAQAALEMABAAAsAwBEAAAwDIEQAAAAMsQAAEAACxDAAQAALAMARAAAMAyBEAAAADLEAABAAAsQwAEAACwTKsJgAsXLlTPnj0VGRmppKQkbdy4sd7a5cuX69prr9VFF12kqKgopaam6h//+IdfTV5enhwOR8B27Nix890KAABAULWKAJifn685c+bo/vvvV2lpqYYOHapx48apvLy8zvoNGzbo2muv1cqVK7Vt2zaNHDlSEyZMUGlpqV9dVFSUKioq/LbIyMiWaAkAACBoHMYYE+xFNCY5OVmDBg3Ss88+6xvr27evJk2apNzc3CbN0b9/f2VkZOjBBx+UdPIK4Jw5c1RVVdXsdXm9XrlcLnk8HkVFRTV7HgAA0HL4/t0KrgDW1NRo27ZtSk9P9xtPT09XcXFxk+Y4ceKEjhw5oi5duviNHz16VAkJCbr44os1fvz4gCuEp6uurpbX6/XbAAAAWpuQD4AHDx7U8ePHFRMT4zceExMjt9vdpDn+8pe/6F//+pemTJniG+vTp4/y8vL09ttva+nSpYqMjNTVV1+tffv21TtPbm6uXC6Xb4uPj29eUwAAAEEU8gHwFIfD4bdvjAkYq8vSpUs1d+5c5efnq1u3br7xlJQU3Xjjjbryyis1dOhQ/e1vf9Pll1+up556qt65cnJy5PF4fNv+/fub3xAAAECQtA32AhoTHR2tsLCwgKt9lZWVAVcFT5efn6+ZM2fqjTfe0OjRoxusbdOmja666qoGrwA6nU45nc6mLx4AACAEhfwVwIiICCUlJamoqMhvvKioSGlpafUet3TpUt188816/fXX9fOf/7zRr2OM0Y4dOxQXF3fWawYAAAhlIX8FUJKys7OVmZmpwYMHKzU1VYsWLVJ5eblmzZol6eSvZg8cOKAlS5ZIOhn+brrpJj3xxBNKSUnxXT1s166dXC6XJGnevHlKSUlRr1695PV69eSTT2rHjh165plngtMkAABAC2kVATAjI0OHDh3S/PnzVVFRocTERK1cuVIJCQmSpIqKCr9nAj7//PP64Ycf9Nvf/la//e1vfeMzZsxQXl6eJKmqqkq33nqr3G63XC6XBg4cqA0bNmjIkCEt2hsAAEBLaxXPAQxVPEcIAIDWh+/freAzgAAAADi3CIAAAACWIQACAABYhgAIAABgGQIgAACAZQiAAAAAliEAAgAAWIYACAAAYBkCIAAAgGUIgAAAAJYhAAIAAFiGAAgAAGAZAiAAAIBlCIAAAACWIQACAABYhgAIAABgGQIgAACAZQiAAAAAliEAAgAAWIYACAAAYBkCIAAAgGUIgAAAAJYhAAIAAFiGAAgAAGAZAiAAAIBlCIAAAACWIQACAABYhgAIAABgGQIgAACAZQiAAAAAliEAAgAAWIYACAAAYBkCIAAAgGUIgAAAAJYhAAIAAFim1QTAhQsXqmfPnoqMjFRSUpI2btzYYP369euVlJSkyMhIXXrppXruuecCagoKCtSvXz85nU7169dPhYWF52v5AAAAIaNVBMD8/HzNmTNH999/v0pLSzV06FCNGzdO5eXlddaXlZXpuuuu09ChQ1VaWqrf//73uvPOO1VQUOCrKSkpUUZGhjIzM7Vz505lZmZqypQp2rx5c0u1BQAAEBQOY4wJ9iIak5ycrEGDBunZZ5/1jfXt21eTJk1Sbm5uQP29996rt99+W3v27PGNzZo1Szt37lRJSYkkKSMjQ16vV6tWrfLVjB07Vp07d9bSpUubtC6v1yuXyyWPx6OoqKjmtgcAAFoQ379bwRXAmpoabdu2Tenp6X7j6enpKi4urvOYkpKSgPoxY8Zo69atqq2tbbCmvjklqbq6Wl6v128DAABobUI+AB48eFDHjx9XTEyM33hMTIzcbnedx7jd7jrrf/jhBx08eLDBmvrmlKTc3Fy5XC7fFh8f35yWAAAAgirkA+ApDofDb98YEzDWWP3p42c6Z05Ojjwej2/bv39/k9cPAAAQKtoGewGNiY6OVlhYWMCVucrKyoAreKfExsbWWd+2bVt17dq1wZr65pQkp9Mpp9PZnDYAAABCRshfAYyIiFBSUpKKior8xouKipSWllbnMampqQH17777rgYPHqzw8PAGa+qbEwAA4EIR8lcAJSk7O1uZmZkaPHiwUlNTtWjRIpWXl2vWrFmSTv5q9sCBA1qyZImkk3f8Pv3008rOzlZWVpZKSkq0ePFiv7t7Z8+erWHDhmnBggW6/vrr9dZbb2nNmjV6//33g9IjAABAS2kVATAjI0OHDh3S/PnzVVFRocTERK1cuVIJCQmSpIqKCr9nAvbs2VMrV67UXXfdpWeeeUbdu3fXk08+qRtuuMFXk5aWpmXLlukPf/iDHnjgAV122WXKz89XcnJyi/cHAADQklrFcwBDFc8RAgCg9eH7dyv4DCAAAADOLQIgAACAZQiAAAAAliEAAgAAWIYACAAAYBkCIAAAgGUIgAAAAJYhAAIAAFiGAAgAAGAZAiAAAIBlCIAAAACWIQACAABYhgAIAABgGQIgAACAZQiAAAAAliEAAgAAWIYACAAAYBkCIAAAgGUIgAAAAJYhAAIAAFiGAAgAAGAZAiAAAIBlCIAAAACWIQACAABYhgAIAABgGQIgAACAZQiAAAAAliEAAgAAWIYACAAAYBkCIAAAgGUIgAAAAJYhAAIAAFiGAAgAAGAZAiAAAIBlCIAAAACWCfkAePjwYWVmZsrlcsnlcikzM1NVVVX11tfW1uree+/VgAED1KFDB3Xv3l033XSTvv76a7+6ESNGyOFw+G1Tp049z90AAAAEX8gHwGnTpmnHjh1avXq1Vq9erR07digzM7Pe+n//+9/avn27HnjgAW3fvl3Lly/Xp59+qokTJwbUZmVlqaKiwrc9//zz57MVAACAkNA22AtoyJ49e7R69Wpt2rRJycnJkqQXXnhBqamp2rt3r3r37h1wjMvlUlFRkd/YU089pSFDhqi8vFyXXHKJb7x9+/aKjY09v00AAACEmJC+AlhSUiKXy+ULf5KUkpIil8ul4uLiJs/j8XjkcDjUqVMnv/HXXntN0dHR6t+/v+655x4dOXLkXC0dAAAgZIX0FUC3261u3boFjHfr1k1ut7tJcxw7dkz33Xefpk2bpqioKN/49OnT1bNnT8XGxmrXrl3KycnRzp07A64e/lh1dbWqq6t9+16v9wy6AQAACA1BuQI4d+7cgBswTt+2bt0qSXI4HAHHG2PqHD9dbW2tpk6dqhMnTmjhwoV+r2VlZWn06NFKTEzU1KlT9eabb2rNmjXavn17vfPl5ub6bkZxuVyKj48/w84BAACCLyhXAG+//fZG77jt0aOHPvzwQ33zzTcBr3377beKiYlp8Pja2lpNmTJFZWVl+uc//+l39a8ugwYNUnh4uPbt26dBgwbVWZOTk6Ps7GzfvtfrJQQCAIBWJygBMDo6WtHR0Y3WpaamyuPxaMuWLRoyZIgkafPmzfJ4PEpLS6v3uFPhb9++fVq7dq26du3a6Nf6+OOPVVtbq7i4uHprnE6nnE5no3MBAACEspC+CaRv374aO3assrKytGnTJm3atElZWVkaP3683x3Affr0UWFhoSTphx9+0H/9139p69ateu2113T8+HG53W653W7V1NRIkj7//HPNnz9fW7du1RdffKGVK1dq8uTJGjhwoK6++uqg9AoAANBSQjoASifv1B0wYIDS09OVnp6uK664Qq+88opfzd69e+XxeCRJX331ld5++2199dVX+tnPfqa4uDjfdurO4YiICL333nsaM2aMevfurTvvvFPp6elas2aNwsLCWrxHAACAluQwxphgL6K18nq9crlc8ng8jX7GEAAAhAa+f7eCK4AAAAA4twiAAAAAliEAAgAAWIYACAAAYBkCIAAAgGUIgAAAAJYhAAIAAFiGAAgAAGAZAiAAAIBlCIAAAACWIQACAABYhgAIAABgGQIgAACAZQiAAAAAliEAAgAAWIYACAAAYBkCIAAAgGUIgAAAAJZpG+wFAAAuDMdPGG0p+06VR46pW8dIDenZRWFtHMFeFoA6EAABAGdt9a4KzXtntyo8x3xjca5IPTShn8YmxgVxZQDqwq+AAQBnZfWuCv3m1e1+4U+S3J5j+s2r27V6V0WQVgagPgRAAECzHT9hNO+d3TJ1vHZqbN47u3X8RF0VAIKFAAgAaLYtZd8FXPn7MSOpwnNMW8q+a7lFAWgUARAA0GyVR+oPf82pA9AyCIAAgGbr1jHynNYBaBkEQABAsw3p2UVxrkjV97AXh07eDTykZ5eWXBaARhAAAQDNFtbGoYcm9JOkgBB4av+hCf14HiAQYgiAAICzMjYxTs/eOEixLv9f88a6IvXsjYN4DiAQgngQNADgrI1NjNO1/WL5SyBAK0EABACcE2FtHEq9rGuwlwGgCfgVMAAAgGUIgAAAAJYhAAIAAFiGAAgAAGAZAiAAAIBlQj4AHj58WJmZmXK5XHK5XMrMzFRVVVWDx9x8881yOBx+W0pKil9NdXW17rjjDkVHR6tDhw6aOHGivvrqq/PYCQAAQGgI+QA4bdo07dixQ6tXr9bq1au1Y8cOZWZmNnrc2LFjVVFR4dtWrlzp9/qcOXNUWFioZcuW6f3339fRo0c1fvx4HT9+/Hy1AgAAEBJC+jmAe/bs0erVq7Vp0yYlJydLkl544QWlpqZq79696t27d73HOp1OxcbG1vmax+PR4sWL9corr2j06NGSpFdffVXx8fFas2aNxowZc+6bAQAACBEhfQWwpKRELpfLF/4kKSUlRS6XS8XFxQ0eu27dOnXr1k2XX365srKyVFlZ6Xtt27Ztqq2tVXp6um+se/fuSkxMbHReAACA1i6krwC63W5169YtYLxbt25yu931Hjdu3DhNnjxZCQkJKisr0wMPPKBrrrlG27Ztk9PplNvtVkREhDp37ux3XExMTIPzVldXq7q62rfv9Xqb0RUAAEBwBeUK4Ny5cwNu0jh927p1qyTJ4Qj8O5LGmDrHT8nIyNDPf/5zJSYmasKECVq1apU+/fRTrVixosF1NTZvbm6u72YUl8ul+Pj4JnYMAAAQOoJyBfD222/X1KlTG6zp0aOHPvzwQ33zzTcBr3377beKiYlp8teLi4tTQkKC9u3bJ0mKjY1VTU2NDh8+7HcVsLKyUmlpafXOk5OTo+zsbN++1+slBAIAgFYnKAEwOjpa0dHRjdalpqbK4/Foy5YtGjJkiCRp8+bN8ng8DQa10x06dEj79+9XXFycJCkpKUnh4eEqKirSlClTJEkVFRXatWuXHn300XrncTqdcjqdTf66AAAAoSikbwLp27evxo4dq6ysLG3atEmbNm1SVlaWxo8f73cHcJ8+fVRYWChJOnr0qO655x6VlJToiy++0Lp16zRhwgRFR0frF7/4hSTJ5XJp5syZuvvuu/Xee++ptLRUN954owYMGOC7KxgAAOBCFdI3gUjSa6+9pjvvvNN3x+7EiRP19NNP+9Xs3btXHo9HkhQWFqaPPvpIS5YsUVVVleLi4jRy5Ejl5+erY8eOvmMef/xxtW3bVlOmTNH333+vUaNGKS8vT2FhYS3XHAAAQBA4jDEm2Itorbxer1wulzwej6KiooK9HAAA0AR8/w7xXwEDAADg3CMAAgAAWIYACAAAYBkCIAAAgGUIgAAAAJYhAAIAAFiGAAgAAGAZAiAAAIBlCIAAAACWIQACAABYhgAIAABgGQIgAACAZQiAAAAAliEAAgAAWIYACAAAYBkCIAAAgGUIgAAAAJYhAAIAAFiGAAgAAGAZAiAAAIBlCIAAAACWIQACAABYhgAIAABgGQIgAACAZQiAAAAAliEAAgAAWIYACAAAYBkCIAAAgGUIgAAAAJYhAAIAAFiGAAgAAGAZAiAAAIBlCIAAAACWIQACAABYhgAIAABgmZAPgIcPH1ZmZqZcLpdcLpcyMzNVVVXV4DEOh6PO7U9/+pOvZsSIEQGvT5069Tx3AwAAEHxtg72AxkybNk1fffWVVq9eLUm69dZblZmZqXfeeafeYyoqKvz2V61apZkzZ+qGG27wG8/KytL8+fN9++3atTuHKwcAAAhNIR0A9+zZo9WrV2vTpk1KTk6WJL3wwgtKTU3V3r171bt37zqPi42N9dt/6623NHLkSF166aV+4+3btw+oBQAAuNCF9K+AS0pK5HK5fOFPklJSUuRyuVRcXNykOb755hutWLFCM2fODHjttddeU3R0tPr376977rlHR44cOWdrBwAACFUhfQXQ7XarW7duAePdunWT2+1u0hwvv/yyOnbsqF/+8pd+49OnT1fPnj0VGxurXbt2KScnRzt37lRRUVG9c1VXV6u6utq37/V6m9gJAABA6AjKFcC5c+fWe6PGqW3r1q2STt7QcTpjTJ3jdXnxxRc1ffp0RUZG+o1nZWVp9OjRSkxM1NSpU/Xmm29qzZo12r59e71z5ebm+m5Gcblcio+PP4OuAQAAQkNQrgDefvvtjd5x26NHD3344Yf65ptvAl779ttvFRMT0+jX2bhxo/bu3av8/PxGawcNGqTw8HDt27dPgwYNqrMmJydH2dnZvn2v10sIBAAArU5QAmB0dLSio6MbrUtNTZXH49GWLVs0ZMgQSdLmzZvl8XiUlpbW6PGLFy9WUlKSrrzyykZrP/74Y9XW1iouLq7eGqfTKafT2ehcAAAAoSykbwLp27evxo4dq6ysLG3atEmbNm1SVlaWxo8f73cHcJ8+fVRYWOh3rNfr1RtvvKFf//rXAfN+/vnnmj9/vrZu3aovvvhCK1eu1OTJkzVw4EBdffXV570vAACAYArpACidvFN3wIABSk9PV3p6uq644gq98sorfjV79+6Vx+PxG1u2bJmMMfrVr34VMGdERITee+89jRkzRr1799add96p9PR0rVmzRmFhYee1HwAAgGBzGGNMsBfRWnm9XrlcLnk8HkVFRQV7OQAAoAn4/t0KrgACAADg3CIAAgAAWIYACAAAYBkCIAAAgGUIgAAAAJYhAAIAAFiGAAgAAGAZAiAAAIBlCIAAAACWIQACAABYhgAIAABgGQIgAACAZQiAAAAAliEAAgAAWIYACAAAYBkCIAAAgGUIgAAAAJYhAAIAAFiGAAgAAGAZAiAAAIBlCIAAAACWIQACAABYhgAIAABgGQIgAACAZQiAAAAAliEAAgAAWIYACAAAYBkCIAAAgGUIgAAAAJYhAAIAAFiGAAgAAGAZAiAAAIBlCIAAAACWIQACAABYhgAIAABgGQIgAACAZUI+AP7xj39UWlqa2rdvr06dOjXpGGOM5s6dq+7du6tdu3YaMWKEPv74Y7+a6upq3XHHHYqOjlaHDh00ceJEffXVV+ehAwAAgNAS8gGwpqZGkydP1m9+85smH/Poo4/qscce09NPP60PPvhAsbGxuvbaa3XkyBFfzZw5c1RYWKhly5bp/fff19GjRzV+/HgdP378fLQBAAAQMhzGGBPsRTRFXl6e5syZo6qqqgbrjDHq3r275syZo3vvvVfSyat9MTExWrBggW677TZ5PB5ddNFFeuWVV5SRkSFJ+vrrrxUfH6+VK1dqzJgxTVqT1+uVy+WSx+NRVFTUWfUHAABaBt+/pbbBXsC5VlZWJrfbrfT0dN+Y0+nU8OHDVVxcrNtuu03btm1TbW2tX0337t2VmJio4uLiegNgdXW1qqurffsej0fSyX9IAACgdTj1fbuVXAM7Ly64AOh2uyVJMTExfuMxMTH68ssvfTURERHq3LlzQM2p4+uSm5urefPmBYzHx8ef7bIBAEALO3TokFwuV7CXERRBCYBz586tM0j92AcffKDBgwc3+2s4HA6/fWNMwNjpGqvJyclRdna2b7+qqkoJCQkqLy+36h+Q1+tVfHy89u/fb9Wlc/qmbxvQN33bwOPx6JJLLlGXLl2CvZSgCUoAvP322zV16tQGa3r06NGsuWNjYyWdvMoXFxfnG6+srPRdFYyNjVVNTY0OHz7sdxWwsrJSaWlp9c7tdDrldDoDxl0ul1VvnFOioqLo2yL0bRf6toutfbdpE/L3wp43QQmA0dHRio6OPi9z9+zZU7GxsSoqKtLAgQMlnbyTeP369VqwYIEkKSkpSeHh4SoqKtKUKVMkSRUVFdq1a5ceffTR87IuAACAUBHynwEsLy/Xd999p/Lych0/flw7duyQJP3Hf/yHfvKTn0iS+vTpo9zcXP3iF7+Qw+HQnDlz9PDDD6tXr17q1auXHn74YbVv317Tpk2TdPKK3cyZM3X33Xera9eu6tKli+655x4NGDBAo0ePDlarAAAALSLkA+CDDz6ol19+2bd/6qre2rVrNWLECEnS3r17fXfkStLvfvc7ff/99/rv//5vHT58WMnJyXr33XfVsWNHX83jjz+utm3basqUKfr+++81atQo5eXlKSwsrMlrczqdeuihh+r8tfCFjL7p2wb0Td82oG+7+v6xVvMcQAAAAJwb9n76EQAAwFIEQAAAAMsQAAEAACxDAAQAALAMAbABf/zjH5WWlqb27durU6dOTTrGGKO5c+eqe/fuateunUaMGKGPP/7Yr6a6ulp33HGHoqOj1aFDB02cOFFfffXVeeigeQ4fPqzMzEy5XC65XC5lZmaqqqqqwWMcDked25/+9CdfzYgRIwJeb+yB4C2pOX3ffPPNAT2lpKT41Vxo57u2tlb33nuvBgwYoA4dOqh79+666aab9PXXX/vVhdr5XrhwoXr27KnIyEglJSVp48aNDdavX79eSUlJioyM1KWXXqrnnnsuoKagoED9+vWT0+lUv379VFhYeL6W32xn0vfy5ct17bXX6qKLLlJUVJRSU1P1j3/8w68mLy+vzvf6sWPHzncrZ+RM+l63bl2dPX3yySd+dRfa+a7r/18Oh0P9+/f31bSG871hwwZNmDBB3bt3l8Ph0N///vdGj7lQ3t9nxaBeDz74oHnsscdMdna2cblcTTrmkUceMR07djQFBQXmo48+MhkZGSYuLs54vV5fzaxZs8xPf/pTU1RUZLZv325GjhxprrzySvPDDz+cp07OzNixY01iYqIpLi42xcXFJjEx0YwfP77BYyoqKvy2F1980TgcDvP555/7aoYPH26ysrL86qqqqs53O03WnL5nzJhhxo4d69fToUOH/GoutPNdVVVlRo8ebfLz880nn3xiSkpKTHJysklKSvKrC6XzvWzZMhMeHm5eeOEFs3v3bjN79mzToUMH8+WXX9ZZ/3//93+mffv2Zvbs2Wb37t3mhRdeMOHh4ebNN9/01RQXF5uwsDDz8MMPmz179piHH37YtG3b1mzatKml2mrUmfY9e/Zss2DBArNlyxbz6aefmpycHBMeHm62b9/uq3nppZdMVFRUwHs+lJxp32vXrjWSzN69e/16+vF79EI831VVVX797t+/33Tp0sU89NBDvprWcL5Xrlxp7r//flNQUGAkmcLCwgbrL5T399kiADbBSy+91KQAeOLECRMbG2seeeQR39ixY8eMy+Uyzz33nDHm5BsuPDzcLFu2zFdz4MAB06ZNG7N69epzvvYztXv3biPJ7x95SUmJkWQ++eSTJs9z/fXXm2uuucZvbPjw4Wb27NnnaqnnVHP7njFjhrn++uvrfd2W871lyxYjye8bTSid7yFDhphZs2b5jfXp08fcd999ddb/7ne/M3369PEbu+2220xKSopvf8qUKWbs2LF+NWPGjDFTp049R6s+e2fad1369etn5s2b59tv6v8Pg+lM+z4VAA8fPlzvnDac78LCQuNwOMwXX3zhG2sN5/vHmhIAL5T399niV8DnUFlZmdxut9LT031jTqdTw4cPV3FxsSRp27Ztqq2t9avp3r27EhMTfTXBVFJSIpfLpeTkZN9YSkqKXC5Xk9f3zTffaMWKFZo5c2bAa6+99pqio6PVv39/3XPPPTpy5Mg5W/vZOJu+161bp27duunyyy9XVlaWKisrfa/ZcL6lk39Y3eFwBHxUIhTOd01NjbZt2+Z3DiQpPT293h5LSkoC6seMGaOtW7eqtra2wZpQOK9S8/o+3YkTJ3TkyBF16dLFb/zo0aNKSEjQxRdfrPHjx6u0tPScrftsnU3fAwcOVFxcnEaNGqW1a9f6vWbD+V68eLFGjx6thIQEv/FQPt/NcSG8v8+FkP9LIK2J2+2WJMXExPiNx8TE6Msvv/TVREREqHPnzgE1p44PJrfbrW7dugWMd+vWrcnre/nll9WxY0f98pe/9BufPn26728179q1Szk5Odq5c6eKiorOydrPRnP7HjdunCZPnqyEhASVlZXpgQce0DXXXKNt27bJ6XRacb6PHTum++67T9OmTfP7Y/Khcr4PHjyo48eP1/m+rK9Ht9tdZ/0PP/yggwcPKi4urt6aUDivUvP6Pt1f/vIX/etf//L9zXTp5J/ezMvL04ABA+T1evXEE0/o6quv1s6dO9WrV69z2kNzNKfvuLg4LVq0SElJSaqurtYrr7yiUaNGad26dRo2bJik+v9NXCjnu6KiQqtWrdLrr7/uNx7q57s5LoT397lgXQCcO3eu5s2b12DNBx98oMGDBzf7azgcDr99Y0zA2OmaUnM2mtq3FLh+6czW9+KLL2r69OmKjIz0G8/KyvL9d2Jionr16qXBgwdr+/btGjRoUJPmPlPnu++MjAzffycmJmrw4MFKSEjQihUrAgLwmcx7tlrqfNfW1mrq1Kk6ceKEFi5c6PdaMM53Q870fVlX/enjzXmvt7TmrnHp0qWaO3eu3nrrLb8fElJSUvxudLr66qs1aNAgPfXUU3ryySfP3cLP0pn03bt3b/Xu3du3n5qaqv379+vPf/6zLwCe6ZzB0tw15uXlqVOnTpo0aZLfeGs532fqQnl/nw3rAuDtt9/e6J2IPXr0aNbcsbGxkk7+dBEXF+cbr6ys9P0kERsbq5qaGh0+fNjvqlBlZaXS0tKa9XWboql9f/jhh/rmm28CXvv2228Dfhqqy8aNG7V3717l5+c3Wjto0CCFh4dr37595y0QtFTfp8TFxSkhIUH79u2TdGGf79raWk2ZMkVlZWX65z//6Xf1ry4tcb7rEh0drbCwsICf3H/8vjxdbGxsnfVt27ZV165dG6w5k38v51Nz+j4lPz9fM2fO1BtvvKHRo0c3WNumTRtdddVVvn/zwXY2ff9YSkqKXn31Vd/+hXy+jTF68cUXlZmZqYiIiAZrQ+18N8eF8P4+J1r+Y4etz5neBLJgwQLfWHV1dZ03geTn5/tqvv7665C7KWDz5s2+sU2bNjX5poAZM2YE3A1an48++shIMuvXr2/2es+Vs+37lIMHDxqn02lefvllY8yFe75ramrMpEmTTP/+/U1lZWWTvlYwz/eQIUPMb37zG7+xvn37NngTSN++ff3GZs2aFfAh8XHjxvnVjB07NqQ+JH6mfRtjzOuvv24iIyMb/SD9KSdOnDCDBw82t9xyy9ks9ZxqTt+nu+GGG8zIkSN9+xfq+Tbm/90E89FHHzX6NULxfP+YmngTyIXw/j5bBMAGfPnll6a0tNTMmzfP/OQnPzGlpaWmtLTUHDlyxFfTu3dvs3z5ct/+I488Ylwul1m+fLn56KOPzK9+9as6HwNz8cUXmzVr1pjt27eba665JuQeC3LFFVeYkpISU1JSYgYMGBDwWJDT+zbGGI/HY9q3b2+effbZgDk/++wzM2/ePPPBBx+YsrIys2LFCtOnTx8zcODAVtv3kSNHzN13322Ki4tNWVmZWbt2rUlNTTU//elPL+jzXVtbayZOnGguvvhis2PHDr9HQ1RXVxtjQu98n3o8xuLFi83u3bvNnDlzTIcOHXx3O953330mMzPTV3/qMRF33XWX2b17t1m8eHHAYyL+93//14SFhZlHHnnE7NmzxzzyyCMh95iIM+379ddfN23btjXPPPNMvY/vmTt3rlm9erX5/PPPTWlpqbnllltM27Zt/X6ICLYz7fvxxx83hYWF5tNPPzW7du0y9913n5FkCgoKfDUX4vk+5cYbbzTJycl1ztkazveRI0d8358lmccee8yUlpb6nkpwob6/zxYBsAEzZswwkgK2tWvX+mokmZdeesm3f+LECfPQQw+Z2NhY43Q6zbBhwwJ+qvr+++/N7bffbrp06WLatWtnxo8fb8rLy1uoq8YdOnTITJ8+3XTs2NF07NjRTJ8+PeDxCKf3bYwxzz//vGnXrl2dz3orLy83w4YNM126dDERERHmsssuM3feeWfAM/OC6Uz7/ve//23S09PNRRddZMLDw80ll1xiZsyYEXAuL7TzXVZWVuf74sfvjVA8388884xJSEgwERERZtCgQX5XImfMmGGGDx/uV79u3TozcOBAExERYXr06FHnDzZvvPGG6d27twkPDzd9+vTxCwyh4kz6Hj58eJ3ndcaMGb6aOXPmmEsuucRERESYiy66yKSnp5vi4uIW7KhpzqTvBQsWmMsuu8xERkaazp07m//8z/80K1asCJjzQjvfxpz8LUW7du3MokWL6pyvNZzvU1cw6/t3eyG/v8+Gw5j//5OPAAAAsALPAQQAALAMARAAAMAyBEAAAADLEAABAAAsQwAEAACwDAEQAADAMgRAAAAAyxAAAQAALEMABAAAsAwBEAAAwDIEQAAAAMsQAAEAACxDAAQAALAMARAAAMAyBEAAAADLEAABAAAsQwAEAACwDAEQAADAMgRAAAAAyxAAAQAALEMABAAAsAwBEAAAwDIEQAAAAMsQAAEAACxDAAQAALAMARAAAMAyBEAAAADLEAABAAAsQwAEAACwDAEQAADAMgRAAAAAy/x/+ImufA0EH1wAAAAASUVORK5CYII=",
+ "text/html": [
+ "\n",
+ " \n",
+ "
\n",
+ " Figure\n",
+ "
\n",
+ "
\n",
+ "
\n",
+ "
\n",
+ " Figure\n",
+ "
\n",
+ "
\n",
+ "