7.model_trackster_graph_net.link.prediction.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "6c07744a",
   "metadata": {},
   "source": [
    "# Graph net for link prediction\n",
    "\n",
    "Question: does including the neighborhood provide a benefit over the pairwise classification?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "e3962d51",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import torch\n",
    "import sys\n",
    "\n",
    "import torch.nn as nn\n",
    "from torch.optim import Adam\n",
    "from torch.optim.lr_scheduler import CosineAnnealingLR\n",
    "\n",
    "from torch.utils.data import random_split\n",
    "from torch_geometric.loader import DataLoader\n",
    "\n",
    "import torch_geometric.utils as geou\n",
    "\n",
    "import torch_geometric.nn as gnn\n",
    "\n",
    "from reco.model import EdgeConvBlock, DynamicEdgeConvBlock\n",
    "\n",
    "from reco.training import precision_recall_curve, roc_auc\n",
    "from reco.loss import QualityFocalLoss\n",
    "from reco.datasetPU import TracksterGraph\n",
    "\n",
    "\n",
    "ds_name = \"CloseByTwoPion\"\n",
    "\n",
    "data_root = \"data\"\n",
    "raw_dir = f\"/Users/ecuba/data/{ds_name}\"\n",
    "\n",
    "# data_root = \"/mnt/ceph/users/ecuba/processed\"\n",
    "# raw_dir = f\"/mnt/ceph/users/ecuba/{ds_name}\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "02f59304-76e4-4132-a379-978a46a042fd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using device: cpu\n"
     ]
    }
   ],
   "source": [
    "# CUDA Setup\n",
    "device = torch.device('cuda' if torch.cuda.is_available() else \"cpu\")\n",
    "print(f\"Using device: {device}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "ba68cd4a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "TracksterGraph(graphs=19968, nodes=563041, radius=100, bigT_e_th=10, lp)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ds = TracksterGraph(\n",
    "    ds_name,\n",
    "    data_root,\n",
    "    raw_dir,\n",
    "    N_FILES=200,\n",
    "    radius=100,\n",
    "    bigT_e_th=10,\n",
    "    link_prediction=True,\n",
    "    pileup=False,\n",
    ")\n",
    "ds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "a65110bd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2421912"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(ds.data.y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "604298f4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train graphs: 17972, Test graphs: 1996, total: 19968\n"
     ]
    }
   ],
   "source": [
    "ds_size = len(ds)\n",
    "test_set_size = ds_size // 20\n",
    "train_set_size = ds_size - test_set_size\n",
    "train_set, test_set = random_split(ds, [train_set_size, test_set_size])\n",
    "print(f\"Train graphs: {len(train_set)}, Test graphs: {len(test_set)}, total: {ds_size}\")\n",
    "\n",
    "train_dl = DataLoader(train_set, batch_size=8, shuffle=True, drop_last=True)\n",
    "test_dl = DataLoader(test_set, batch_size=8, shuffle=True, drop_last=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "3f4a0b99-bef3-4c82-a0ac-d88e1521382b",
   "metadata": {},
   "outputs": [],
   "source": [
    "class LinkPredictionNet(nn.Module):\n",
    "    def __init__(self, input_dim, output_dim=1, dropout=0.1):\n",
    "        super(LinkPredictionNet, self).__init__()\n",
    "        self.norm = gnn.LayerNorm(input_dim)\n",
    "        \n",
    "        hdim1 = 32\n",
    "        hdim2 = 64\n",
    "        hdim_fc = 128\n",
    "\n",
    "        self.edgeconv1 = DynamicEdgeConvBlock(input_dim, hdim1, k=7)\n",
    "        self.edgeconv2 = DynamicEdgeConvBlock(hdim1, hdim2, k=7)\n",
    "        \n",
    "        # Edge features from node embeddings for classification        \n",
    "        self.nodenetwork = nn.Sequential(\n",
    "            nn.Linear(2 * hdim2, hdim_fc),\n",
    "            nn.Sigmoid(),\n",
    "            nn.Dropout(dropout),\n",
    "            nn.Linear(hdim_fc, output_dim),\n",
    "        )\n",
    "\n",
    "    def forward(self, X, edge_index, batch=None):\n",
    "        \n",
    "        #(prepared_edges, _) = geou.add_self_loops(edge_index)  \n",
    "        #undirected_index = geou.to_undirected(prepared_edges)\n",
    "\n",
    "        H0 = X #self.norm(X, batch)\n",
    "        H1 = self.edgeconv1(H0, batch=batch)\n",
    "        H2 = self.edgeconv2(H1, batch=batch)\n",
    "        \n",
    "        H3 = H2\n",
    "\n",
    "        src, dst = edge_index\n",
    "        edges = torch.cat([H3[src], H3[dst]], dim=-1)\n",
    "        return self.nodenetwork(edges).squeeze(-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "2539cf65",
   "metadata": {},
   "outputs": [],
   "source": [
    "def train_link_prediction(model, device, optimizer, loss_func, train_dl):\n",
    "    train_loss = []\n",
    "    model.train()\n",
    "\n",
    "    for data in train_dl:\n",
    "        data = data.to(device)\n",
    "        optimizer.zero_grad()\n",
    "\n",
    "        seg_pred = model(data.x, data.edge_index, batch=data.batch)        \n",
    "        loss = loss_func(seg_pred, data.y)\n",
    "        \n",
    "        loss.backward()\n",
    "        \n",
    "        optimizer.step()\n",
    "        train_loss.append(loss.item())\n",
    "\n",
    "    return np.mean(train_loss)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "259ee8b3",
   "metadata": {},
   "outputs": [],
   "source": [
    "@torch.no_grad()\n",
    "def test_link_prediction(model, device, loss_func, test_dl):\n",
    "    test_loss = []\n",
    "    model.eval()\n",
    "    for data in test_dl:\n",
    "        data = data.to(device)\n",
    "        seg_pred = model(data.x, data.edge_index, data.batch)\n",
    "        loss = loss_func(seg_pred, data.y)\n",
    "        test_loss.append(loss.item())\n",
    "\n",
    "    return np.mean(test_loss)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "b85638de",
   "metadata": {},
   "outputs": [],
   "source": [
    "model = LinkPredictionNet(input_dim=ds.data.x.shape[1])\n",
    "epochs = 51\n",
    "model_path = f\"models/LinkPredictionNet.PnetL.ns.{epochs}e-{ds_name}.r{ds.RADIUS}.e{ds.bigT_e_th}.f{ds.N_FILES}.pt\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "2d3f1e82-74c5-4629-9b20-f05d7662aaf1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "skipping\n"
     ]
    }
   ],
   "source": [
    "%%script echo skipping\n",
    "# alpha - percentage of negative edges\n",
    "loss_func = nn.BCEWithLogitsLoss()\n",
    "\n",
    "model = model.to(device)\n",
    "optimizer = Adam(model.parameters(), lr=1e-3)\n",
    "scheduler = CosineAnnealingLR(optimizer, epochs, eta_min=1e-5)\n",
    "\n",
    "for epoch in range(epochs):\n",
    "\n",
    "    train_loss = train_link_prediction(\n",
    "        model,\n",
    "        device,\n",
    "        optimizer,\n",
    "        loss_func,\n",
    "        train_dl\n",
    "    )\n",
    "    \n",
    "    scheduler.step()\n",
    "\n",
    "    if epoch % 2 == 0:\n",
    "        test_loss = test_link_prediction(model, device, loss_func, test_dl)\n",
    "        print(\n",
    "            f\"Epoch {epoch}:\\t train loss: {train_loss:.3f}\\t validation loss: {test_loss:.3f}\",\n",
    "            file=sys.stderr\n",
    "        )\n",
    "        \n",
    "torch.save(model.state_dict(), model_path)\n",
    "print(model_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "0b9cbaa6",
   "metadata": {},
   "outputs": [],
   "source": [
    "#%%script echo skipping\n",
    "model.load_state_dict(torch.load(\n",
    "    model_path,\n",
    "    map_location=device\n",
    "))\n",
    "model = model.to(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "c86c98db-0ac0-4a0c-84fd-4c50df507528",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.959385396796905\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[24], line 3\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[39m#%%script echo skipping\u001b[39;00m\n\u001b[1;32m      2\u001b[0m \u001b[39mprint\u001b[39m(roc_auc(model, device, test_dl))\n\u001b[0;32m----> 3\u001b[0m precision_recall_curve(model, device, test_dl, focus_metric\u001b[39m=\u001b[39;49m\u001b[39m\"\u001b[39;49m\u001b[39mfbeta\u001b[39;49m\u001b[39m\"\u001b[39;49m, beta\u001b[39m=\u001b[39;49m\u001b[39m1\u001b[39;49m, step\u001b[39m=\u001b[39;49m\u001b[39m3\u001b[39;49m, truth_threshold\u001b[39m=\u001b[39;49m\u001b[39m0.5\u001b[39;49m)\n",
      "File \u001b[0;32m~/devel/pattern-reco/ve/lib/python3.9/site-packages/torch/autograd/grad_mode.py:27\u001b[0m, in \u001b[0;36m_DecoratorContextManager.__call__.<locals>.decorate_context\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m     24\u001b[0m \u001b[39m@functools\u001b[39m\u001b[39m.\u001b[39mwraps(func)\n\u001b[1;32m     25\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39mdecorate_context\u001b[39m(\u001b[39m*\u001b[39margs, \u001b[39m*\u001b[39m\u001b[39m*\u001b[39mkwargs):\n\u001b[1;32m     26\u001b[0m     \u001b[39mwith\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mclone():\n\u001b[0;32m---> 27\u001b[0m         \u001b[39mreturn\u001b[39;00m func(\u001b[39m*\u001b[39;49margs, \u001b[39m*\u001b[39;49m\u001b[39m*\u001b[39;49mkwargs)\n",
      "File \u001b[0;32m~/devel/pattern-reco/reco/training.py:160\u001b[0m, in \u001b[0;36mprecision_recall_curve\u001b[0;34m(model, device, test_dl, beta, truth_threshold, step, focus_metric)\u001b[0m\n\u001b[1;32m    158\u001b[0m pred \u001b[39m=\u001b[39m []\n\u001b[1;32m    159\u001b[0m lab \u001b[39m=\u001b[39m []\n\u001b[0;32m--> 160\u001b[0m \u001b[39mfor\u001b[39;00m data \u001b[39min\u001b[39;00m test_dl:\n\u001b[1;32m    162\u001b[0m     \u001b[39mif\u001b[39;00m \u001b[39misinstance\u001b[39m(data, Data):\n\u001b[1;32m    163\u001b[0m         \u001b[39m# graph dataset\u001b[39;00m\n\u001b[1;32m    164\u001b[0m         l \u001b[39m=\u001b[39m data\u001b[39m.\u001b[39my\u001b[39m.\u001b[39mreshape(\u001b[39m-\u001b[39m\u001b[39m1\u001b[39m)\n",
      "File \u001b[0;32m~/devel/pattern-reco/ve/lib/python3.9/site-packages/torch/utils/data/dataloader.py:648\u001b[0m, in \u001b[0;36m_BaseDataLoaderIter.__next__\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    647\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39m__next__\u001b[39m(\u001b[39mself\u001b[39m) \u001b[39m-\u001b[39m\u001b[39m>\u001b[39m Any:\n\u001b[0;32m--> 648\u001b[0m     \u001b[39mwith\u001b[39;00m torch\u001b[39m.\u001b[39mautograd\u001b[39m.\u001b[39mprofiler\u001b[39m.\u001b[39mrecord_function(\u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_profile_name):\n\u001b[1;32m    649\u001b[0m         \u001b[39mif\u001b[39;00m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_sampler_iter \u001b[39mis\u001b[39;00m \u001b[39mNone\u001b[39;00m:\n\u001b[1;32m    650\u001b[0m             \u001b[39m# TODO(https://github.com/pytorch/pytorch/issues/76750)\u001b[39;00m\n\u001b[1;32m    651\u001b[0m             \u001b[39mself\u001b[39m\u001b[39m.\u001b[39m_reset()  \u001b[39m# type: ignore[call-arg]\u001b[39;00m\n",
      "File \u001b[0;32m~/devel/pattern-reco/ve/lib/python3.9/site-packages/torch/autograd/profiler.py:446\u001b[0m, in \u001b[0;36mrecord_function.__enter__\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    445\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39m__enter__\u001b[39m(\u001b[39mself\u001b[39m):\n\u001b[0;32m--> 446\u001b[0m     \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mhandle \u001b[39m=\u001b[39m torch\u001b[39m.\u001b[39;49mops\u001b[39m.\u001b[39;49mprofiler\u001b[39m.\u001b[39;49m_record_function_enter(\u001b[39mself\u001b[39m\u001b[39m.\u001b[39mname, \u001b[39mself\u001b[39m\u001b[39m.\u001b[39margs)\n\u001b[1;32m    447\u001b[0m     \u001b[39mreturn\u001b[39;00m \u001b[39mself\u001b[39m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "#%%script echo skipping\n",
    "print(roc_auc(model, device, test_dl))\n",
    "precision_recall_curve(model, device, test_dl, focus_metric=\"fbeta\", beta=1, step=3, truth_threshold=0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f5302623",
   "metadata": {},
   "outputs": [],
   "source": [
    "del ds"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "505ae0f3",
   "metadata": {},
   "source": [
    "--> overfits: need more data!!"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "eb8ce4e2",
   "metadata": {},
   "source": [
    "## Eval"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "fb3f12d6",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from reco.data import get_event_data\n",
    "from reco.evaluation import model_evaluation\n",
    "\n",
    "file_name = f\"{raw_dir}/new_ntuples_15101852_0.root\"\n",
    "cluster_data, trackster_data, simtrackster_data, assoc_data = get_event_data(file_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "6b5ffc46",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " --- Radius threshold: 5 ---\n",
      "Event 0:\n",
      "\tclue3d_to_sim:\tP: 0.988 R: 0.229 F: 0.594\n",
      "\ttarget_to_sim:\tP: 0.987 R: 0.544 F: 0.849\n",
      "\treco_to_sim:\tP: 0.987 R: 0.544 F: 0.849\n",
      "\t|S| = 2 |T| = 22 |R| = 22\n",
      "Event 1:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.283 F: 0.664\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.348 F: 0.728\n",
      "\treco_to_sim:\tP: 1.000 R: 0.348 F: 0.728\n",
      "\t|S| = 2 |T| = 11 |R| = 11\n",
      "Event 2:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.396 F: 0.766\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.402 F: 0.771\n",
      "\treco_to_sim:\tP: 1.000 R: 0.402 F: 0.771\n",
      "\t|S| = 2 |T| = 41 |R| = 41\n",
      "Event 3:\n",
      "\tclue3d_to_sim:\tP: 0.995 R: 0.383 F: 0.754\n",
      "\ttarget_to_sim:\tP: 0.994 R: 0.648 F: 0.898\n",
      "\treco_to_sim:\tP: 0.994 R: 0.648 F: 0.898\n",
      "\t|S| = 2 |T| = 18 |R| = 18\n",
      "Event 4:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.250 F: 0.625\n",
      "\ttarget_to_sim:\tP: 0.997 R: 0.602 F: 0.881\n",
      "\treco_to_sim:\tP: 0.997 R: 0.602 F: 0.881\n",
      "\t|S| = 2 |T| = 35 |R| = 35\n",
      "Event 5:\n",
      "\tclue3d_to_sim:\tP: 0.996 R: 0.551 F: 0.858\n",
      "\ttarget_to_sim:\tP: 0.996 R: 0.673 F: 0.909\n",
      "\treco_to_sim:\tP: 0.996 R: 0.673 F: 0.909\n",
      "\t|S| = 2 |T| = 16 |R| = 16\n",
      "Event 6:\n",
      "\tclue3d_to_sim:\tP: 0.992 R: 0.219 F: 0.582\n",
      "\ttarget_to_sim:\tP: 0.992 R: 0.566 F: 0.862\n",
      "\treco_to_sim:\tP: 0.992 R: 0.566 F: 0.862\n",
      "\t|S| = 2 |T| = 18 |R| = 18\n",
      "Event 7:\n",
      "\tclue3d_to_sim:\tP: 0.974 R: 0.112 F: 0.385\n",
      "\ttarget_to_sim:\tP: 0.974 R: 0.186 F: 0.526\n",
      "\treco_to_sim:\tP: 0.925 R: 0.224 F: 0.569\n",
      "\t|S| = 2 |T| = 5 |R| = 3\n",
      "Event 8:\n",
      "\tclue3d_to_sim:\tP: 0.997 R: 0.413 F: 0.777\n",
      "\ttarget_to_sim:\tP: 0.997 R: 0.535 F: 0.850\n",
      "\treco_to_sim:\tP: 0.997 R: 0.535 F: 0.850\n",
      "\t|S| = 2 |T| = 19 |R| = 19\n",
      "Event 9:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.177 F: 0.518\n",
      "\ttarget_to_sim:\tP: 0.998 R: 0.410 F: 0.776\n",
      "\treco_to_sim:\tP: 0.998 R: 0.410 F: 0.776\n",
      "\t|S| = 2 |T| = 15 |R| = 15\n",
      "Event 10:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.340 F: 0.720\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.384 F: 0.757\n",
      "\treco_to_sim:\tP: 1.000 R: 0.340 F: 0.720\n",
      "\t|S| = 2 |T| = 13 |R| = 15\n",
      "Event 11:\n",
      "\tclue3d_to_sim:\tP: 0.999 R: 0.524 F: 0.846\n",
      "\ttarget_to_sim:\tP: 0.999 R: 0.613 F: 0.888\n",
      "\treco_to_sim:\tP: 0.999 R: 0.613 F: 0.888\n",
      "\t|S| = 2 |T| = 24 |R| = 24\n",
      "Event 12:\n",
      "\tclue3d_to_sim:\tP: 0.984 R: 0.277 F: 0.651\n",
      "\ttarget_to_sim:\tP: 0.977 R: 0.549 F: 0.845\n",
      "\treco_to_sim:\tP: 0.970 R: 0.546 F: 0.840\n",
      "\t|S| = 2 |T| = 13 |R| = 13\n",
      "Event 13:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.258 F: 0.635\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.558 F: 0.863\n",
      "\treco_to_sim:\tP: 1.000 R: 0.558 F: 0.863\n",
      "\t|S| = 2 |T| = 17 |R| = 17\n",
      "Event 14:\n",
      "\tclue3d_to_sim:\tP: 0.976 R: 0.155 F: 0.475\n",
      "\ttarget_to_sim:\tP: 0.977 R: 0.333 F: 0.704\n",
      "\treco_to_sim:\tP: 0.974 R: 0.340 F: 0.709\n",
      "\t|S| = 2 |T| = 16 |R| = 15\n",
      "Event 15:\n",
      "\tclue3d_to_sim:\tP: 0.996 R: 0.209 F: 0.568\n",
      "\ttarget_to_sim:\tP: 0.996 R: 0.539 F: 0.852\n",
      "\treco_to_sim:\tP: 0.996 R: 0.539 F: 0.852\n",
      "\t|S| = 2 |T| = 18 |R| = 18\n",
      "Event 16:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.376 F: 0.750\n",
      "\ttarget_to_sim:\tP: 0.998 R: 0.422 F: 0.784\n",
      "\treco_to_sim:\tP: 0.998 R: 0.422 F: 0.784\n",
      "\t|S| = 2 |T| = 11 |R| = 11\n",
      "Event 17:\n",
      "\tclue3d_to_sim:\tP: 0.985 R: 0.268 F: 0.642\n",
      "\ttarget_to_sim:\tP: 0.984 R: 0.533 F: 0.842\n",
      "\treco_to_sim:\tP: 0.984 R: 0.535 F: 0.842\n",
      "\t|S| = 2 |T| = 26 |R| = 25\n",
      "Event 18:\n",
      "\tclue3d_to_sim:\tP: 0.989 R: 0.317 F: 0.695\n",
      "\ttarget_to_sim:\tP: 0.989 R: 0.685 F: 0.909\n",
      "\treco_to_sim:\tP: 0.989 R: 0.685 F: 0.909\n",
      "\t|S| = 2 |T| = 11 |R| = 11\n",
      "Event 19:\n",
      "\tclue3d_to_sim:\tP: 0.991 R: 0.277 F: 0.654\n",
      "\ttarget_to_sim:\tP: 0.989 R: 0.362 F: 0.735\n",
      "\treco_to_sim:\tP: 0.989 R: 0.362 F: 0.735\n",
      "\t|S| = 2 |T| = 19 |R| = 19\n",
      "-----\n",
      "mean clue3d_to_sim:\tP: 0.993 R: 0.301 F: 0.658\n",
      "mean target_to_sim:\tP: 0.992 R: 0.495 F: 0.811\n",
      "mean reco_to_sim:\tP: 0.989 R: 0.495 F: 0.812\n",
      "mean n_tracksters:\tP: 2.000 R: 18.400 F: 18.300\n",
      " --- Radius threshold: 10 ---\n",
      "Event 0:\n",
      "\tclue3d_to_sim:\tP: 0.988 R: 0.229 F: 0.594\n",
      "\ttarget_to_sim:\tP: 0.981 R: 0.640 F: 0.886\n",
      "\treco_to_sim:\tP: 0.593 R: 0.664 F: 0.606\n",
      "\t|S| = 2 |T| = 9 |R| = 7\n",
      "Event 1:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.283 F: 0.664\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.649 F: 0.903\n",
      "\treco_to_sim:\tP: 1.000 R: 0.649 F: 0.903\n",
      "\t|S| = 2 |T| = 3 |R| = 3\n",
      "Event 2:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.396 F: 0.766\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.575 F: 0.871\n",
      "\treco_to_sim:\tP: 1.000 R: 0.575 F: 0.871\n",
      "\t|S| = 2 |T| = 17 |R| = 17\n",
      "Event 3:\n",
      "\tclue3d_to_sim:\tP: 0.995 R: 0.383 F: 0.754\n",
      "\ttarget_to_sim:\tP: 0.639 R: 0.753 F: 0.659\n",
      "\treco_to_sim:\tP: 0.992 R: 0.745 F: 0.930\n",
      "\t|S| = 2 |T| = 5 |R| = 6\n",
      "Event 4:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.250 F: 0.625\n",
      "\ttarget_to_sim:\tP: 0.997 R: 0.742 F: 0.933\n",
      "\treco_to_sim:\tP: 0.997 R: 0.742 F: 0.933\n",
      "\t|S| = 2 |T| = 14 |R| = 14\n",
      "Event 5:\n",
      "\tclue3d_to_sim:\tP: 0.996 R: 0.551 F: 0.858\n",
      "\ttarget_to_sim:\tP: 0.996 R: 0.811 F: 0.953\n",
      "\treco_to_sim:\tP: 0.996 R: 0.811 F: 0.953\n",
      "\t|S| = 2 |T| = 5 |R| = 5\n",
      "Event 6:\n",
      "\tclue3d_to_sim:\tP: 0.992 R: 0.219 F: 0.582\n",
      "\ttarget_to_sim:\tP: 0.992 R: 0.722 F: 0.923\n",
      "\treco_to_sim:\tP: 0.851 R: 0.738 F: 0.826\n",
      "\t|S| = 2 |T| = 3 |R| = 1\n",
      "Event 7:\n",
      "\tclue3d_to_sim:\tP: 0.974 R: 0.112 F: 0.385\n",
      "\ttarget_to_sim:\tP: 0.968 R: 0.216 F: 0.571\n",
      "\treco_to_sim:\tP: 0.923 R: 0.269 F: 0.621\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 8:\n",
      "\tclue3d_to_sim:\tP: 0.997 R: 0.413 F: 0.777\n",
      "\ttarget_to_sim:\tP: 0.996 R: 0.755 F: 0.936\n",
      "\treco_to_sim:\tP: 0.996 R: 0.748 F: 0.934\n",
      "\t|S| = 2 |T| = 3 |R| = 4\n",
      "Event 9:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.177 F: 0.518\n",
      "\ttarget_to_sim:\tP: 0.998 R: 0.466 F: 0.813\n",
      "\treco_to_sim:\tP: 0.998 R: 0.466 F: 0.813\n",
      "\t|S| = 2 |T| = 6 |R| = 6\n",
      "Event 10:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.340 F: 0.720\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.456 F: 0.807\n",
      "\treco_to_sim:\tP: 1.000 R: 0.456 F: 0.807\n",
      "\t|S| = 2 |T| = 11 |R| = 11\n",
      "Event 11:\n",
      "\tclue3d_to_sim:\tP: 0.999 R: 0.524 F: 0.846\n",
      "\ttarget_to_sim:\tP: 0.999 R: 0.680 F: 0.913\n",
      "\treco_to_sim:\tP: 0.999 R: 0.680 F: 0.913\n",
      "\t|S| = 2 |T| = 12 |R| = 12\n",
      "Event 12:\n",
      "\tclue3d_to_sim:\tP: 0.984 R: 0.277 F: 0.651\n",
      "\ttarget_to_sim:\tP: 0.975 R: 0.612 F: 0.872\n",
      "\treco_to_sim:\tP: 0.737 R: 0.639 F: 0.715\n",
      "\t|S| = 2 |T| = 3 |R| = 1\n",
      "Event 13:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.258 F: 0.635\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.711 F: 0.924\n",
      "\treco_to_sim:\tP: 1.000 R: 0.711 F: 0.924\n",
      "\t|S| = 2 |T| = 7 |R| = 7\n",
      "Event 14:\n",
      "\tclue3d_to_sim:\tP: 0.976 R: 0.155 F: 0.475\n",
      "\ttarget_to_sim:\tP: 0.820 R: 0.404 F: 0.680\n",
      "\treco_to_sim:\tP: 0.854 R: 0.425 F: 0.711\n",
      "\t|S| = 2 |T| = 4 |R| = 4\n",
      "Event 15:\n",
      "\tclue3d_to_sim:\tP: 0.996 R: 0.209 F: 0.568\n",
      "\ttarget_to_sim:\tP: 0.996 R: 0.701 F: 0.919\n",
      "\treco_to_sim:\tP: 0.996 R: 0.701 F: 0.919\n",
      "\t|S| = 2 |T| = 7 |R| = 7\n",
      "Event 16:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.376 F: 0.750\n",
      "\ttarget_to_sim:\tP: 0.998 R: 0.507 F: 0.836\n",
      "\treco_to_sim:\tP: 0.998 R: 0.507 F: 0.836\n",
      "\t|S| = 2 |T| = 8 |R| = 8\n",
      "Event 17:\n",
      "\tclue3d_to_sim:\tP: 0.985 R: 0.268 F: 0.642\n",
      "\ttarget_to_sim:\tP: 0.983 R: 0.685 F: 0.904\n",
      "\treco_to_sim:\tP: 0.975 R: 0.684 F: 0.899\n",
      "\t|S| = 2 |T| = 6 |R| = 4\n",
      "Event 18:\n",
      "\tclue3d_to_sim:\tP: 0.989 R: 0.317 F: 0.695\n",
      "\ttarget_to_sim:\tP: 0.987 R: 0.757 F: 0.931\n",
      "\treco_to_sim:\tP: 0.945 R: 0.775 F: 0.905\n",
      "\t|S| = 2 |T| = 7 |R| = 5\n",
      "Event 19:\n",
      "\tclue3d_to_sim:\tP: 0.991 R: 0.277 F: 0.654\n",
      "\ttarget_to_sim:\tP: 0.988 R: 0.493 F: 0.823\n",
      "\treco_to_sim:\tP: 0.986 R: 0.493 F: 0.822\n",
      "\t|S| = 2 |T| = 5 |R| = 4\n",
      "-----\n",
      "mean clue3d_to_sim:\tP: 0.993 R: 0.301 F: 0.658\n",
      "mean target_to_sim:\tP: 0.966 R: 0.617 F: 0.853\n",
      "mean reco_to_sim:\tP: 0.942 R: 0.624 F: 0.842\n",
      "mean n_tracksters:\tP: 2.000 R: 6.850 F: 6.350\n",
      " --- Radius threshold: 20 ---\n",
      "Event 0:\n",
      "\tclue3d_to_sim:\tP: 0.988 R: 0.229 F: 0.594\n",
      "\ttarget_to_sim:\tP: 0.981 R: 0.682 F: 0.902\n",
      "\treco_to_sim:\tP: 0.586 R: 0.693 F: 0.605\n",
      "\t|S| = 2 |T| = 3 |R| = 3\n",
      "Event 1:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.283 F: 0.664\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.659 F: 0.906\n",
      "\treco_to_sim:\tP: 1.000 R: 0.659 F: 0.906\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 2:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.396 F: 0.766\n",
      "\ttarget_to_sim:\tP: 0.999 R: 0.814 F: 0.956\n",
      "\treco_to_sim:\tP: 0.999 R: 0.814 F: 0.956\n",
      "\t|S| = 2 |T| = 3 |R| = 3\n",
      "Event 3:\n",
      "\tclue3d_to_sim:\tP: 0.995 R: 0.383 F: 0.754\n",
      "\ttarget_to_sim:\tP: 0.631 R: 0.792 F: 0.658\n",
      "\treco_to_sim:\tP: 0.631 R: 0.792 F: 0.658\n",
      "\t|S| = 2 |T| = 1 |R| = 1\n",
      "Event 4:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.250 F: 0.625\n",
      "\ttarget_to_sim:\tP: 0.997 R: 0.796 F: 0.949\n",
      "\treco_to_sim:\tP: 0.992 R: 0.795 F: 0.945\n",
      "\t|S| = 2 |T| = 4 |R| = 3\n",
      "Event 5:\n",
      "\tclue3d_to_sim:\tP: 0.996 R: 0.551 F: 0.858\n",
      "\ttarget_to_sim:\tP: 0.995 R: 0.831 F: 0.957\n",
      "\treco_to_sim:\tP: 0.995 R: 0.831 F: 0.957\n",
      "\t|S| = 2 |T| = 3 |R| = 3\n",
      "Event 6:\n",
      "\tclue3d_to_sim:\tP: 0.992 R: 0.219 F: 0.582\n",
      "\ttarget_to_sim:\tP: 0.992 R: 0.727 F: 0.925\n",
      "\treco_to_sim:\tP: 0.851 R: 0.738 F: 0.826\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 7:\n",
      "\tclue3d_to_sim:\tP: 0.974 R: 0.112 F: 0.385\n",
      "\ttarget_to_sim:\tP: 0.968 R: 0.216 F: 0.571\n",
      "\treco_to_sim:\tP: 0.923 R: 0.269 F: 0.621\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 8:\n",
      "\tclue3d_to_sim:\tP: 0.997 R: 0.413 F: 0.777\n",
      "\ttarget_to_sim:\tP: 0.996 R: 0.755 F: 0.936\n",
      "\treco_to_sim:\tP: 0.996 R: 0.748 F: 0.934\n",
      "\t|S| = 2 |T| = 3 |R| = 4\n",
      "Event 9:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.177 F: 0.518\n",
      "\ttarget_to_sim:\tP: 0.998 R: 0.562 F: 0.864\n",
      "\treco_to_sim:\tP: 0.705 R: 0.562 F: 0.671\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 10:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.340 F: 0.720\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.456 F: 0.807\n",
      "\treco_to_sim:\tP: 1.000 R: 0.456 F: 0.807\n",
      "\t|S| = 2 |T| = 11 |R| = 11\n",
      "Event 11:\n",
      "\tclue3d_to_sim:\tP: 0.999 R: 0.524 F: 0.846\n",
      "\ttarget_to_sim:\tP: 0.999 R: 0.819 F: 0.957\n",
      "\treco_to_sim:\tP: 0.999 R: 0.819 F: 0.957\n",
      "\t|S| = 2 |T| = 4 |R| = 4\n",
      "Event 12:\n",
      "\tclue3d_to_sim:\tP: 0.984 R: 0.277 F: 0.651\n",
      "\ttarget_to_sim:\tP: 0.975 R: 0.616 F: 0.873\n",
      "\treco_to_sim:\tP: 0.737 R: 0.639 F: 0.715\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 13:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.258 F: 0.635\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.757 F: 0.939\n",
      "\treco_to_sim:\tP: 1.000 R: 0.757 F: 0.939\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 14:\n",
      "\tclue3d_to_sim:\tP: 0.976 R: 0.155 F: 0.475\n",
      "\ttarget_to_sim:\tP: 0.785 R: 0.605 F: 0.741\n",
      "\treco_to_sim:\tP: 0.785 R: 0.605 F: 0.741\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 15:\n",
      "\tclue3d_to_sim:\tP: 0.996 R: 0.209 F: 0.568\n",
      "\ttarget_to_sim:\tP: 0.993 R: 0.745 F: 0.931\n",
      "\treco_to_sim:\tP: 0.993 R: 0.745 F: 0.931\n",
      "\t|S| = 2 |T| = 4 |R| = 4\n",
      "Event 16:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.376 F: 0.750\n",
      "\ttarget_to_sim:\tP: 0.997 R: 0.656 F: 0.903\n",
      "\treco_to_sim:\tP: 0.997 R: 0.656 F: 0.903\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 17:\n",
      "\tclue3d_to_sim:\tP: 0.985 R: 0.268 F: 0.642\n",
      "\ttarget_to_sim:\tP: 0.682 R: 0.781 F: 0.700\n",
      "\treco_to_sim:\tP: 0.975 R: 0.758 F: 0.922\n",
      "\t|S| = 2 |T| = 1 |R| = 2\n",
      "Event 18:\n",
      "\tclue3d_to_sim:\tP: 0.989 R: 0.317 F: 0.695\n",
      "\ttarget_to_sim:\tP: 0.987 R: 0.779 F: 0.937\n",
      "\treco_to_sim:\tP: 0.981 R: 0.757 F: 0.927\n",
      "\t|S| = 2 |T| = 5 |R| = 5\n",
      "Event 19:\n",
      "\tclue3d_to_sim:\tP: 0.991 R: 0.277 F: 0.654\n",
      "\ttarget_to_sim:\tP: 0.987 R: 0.494 F: 0.823\n",
      "\treco_to_sim:\tP: 0.986 R: 0.493 F: 0.822\n",
      "\t|S| = 2 |T| = 4 |R| = 4\n",
      "-----\n",
      "mean clue3d_to_sim:\tP: 0.993 R: 0.301 F: 0.658\n",
      "mean target_to_sim:\tP: 0.948 R: 0.677 F: 0.862\n",
      "mean reco_to_sim:\tP: 0.907 R: 0.679 F: 0.837\n",
      "mean n_tracksters:\tP: 2.000 R: 3.100 F: 3.000\n",
      " --- Radius threshold: 30 ---\n",
      "Event 0:\n",
      "\tclue3d_to_sim:\tP: 0.988 R: 0.229 F: 0.594\n",
      "\ttarget_to_sim:\tP: 0.981 R: 0.686 F: 0.903\n",
      "\treco_to_sim:\tP: 0.585 R: 0.697 F: 0.605\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 1:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.283 F: 0.664\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.659 F: 0.906\n",
      "\treco_to_sim:\tP: 1.000 R: 0.659 F: 0.906\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 2:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.396 F: 0.766\n",
      "\ttarget_to_sim:\tP: 0.999 R: 0.814 F: 0.956\n",
      "\treco_to_sim:\tP: 0.999 R: 0.814 F: 0.956\n",
      "\t|S| = 2 |T| = 2 |R| = 3\n",
      "Event 3:\n",
      "\tclue3d_to_sim:\tP: 0.995 R: 0.383 F: 0.754\n",
      "\ttarget_to_sim:\tP: 0.631 R: 0.792 F: 0.658\n",
      "\treco_to_sim:\tP: 0.631 R: 0.792 F: 0.658\n",
      "\t|S| = 2 |T| = 1 |R| = 1\n",
      "Event 4:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.250 F: 0.625\n",
      "\ttarget_to_sim:\tP: 0.544 R: 0.842 F: 0.586\n",
      "\treco_to_sim:\tP: 0.992 R: 0.833 F: 0.955\n",
      "\t|S| = 2 |T| = 1 |R| = 2\n",
      "Event 5:\n",
      "\tclue3d_to_sim:\tP: 0.996 R: 0.551 F: 0.858\n",
      "\ttarget_to_sim:\tP: 0.995 R: 0.831 F: 0.957\n",
      "\treco_to_sim:\tP: 0.995 R: 0.831 F: 0.957\n",
      "\t|S| = 2 |T| = 3 |R| = 3\n",
      "Event 6:\n",
      "\tclue3d_to_sim:\tP: 0.992 R: 0.219 F: 0.582\n",
      "\ttarget_to_sim:\tP: 0.992 R: 0.727 F: 0.925\n",
      "\treco_to_sim:\tP: 0.851 R: 0.738 F: 0.826\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 7:\n",
      "\tclue3d_to_sim:\tP: 0.974 R: 0.112 F: 0.385\n",
      "\ttarget_to_sim:\tP: 0.968 R: 0.216 F: 0.571\n",
      "\treco_to_sim:\tP: 0.923 R: 0.269 F: 0.621\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 8:\n",
      "\tclue3d_to_sim:\tP: 0.997 R: 0.413 F: 0.777\n",
      "\ttarget_to_sim:\tP: 0.996 R: 0.755 F: 0.936\n",
      "\treco_to_sim:\tP: 0.996 R: 0.748 F: 0.934\n",
      "\t|S| = 2 |T| = 3 |R| = 4\n",
      "Event 9:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.177 F: 0.518\n",
      "\ttarget_to_sim:\tP: 0.998 R: 0.562 F: 0.864\n",
      "\treco_to_sim:\tP: 0.705 R: 0.562 F: 0.671\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 10:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.340 F: 0.720\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.500 F: 0.833\n",
      "\treco_to_sim:\tP: 1.000 R: 0.500 F: 0.833\n",
      "\t|S| = 2 |T| = 10 |R| = 10\n",
      "Event 11:\n",
      "\tclue3d_to_sim:\tP: 0.999 R: 0.524 F: 0.846\n",
      "\ttarget_to_sim:\tP: 0.999 R: 0.829 F: 0.960\n",
      "\treco_to_sim:\tP: 0.999 R: 0.829 F: 0.960\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 12:\n",
      "\tclue3d_to_sim:\tP: 0.984 R: 0.277 F: 0.651\n",
      "\ttarget_to_sim:\tP: 0.975 R: 0.616 F: 0.873\n",
      "\treco_to_sim:\tP: 0.737 R: 0.639 F: 0.715\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 13:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.258 F: 0.635\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.757 F: 0.939\n",
      "\treco_to_sim:\tP: 1.000 R: 0.757 F: 0.939\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 14:\n",
      "\tclue3d_to_sim:\tP: 0.976 R: 0.155 F: 0.475\n",
      "\ttarget_to_sim:\tP: 0.776 R: 0.650 F: 0.747\n",
      "\treco_to_sim:\tP: 0.776 R: 0.650 F: 0.747\n",
      "\t|S| = 2 |T| = 1 |R| = 1\n",
      "Event 15:\n",
      "\tclue3d_to_sim:\tP: 0.996 R: 0.209 F: 0.568\n",
      "\ttarget_to_sim:\tP: 0.993 R: 0.745 F: 0.931\n",
      "\treco_to_sim:\tP: 0.993 R: 0.745 F: 0.931\n",
      "\t|S| = 2 |T| = 4 |R| = 4\n",
      "Event 16:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.376 F: 0.750\n",
      "\ttarget_to_sim:\tP: 0.997 R: 0.656 F: 0.903\n",
      "\treco_to_sim:\tP: 0.997 R: 0.656 F: 0.903\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 17:\n",
      "\tclue3d_to_sim:\tP: 0.985 R: 0.268 F: 0.642\n",
      "\ttarget_to_sim:\tP: 0.682 R: 0.781 F: 0.700\n",
      "\treco_to_sim:\tP: 0.975 R: 0.758 F: 0.922\n",
      "\t|S| = 2 |T| = 1 |R| = 2\n",
      "Event 18:\n",
      "\tclue3d_to_sim:\tP: 0.989 R: 0.317 F: 0.695\n",
      "\ttarget_to_sim:\tP: 0.987 R: 0.779 F: 0.937\n",
      "\treco_to_sim:\tP: 0.981 R: 0.757 F: 0.927\n",
      "\t|S| = 2 |T| = 5 |R| = 5\n",
      "Event 19:\n",
      "\tclue3d_to_sim:\tP: 0.991 R: 0.277 F: 0.654\n",
      "\ttarget_to_sim:\tP: 0.987 R: 0.494 F: 0.823\n",
      "\treco_to_sim:\tP: 0.986 R: 0.493 F: 0.822\n",
      "\t|S| = 2 |T| = 4 |R| = 4\n",
      "-----\n",
      "mean clue3d_to_sim:\tP: 0.993 R: 0.301 F: 0.658\n",
      "mean target_to_sim:\tP: 0.925 R: 0.684 F: 0.845\n",
      "mean reco_to_sim:\tP: 0.906 R: 0.686 F: 0.839\n",
      "mean n_tracksters:\tP: 2.000 R: 2.650 F: 2.700\n",
      " --- Radius threshold: 50 ---\n",
      "Event 0:\n",
      "\tclue3d_to_sim:\tP: 0.988 R: 0.229 F: 0.594\n",
      "\ttarget_to_sim:\tP: 0.981 R: 0.686 F: 0.903\n",
      "\treco_to_sim:\tP: 0.585 R: 0.697 F: 0.605\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 1:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.283 F: 0.664\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.659 F: 0.906\n",
      "\treco_to_sim:\tP: 1.000 R: 0.659 F: 0.906\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 2:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.396 F: 0.766\n",
      "\ttarget_to_sim:\tP: 0.999 R: 0.814 F: 0.956\n",
      "\treco_to_sim:\tP: 0.999 R: 0.814 F: 0.956\n",
      "\t|S| = 2 |T| = 2 |R| = 3\n",
      "Event 3:\n",
      "\tclue3d_to_sim:\tP: 0.995 R: 0.383 F: 0.754\n",
      "\ttarget_to_sim:\tP: 0.631 R: 0.792 F: 0.658\n",
      "\treco_to_sim:\tP: 0.631 R: 0.792 F: 0.658\n",
      "\t|S| = 2 |T| = 1 |R| = 1\n",
      "Event 4:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.250 F: 0.625\n",
      "\ttarget_to_sim:\tP: 0.544 R: 0.842 F: 0.586\n",
      "\treco_to_sim:\tP: 0.992 R: 0.833 F: 0.955\n",
      "\t|S| = 2 |T| = 1 |R| = 2\n",
      "Event 5:\n",
      "\tclue3d_to_sim:\tP: 0.996 R: 0.551 F: 0.858\n",
      "\ttarget_to_sim:\tP: 0.995 R: 0.839 F: 0.959\n",
      "\treco_to_sim:\tP: 0.995 R: 0.839 F: 0.959\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 6:\n",
      "\tclue3d_to_sim:\tP: 0.992 R: 0.219 F: 0.582\n",
      "\ttarget_to_sim:\tP: 0.992 R: 0.727 F: 0.925\n",
      "\treco_to_sim:\tP: 0.851 R: 0.738 F: 0.826\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 7:\n",
      "\tclue3d_to_sim:\tP: 0.974 R: 0.112 F: 0.385\n",
      "\ttarget_to_sim:\tP: 0.968 R: 0.216 F: 0.571\n",
      "\treco_to_sim:\tP: 0.923 R: 0.269 F: 0.621\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 8:\n",
      "\tclue3d_to_sim:\tP: 0.997 R: 0.413 F: 0.777\n",
      "\ttarget_to_sim:\tP: 0.996 R: 0.761 F: 0.938\n",
      "\treco_to_sim:\tP: 0.996 R: 0.755 F: 0.936\n",
      "\t|S| = 2 |T| = 2 |R| = 3\n",
      "Event 9:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.177 F: 0.518\n",
      "\ttarget_to_sim:\tP: 0.998 R: 0.562 F: 0.864\n",
      "\treco_to_sim:\tP: 0.704 R: 0.565 F: 0.671\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 10:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.340 F: 0.720\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.530 F: 0.849\n",
      "\treco_to_sim:\tP: 1.000 R: 0.530 F: 0.849\n",
      "\t|S| = 2 |T| = 9 |R| = 9\n",
      "Event 11:\n",
      "\tclue3d_to_sim:\tP: 0.999 R: 0.524 F: 0.846\n",
      "\ttarget_to_sim:\tP: 0.999 R: 0.829 F: 0.960\n",
      "\treco_to_sim:\tP: 0.999 R: 0.829 F: 0.960\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 12:\n",
      "\tclue3d_to_sim:\tP: 0.984 R: 0.277 F: 0.651\n",
      "\ttarget_to_sim:\tP: 0.975 R: 0.616 F: 0.873\n",
      "\treco_to_sim:\tP: 0.737 R: 0.639 F: 0.715\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 13:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.258 F: 0.635\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.757 F: 0.939\n",
      "\treco_to_sim:\tP: 1.000 R: 0.757 F: 0.939\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 14:\n",
      "\tclue3d_to_sim:\tP: 0.976 R: 0.155 F: 0.475\n",
      "\ttarget_to_sim:\tP: 0.776 R: 0.650 F: 0.747\n",
      "\treco_to_sim:\tP: 0.776 R: 0.650 F: 0.747\n",
      "\t|S| = 2 |T| = 1 |R| = 1\n",
      "Event 15:\n",
      "\tclue3d_to_sim:\tP: 0.996 R: 0.209 F: 0.568\n",
      "\ttarget_to_sim:\tP: 0.993 R: 0.752 F: 0.933\n",
      "\treco_to_sim:\tP: 0.993 R: 0.752 F: 0.933\n",
      "\t|S| = 2 |T| = 3 |R| = 3\n",
      "Event 16:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.376 F: 0.750\n",
      "\ttarget_to_sim:\tP: 0.997 R: 0.656 F: 0.903\n",
      "\treco_to_sim:\tP: 0.997 R: 0.656 F: 0.903\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 17:\n",
      "\tclue3d_to_sim:\tP: 0.985 R: 0.268 F: 0.642\n",
      "\ttarget_to_sim:\tP: 0.682 R: 0.781 F: 0.700\n",
      "\treco_to_sim:\tP: 0.975 R: 0.758 F: 0.922\n",
      "\t|S| = 2 |T| = 1 |R| = 2\n",
      "Event 18:\n",
      "\tclue3d_to_sim:\tP: 0.989 R: 0.317 F: 0.695\n",
      "\ttarget_to_sim:\tP: 0.987 R: 0.779 F: 0.937\n",
      "\treco_to_sim:\tP: 0.980 R: 0.788 F: 0.934\n",
      "\t|S| = 2 |T| = 5 |R| = 4\n",
      "Event 19:\n",
      "\tclue3d_to_sim:\tP: 0.991 R: 0.277 F: 0.654\n",
      "\ttarget_to_sim:\tP: 0.987 R: 0.499 F: 0.826\n",
      "\treco_to_sim:\tP: 0.986 R: 0.498 F: 0.825\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "-----\n",
      "mean clue3d_to_sim:\tP: 0.993 R: 0.301 F: 0.658\n",
      "mean target_to_sim:\tP: 0.925 R: 0.687 F: 0.847\n",
      "mean reco_to_sim:\tP: 0.906 R: 0.691 F: 0.841\n",
      "mean n_tracksters:\tP: 2.000 R: 2.350 F: 2.300\n",
      " --- Radius threshold: 70 ---\n",
      "Event 0:\n",
      "\tclue3d_to_sim:\tP: 0.988 R: 0.229 F: 0.594\n",
      "\ttarget_to_sim:\tP: 0.981 R: 0.686 F: 0.903\n",
      "\treco_to_sim:\tP: 0.585 R: 0.697 F: 0.605\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 1:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.283 F: 0.664\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.659 F: 0.906\n",
      "\treco_to_sim:\tP: 1.000 R: 0.659 F: 0.906\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 2:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.396 F: 0.766\n",
      "\ttarget_to_sim:\tP: 0.999 R: 0.814 F: 0.956\n",
      "\treco_to_sim:\tP: 0.999 R: 0.814 F: 0.956\n",
      "\t|S| = 2 |T| = 2 |R| = 3\n",
      "Event 3:\n",
      "\tclue3d_to_sim:\tP: 0.995 R: 0.383 F: 0.754\n",
      "\ttarget_to_sim:\tP: 0.631 R: 0.792 F: 0.658\n",
      "\treco_to_sim:\tP: 0.631 R: 0.792 F: 0.658\n",
      "\t|S| = 2 |T| = 1 |R| = 1\n",
      "Event 4:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.250 F: 0.625\n",
      "\ttarget_to_sim:\tP: 0.544 R: 0.842 F: 0.586\n",
      "\treco_to_sim:\tP: 0.992 R: 0.833 F: 0.955\n",
      "\t|S| = 2 |T| = 1 |R| = 2\n",
      "Event 5:\n",
      "\tclue3d_to_sim:\tP: 0.996 R: 0.551 F: 0.858\n",
      "\ttarget_to_sim:\tP: 0.995 R: 0.839 F: 0.959\n",
      "\treco_to_sim:\tP: 0.995 R: 0.839 F: 0.959\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 6:\n",
      "\tclue3d_to_sim:\tP: 0.992 R: 0.219 F: 0.582\n",
      "\ttarget_to_sim:\tP: 0.992 R: 0.727 F: 0.925\n",
      "\treco_to_sim:\tP: 0.851 R: 0.738 F: 0.826\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 7:\n",
      "\tclue3d_to_sim:\tP: 0.974 R: 0.112 F: 0.385\n",
      "\ttarget_to_sim:\tP: 0.968 R: 0.216 F: 0.571\n",
      "\treco_to_sim:\tP: 0.923 R: 0.269 F: 0.621\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 8:\n",
      "\tclue3d_to_sim:\tP: 0.997 R: 0.413 F: 0.777\n",
      "\ttarget_to_sim:\tP: 0.996 R: 0.761 F: 0.938\n",
      "\treco_to_sim:\tP: 0.996 R: 0.755 F: 0.936\n",
      "\t|S| = 2 |T| = 2 |R| = 3\n",
      "Event 9:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.177 F: 0.518\n",
      "\ttarget_to_sim:\tP: 0.998 R: 0.562 F: 0.864\n",
      "\treco_to_sim:\tP: 0.704 R: 0.565 F: 0.671\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 10:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.340 F: 0.720\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.599 F: 0.882\n",
      "\treco_to_sim:\tP: 1.000 R: 0.599 F: 0.882\n",
      "\t|S| = 2 |T| = 8 |R| = 8\n",
      "Event 11:\n",
      "\tclue3d_to_sim:\tP: 0.999 R: 0.524 F: 0.846\n",
      "\ttarget_to_sim:\tP: 0.999 R: 0.829 F: 0.960\n",
      "\treco_to_sim:\tP: 0.999 R: 0.829 F: 0.960\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 12:\n",
      "\tclue3d_to_sim:\tP: 0.984 R: 0.277 F: 0.651\n",
      "\ttarget_to_sim:\tP: 0.975 R: 0.616 F: 0.873\n",
      "\treco_to_sim:\tP: 0.737 R: 0.639 F: 0.715\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 13:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.258 F: 0.635\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.757 F: 0.939\n",
      "\treco_to_sim:\tP: 1.000 R: 0.757 F: 0.939\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 14:\n",
      "\tclue3d_to_sim:\tP: 0.976 R: 0.155 F: 0.475\n",
      "\ttarget_to_sim:\tP: 0.776 R: 0.650 F: 0.747\n",
      "\treco_to_sim:\tP: 0.776 R: 0.650 F: 0.747\n",
      "\t|S| = 2 |T| = 1 |R| = 1\n",
      "Event 15:\n",
      "\tclue3d_to_sim:\tP: 0.996 R: 0.209 F: 0.568\n",
      "\ttarget_to_sim:\tP: 0.993 R: 0.758 F: 0.935\n",
      "\treco_to_sim:\tP: 0.993 R: 0.758 F: 0.935\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 16:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.376 F: 0.750\n",
      "\ttarget_to_sim:\tP: 0.997 R: 0.656 F: 0.903\n",
      "\treco_to_sim:\tP: 0.997 R: 0.656 F: 0.903\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 17:\n",
      "\tclue3d_to_sim:\tP: 0.985 R: 0.268 F: 0.642\n",
      "\ttarget_to_sim:\tP: 0.682 R: 0.781 F: 0.700\n",
      "\treco_to_sim:\tP: 0.975 R: 0.758 F: 0.922\n",
      "\t|S| = 2 |T| = 1 |R| = 2\n",
      "Event 18:\n",
      "\tclue3d_to_sim:\tP: 0.989 R: 0.317 F: 0.695\n",
      "\ttarget_to_sim:\tP: 0.987 R: 0.779 F: 0.937\n",
      "\treco_to_sim:\tP: 0.963 R: 0.790 F: 0.922\n",
      "\t|S| = 2 |T| = 5 |R| = 3\n",
      "Event 19:\n",
      "\tclue3d_to_sim:\tP: 0.991 R: 0.277 F: 0.654\n",
      "\ttarget_to_sim:\tP: 0.987 R: 0.499 F: 0.826\n",
      "\treco_to_sim:\tP: 0.986 R: 0.498 F: 0.825\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "-----\n",
      "mean clue3d_to_sim:\tP: 0.993 R: 0.301 F: 0.658\n",
      "mean target_to_sim:\tP: 0.925 R: 0.691 F: 0.848\n",
      "mean reco_to_sim:\tP: 0.905 R: 0.695 F: 0.842\n",
      "mean n_tracksters:\tP: 2.000 R: 2.250 F: 2.150\n",
      " --- Radius threshold: 100 ---\n",
      "Event 0:\n",
      "\tclue3d_to_sim:\tP: 0.988 R: 0.229 F: 0.594\n",
      "\ttarget_to_sim:\tP: 0.981 R: 0.686 F: 0.903\n",
      "\treco_to_sim:\tP: 0.585 R: 0.697 F: 0.605\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 1:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.283 F: 0.664\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.659 F: 0.906\n",
      "\treco_to_sim:\tP: 1.000 R: 0.659 F: 0.906\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 2:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.396 F: 0.766\n",
      "\ttarget_to_sim:\tP: 0.999 R: 0.814 F: 0.956\n",
      "\treco_to_sim:\tP: 0.999 R: 0.814 F: 0.956\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 3:\n",
      "\tclue3d_to_sim:\tP: 0.995 R: 0.383 F: 0.754\n",
      "\ttarget_to_sim:\tP: 0.631 R: 0.792 F: 0.658\n",
      "\treco_to_sim:\tP: 0.631 R: 0.792 F: 0.658\n",
      "\t|S| = 2 |T| = 1 |R| = 1\n",
      "Event 4:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.250 F: 0.625\n",
      "\ttarget_to_sim:\tP: 0.544 R: 0.842 F: 0.586\n",
      "\treco_to_sim:\tP: 0.992 R: 0.833 F: 0.955\n",
      "\t|S| = 2 |T| = 1 |R| = 2\n",
      "Event 5:\n",
      "\tclue3d_to_sim:\tP: 0.996 R: 0.551 F: 0.858\n",
      "\ttarget_to_sim:\tP: 0.995 R: 0.839 F: 0.959\n",
      "\treco_to_sim:\tP: 0.995 R: 0.839 F: 0.959\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 6:\n",
      "\tclue3d_to_sim:\tP: 0.992 R: 0.219 F: 0.582\n",
      "\ttarget_to_sim:\tP: 0.992 R: 0.727 F: 0.925\n",
      "\treco_to_sim:\tP: 0.851 R: 0.738 F: 0.826\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 7:\n",
      "\tclue3d_to_sim:\tP: 0.974 R: 0.112 F: 0.385\n",
      "\ttarget_to_sim:\tP: 0.968 R: 0.216 F: 0.571\n",
      "\treco_to_sim:\tP: 0.923 R: 0.269 F: 0.621\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 8:\n",
      "\tclue3d_to_sim:\tP: 0.997 R: 0.413 F: 0.777\n",
      "\ttarget_to_sim:\tP: 0.996 R: 0.761 F: 0.938\n",
      "\treco_to_sim:\tP: 0.996 R: 0.755 F: 0.936\n",
      "\t|S| = 2 |T| = 2 |R| = 3\n",
      "Event 9:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.177 F: 0.518\n",
      "\ttarget_to_sim:\tP: 0.998 R: 0.562 F: 0.864\n",
      "\treco_to_sim:\tP: 0.704 R: 0.565 F: 0.671\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 10:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.340 F: 0.720\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.599 F: 0.882\n",
      "\treco_to_sim:\tP: 1.000 R: 0.599 F: 0.882\n",
      "\t|S| = 2 |T| = 8 |R| = 8\n",
      "Event 11:\n",
      "\tclue3d_to_sim:\tP: 0.999 R: 0.524 F: 0.846\n",
      "\ttarget_to_sim:\tP: 0.999 R: 0.829 F: 0.960\n",
      "\treco_to_sim:\tP: 0.999 R: 0.829 F: 0.960\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 12:\n",
      "\tclue3d_to_sim:\tP: 0.984 R: 0.277 F: 0.651\n",
      "\ttarget_to_sim:\tP: 0.975 R: 0.616 F: 0.873\n",
      "\treco_to_sim:\tP: 0.737 R: 0.639 F: 0.715\n",
      "\t|S| = 2 |T| = 2 |R| = 1\n",
      "Event 13:\n",
      "\tclue3d_to_sim:\tP: 1.000 R: 0.258 F: 0.635\n",
      "\ttarget_to_sim:\tP: 1.000 R: 0.757 F: 0.939\n",
      "\treco_to_sim:\tP: 1.000 R: 0.757 F: 0.939\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 14:\n",
      "\tclue3d_to_sim:\tP: 0.976 R: 0.155 F: 0.475\n",
      "\ttarget_to_sim:\tP: 0.776 R: 0.650 F: 0.747\n",
      "\treco_to_sim:\tP: 0.776 R: 0.650 F: 0.747\n",
      "\t|S| = 2 |T| = 1 |R| = 1\n",
      "Event 15:\n",
      "\tclue3d_to_sim:\tP: 0.996 R: 0.209 F: 0.568\n",
      "\ttarget_to_sim:\tP: 0.993 R: 0.758 F: 0.935\n",
      "\treco_to_sim:\tP: 0.993 R: 0.758 F: 0.935\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 16:\n",
      "\tclue3d_to_sim:\tP: 0.998 R: 0.376 F: 0.750\n",
      "\ttarget_to_sim:\tP: 0.997 R: 0.656 F: 0.903\n",
      "\treco_to_sim:\tP: 0.997 R: 0.656 F: 0.903\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "Event 17:\n",
      "\tclue3d_to_sim:\tP: 0.985 R: 0.268 F: 0.642\n",
      "\ttarget_to_sim:\tP: 0.682 R: 0.781 F: 0.700\n",
      "\treco_to_sim:\tP: 0.975 R: 0.758 F: 0.922\n",
      "\t|S| = 2 |T| = 1 |R| = 2\n",
      "Event 18:\n",
      "\tclue3d_to_sim:\tP: 0.989 R: 0.317 F: 0.695\n",
      "\ttarget_to_sim:\tP: 0.987 R: 0.779 F: 0.937\n",
      "\treco_to_sim:\tP: 0.963 R: 0.790 F: 0.922\n",
      "\t|S| = 2 |T| = 5 |R| = 3\n",
      "Event 19:\n",
      "\tclue3d_to_sim:\tP: 0.991 R: 0.277 F: 0.654\n",
      "\ttarget_to_sim:\tP: 0.987 R: 0.499 F: 0.826\n",
      "\treco_to_sim:\tP: 0.986 R: 0.498 F: 0.825\n",
      "\t|S| = 2 |T| = 2 |R| = 2\n",
      "-----\n",
      "mean clue3d_to_sim:\tP: 0.993 R: 0.301 F: 0.658\n",
      "mean target_to_sim:\tP: 0.925 R: 0.691 F: 0.848\n",
      "mean reco_to_sim:\tP: 0.905 R: 0.695 F: 0.842\n",
      "mean n_tracksters:\tP: 2.000 R: 2.250 F: 2.100\n"
     ]
    }
   ],
   "source": [
    "r_ranges = [5, 10, 20, 30, 50, 70, 100]\n",
    "results = []\n",
    "\n",
    "max_events = 20\n",
    "clue3D_F = []\n",
    "target_F = []\n",
    "model_reco_F = []\n",
    "for r in r_ranges:\n",
    "    print(f\" --- Radius threshold: {r} ---\")\n",
    "    result = model_evaluation(\n",
    "        cluster_data,\n",
    "        trackster_data,\n",
    "        simtrackster_data,\n",
    "        assoc_data,\n",
    "        model,\n",
    "        decision_th=0.5,\n",
    "        radius=r,\n",
    "        max_events=max_events,\n",
    "        bigT_e_th=10,\n",
    "        pileup=False,\n",
    "        graph=True,\n",
    "        link_prediction=True,\n",
    "    )\n",
    "    model_reco_F.append(np.sum(np.array(result[\"reco_to_sim\"])[:,2]) / max_events)\n",
    "    clue3D_F.append(np.sum(np.array(result[\"clue3d_to_sim\"])[:,2]) / max_events)\n",
    "    target_F.append(np.sum(np.array(result[\"target_to_sim\"])[:,2]) / max_events)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "972e33d3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.657833488850464, 0.657833488850464, 0.657833488850464, 0.657833488850464, 0.657833488850464, 0.657833488850464, 0.657833488850464]\n",
      "[0.8113849590959304, 0.8527976885416955, 0.8617172846489733, 0.8453846514718407, 0.8466663904179444, 0.8483862364330801, 0.8483862364330801]\n",
      "[0.8116899527149318, 0.8419699409854798, 0.8371109602498068, 0.839397226292712, 0.8410772163689512, 0.8422074125891056, 0.8422074125891056]\n"
     ]
    }
   ],
   "source": [
    "print(clue3D_F)\n",
    "print(target_F)\n",
    "print(model_reco_F)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "id": "bc9b49b1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAr4AAAHECAYAAADf1ilPAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8o6BhiAAAACXBIWXMAAA9hAAAPYQGoP6dpAAB2qklEQVR4nO3deVxU9f4/8NeZYYYBhn0HUVAJFwRcEPclSdQ0tW63zL5qeW1T08gWW9TKG1paZnW1BVPvL9PqlpUpZSSWSrig5Urihsoi+7Azy/n9gRwdGRAUmIF5PR+Pecic8znnfGYO4IvPvM/nCKIoiiAiIiIiaudk5u4AEREREVFrYPAlIiIiIqvA4EtEREREVoHBl4iIiIisAoMvEREREVkFBl8iIiIisgoMvkRERERkFRh8iYiIiMgqMPgSERERkVVg8CUiIiIiq2BRwffDDz9EYGAgVCoVoqKisH///nrbfvLJJxg6dChcXV3h6uqK6OjoOu1nzJgBQRCMHmPGjGnpl0FEREREFshigu+WLVsQGxuLxYsXIzU1FeHh4YiJicGVK1dMtk9KSsKUKVOwa9cuJCcnIyAgAKNHj8bly5eN2o0ZMwZZWVnS44svvmiNl0NEREREFkYQRVE0dycAICoqCpGRkfjggw8AAAaDAQEBAZg7dy5efPHFm26v1+vh6uqKDz74ANOmTQNQM+JbVFSErVu33lKfDAYDMjMz4ejoCEEQbmkfRERERNRyRFFESUkJ/Pz8IJM1PKZr00p9alB1dTUOHTqEhQsXSstkMhmio6ORnJzcqH2Ul5dDq9XCzc3NaHlSUhK8vLzg6uqKO++8E0uXLoW7u7vJfVRVVaGqqkp6fvnyZfTo0eMWXhERERERtaaLFy+iQ4cODbaxiOCbl5cHvV4Pb29vo+Xe3t44depUo/bxwgsvwM/PD9HR0dKyMWPG4N5770VQUBDOnDmDl156CWPHjkVycjLkcnmdfcTFxeG1116rs/zixYtwcnJq4qsiIiIiopam0WgQEBAAR0fHm7a1iOB7u5YtW4bNmzcjKSkJKpVKWv7ggw9KX/fq1QthYWHo0qULkpKSMGrUqDr7WbhwIWJjY6XntW+kk5MTgy8RERGRBWtMWapFXNzm4eEBuVyOnJwco+U5OTnw8fFpcNsVK1Zg2bJl+PnnnxEWFtZg286dO8PDwwPp6ekm19va2kohl2GXiIiIqH2xiOCrVCrRt29fJCYmSssMBgMSExMxcODAerd766238MYbbyAhIQH9+vW76XEuXbqE/Px8+Pr6Nku/iYiIiKjtsIjgCwCxsbH45JNPsGHDBpw8eRJPPvkkysrK8MgjjwAApk2bZnTx2/Lly/Hqq69i3bp1CAwMRHZ2NrKzs1FaWgoAKC0txXPPPYc//vgD58+fR2JiIiZOnIiuXbsiJibGLK+RiIiIiMzHYmp8H3jgAeTm5mLRokXIzs5GREQEEhISpAveMjIyjKaoWLNmDaqrq/GPf/zDaD+LFy/GkiVLIJfL8ddff2HDhg0oKiqCn58fRo8ejTfeeAO2trat+tqIiIiIyPwsZh5fS6TRaODs7Izi4mLW+xIRERFZoKbkNYspdSAiIiIiakkMvkRERERkFRh8iYiIiMgqMPgSERERkVWwmFkdiIiIiKht0uZnQFeSV+96G0cPKNw7tmKP6umHuTtARERERG2XNj8D6S+EQNRW1ttGUKjQdXma2cMvSx2IiIiI6JbpSvIaDL0AIGorGxwRbi0c8SUiohbVVj4CJaL2j8GXiIhaTFv6CJToRqIoAqIIGPQQDXpANEBQ2kEQBACAvkIDsapcWica9DVtRT1gMEDpEwxBJgcAVOeeh15zBaJouNrGYLRf+5BhkClq7ixbmfEXqnNOG+/zumM4Rf4DcntnAED56X2oOLu/Tpvaf91GPgEbFx8AQOnxX1D653aj/V7/tceEl6D07goAKDnyI4r2rK/TprbfXv9cDrvAPgCAspO7WvO03BYGXyIiajFN+QiUwdc0URSloCWKIsSqsjohpPa5oFDBxslT2rYq8yREva5OGIJBD5mdE1QdQqW2pccTIeqqTAYduYMb1KF3SW2L930OQ1WpiX4YYOPoAZehM6S2+T+/VxP4DAZArO1rzddyB3d4Tl4stb3yzSJoc89LwfH6ECmzd4H/rM+kttn/72lUXjx6XeCrOT4Megi29gh6+Xep7eVPZqA87Xej49fsVw8IMoS8nyO1vfTBP6FJ3Sq9DzfqHl8F2CgBAFnrn4Dmjy/qPXch/ymE3MEFAJD3w79RtPvTetsGv3sJMjd/AEDRb/Eo2Lm63rb2IUOl4Fv654/I++HNets69pkoBd+KMyko+Ondetu6DJspBd/qnNMoOfB1vW31132KYygrvPa1IEOGxwCUqLzhWJmDjnl/QGbifTQXBl8iIjK7ijMp0JfkSQEGBgNsA8Kg9AwEAOiKslF6/JeaIGJiZMsueBDsOvUGAGgLLqN470YTo3A12zqEjoa65yipbd73S6V1N7Z3jJgA54FTatoWZiLrs8caHIVzj5lf019NLi4sH2Xy+KJBD+eoB+D94NsAakYNTz8baBRgr9/OacAUdHhyU80bZdDj1OOO9b6P6ogJ6PjM99LzM6+EA3qtybYOPUah0wu/SM8vffAPGMqLTLa16zLAKPjmfPkCdIWXTba17dDLKPgWJv4H1dl/m2yr8OxsFHxLD/+AyowjJtvKnb2NnlecT0XF6b0m28pUaqPnuqIsaHPPmmwLwfhyJ9Ggq/c9q1lvgCBtKgcEAZDJIQiymn9l8pp9yuRGwVmudq/54+6GNrVf144MA4DSuyvs7hhivO66f2VKe6mtbUA4nAZMMW5z3dc2aneprX3wILjf/YJRu+u3u/6PT4fuI+Hzfx/U29b2uj+aVF0HAgBO+o9DQu+l0Nj7S+ucyi9jzOFX0P3y9nrf09bE4EtERM3GUFWGqsxTqL6SjuqcdFScTWnUdtkbn6qzzGf6Grjd+QQAoPLycWR+/H/1bu/94Irrgu9FXPn6pXrbymwdpOCrLytA4a619bZVuHeUgq9YXYHSP3+st62qY8S1JwY9qi4drbetcc2zYDRiVodBf13Teq5Jvz5EXcfGyQuirrpOGIJMDhsX3zr9N1SWGAedq/u19e9h1FbdKwb60oK6gU8mh8ItwKit86D/g74k17iPV7eTXxfKAMAtZj50mlwIsqtthdrAJYOgdDBq6zn5NehL82vaCrX7vfr11RHZWt5T3oGhUiPtDzLZ1a9ldd4z3xkfwefh1XVD6dXnwtVyBADwe2wj/B//r+lzcgPvfy6D9z+XNaqt211z4XbX3Ea1dY76J5yj/tmotg7dR8Kh+8hGtVV1DIeqY3ij2ipcfHHSfxy+HBRfZ53GzhdfDorHP/fNRFCj9tayBFEURXN3wlJpNBo4OzujuLgYTk5O5u4OEZHZiaIIfVkBqnPSoc2pCbfqiPGwC+oLANAc+BqXPrjf5LYNfQSq8OoCmUptFEbcxz4Hp8j7AACVF44gZ8tzpsOIIIfzoIfh2OceAEB17jnkfb/UKAzVbFPztTo0BupeowEAOs0VFCauuRaGbhhZswvqB/s7hgAA9BUl0Bz4ysTxa0Ka0quLFBREXTXK0n6T1t3YXu7oKY1miwZDTT1nPSNrgtIOcjsn6f0XqytuaMMJmsi8ys6lYv5BARo7X9N/nIkGOFVkYVU/EQ5BfZr9+E3Jawy+DWDwJSJrJIoioNdBsFEAACovHUPe90tRnZOO6ivpMJQXG7X3nrIS7mNia9peOIILK8dA6dkFSu+uEJT2KNq19qYfgQa9dki6UIbI3ERRhAjAINY89KIofV3zEK8uv/a1QQQMuOH5DetN7gdXy5lFwADj9cbtr99PTR+vHbNmvd5EW1PHFVFP/+vs59p6EYDeIF53vGvrqvUGVBuEht5SAMBL3asR3rlDs5+vpuQ1ljoQEVkh0WCArvCyVJJQG2pr/j0Dr/uWSvWqoq4ampQtRtvbuHWA0qsrlN5dYevfU1qu6hSBkNXZ0vOK86lI/jujTXwE2p4Y6glMdULQdW3rC0Emn6P+EHTT49bp383CV0PHMQ6DpkKo3oCrIbbhMFh7nNr21BQ3D70AUGrr0cL9uDkGXyKidkrU66DNz0D1lTOozkmHbYdQOIQMBQBUXkjFuSWR9W5bnXNa+trW5w54P7gCSu+uUHh1hdKrM2RKu0b1wSACCb2X1jy58SNQQQaIBiT0Xoox13342HAoMg5BTR6JMxiPkN1sP9e2N+6DqRExU+HQuD+Ne216AxoRDk0EvuveF2pZAgCZAMgFQCYIkAm47iFcXX7tudReJkAG43XX9mO8LwEC5LKry2G8Tm7imHWeX3cc4/YNbIsb+3jz48gF4EyRFh/+WXLT981FZf6yHAZfIqI27PqprnTFOcj74d/XRnDzzgF6ndTW7a65UvBVenUB5AooPYOg9O4qjd4qvGpKFJQegdJ2MpUa7mOfbVKfiqsMuFJhwIEib2jsFfU3FmTQ2Ptj1gkROHkFeoa2Flcbwm4MMnVCHG4MOsahx2T4AiCX1YS2m4XBOvu52i+hgTBY57gmjlP/cW94bSZCaGOOKQDSzxzV8HWQ44u0MhRU1j9tmbtKhu5uDfwuaCUMvkREFs5QVY7q3LM1F5RdHb2tLUtw7HcffKasqGkok6Ng5/tG2woKWyg9u0Dh3RW2HcKk5XIHV3T/pByC/Nb+G6jQGXCl3IAr5fqaR4X+2tflelRL//817j86PYRGfb5sHNpMBDPUE5BMtb0ubNU3EmcyTDUwElffCJm8UX1saCTu6n5kpo5Zfx/kJo7J0EbNTSYImNFDjXdSNfW2md5DDZkFfO8x+BIRWQB9hQbVV85Am5MOmYObNN2WriQPf8/xrHe76uw06Wu52h0e97wChXtHaRTXxtW/3qv+Gwq9OoOIvAp9Tbi9IdReKdejRNtwShUAuNvJ4GAj4EKJvsG2ADCvtxO6uSmujcSZDJWwiP84iaiuKF8VYvsA60+UGo38uqtkmN5DjShflRl7dw2DLxFRKxN11cj7cbnRRWV6zRVpvbr3PVLwlavdIVM5AjIZlN7BUkmCVJ7gEyxtJwgCvO57o3F9uK4cIeeGUJtboUdeheGmA7COCgGe9nJ42cvhbS+Hp13Nv172MnjYyWEjE2AQRcz+Nf+mH4EO8LVlqCVq46J8VYj0scXJAi2KKg1wuVreYEk/2wy+RETNRBRF6EtyjWdJqL2wzK/HtdutyhXI+3E5xKoyo+3ljh5QehnPkiAIAoLfvQiZnVOTP6JufDmCaQoZ4GUvh5ddTbi99pDBy04Oe8XNL1RpSx+BEtHtkwkCerorb97QTBh8iYiaQDQYoCvKQvWVdIh6nTQyK4oiTs/zg6442/R21RXS14IgwH1MLASF6toIrlcXyO2dTW5b3/LmKkcwCrZ218Kti62sWepB28pHoETU/jH4EhE1oPC3dai6fKLmorKrI7i1IVbVMQLqNw4DqAmzckdP6DQ5ULgFQHHdTAk1JQl3GO3X697Xb3rs2nKEHBPBtqnlCN5Xg62pcoTW0BY+AiWi9o/Bl4iskqjTojrv/LWZEq7OkiAoVAiY+7XUrmDHSlRlnjDeWCaH0jMICq8uRos7PrujpiZX2fgRzBvLEXKuhtrbKUfwtpfDswnlCK3F0j8CJaL2j8GXiFqNNj8DupK8etfbOHpA4d6x2Y5nqK6ENu8cdJorcOg2XFp+4a3RKDv5K2CoO9uAzM7JaG5cpwFToC/NN7qgTOHeUbqd7/UUbv51lrWVcgQiImvA4EtErUKbn4H0F0IgaivrbSMoVOi6PO2Wwm/psZ2ovHDYaI5bXeElQBQhs3dGyH8KpYAo2CgBgx6C0q6mDKH2pg1X70wGUQSutvWc+EqDx21P5QhERO0dgy8RtQpdSV6DoRcARG0ldCV5dYKvvqyoZnaEq4G2Oicdek0OOj67XWpT8NO7KP1rR519ylSOUHp2gVhdDsHWAQDg8/BqCI98DBsX30aNllpTOQIRUXvG4EtEFiv7/81DcfLn0Jfmm1yvLy2AXO0GAHDoeRdkDq515rmVO3rUCbdKr85Gz1ukHEEKuixHICKyFAy+RNTqDIIMGR4DUKLyhmNlDjrm/QGZWDNsaqgqldqJBp0Uem2cfa7NlHC1NEFQ2Ept3cc8U+/xGipHuFKhRz7LEYiIrAKDLxG1qpP+45DQeyk09tcuBHMqv4wxh19B98vboS8tkJa7jZ4P1xGPQenVBTKVusH9shyBiIhuhsGXiFrNSf9x+HJQfJ3lGjtffDkoHv/cNxNB19X32l53O16WIxAR0e1i8CWiFlOV/TcKEt6FKrAvlJ36IKH30poVwg2jp4IMEA3Y0ftNdCmTo+RyJcsRiIio2TH4ElGzEkUR5X/vQUHCSpQc/h4QRSg8O6Nk+rfQ2PvUv6EgQ4m9L+LOA4DGZJPacgTvq6GW5QhERNQUDL5E1CxEvQ6ag98gP2ElKs/ul5arIybAfewCHIRbo/ajlovo6KxkOQIRETU7Bl8iahZZG2ejKOljAICgsIXz4Olwj3kGtn7dUK41IP1UKYCG5/EFgNhIV97WloiIWgSDLxHdEm1hJgSZHDbO3gAAl8H/h5JD38B11Gy4jXoKNk5euFKuR8KJEvx6sRIVuptV6ALuKhm6u9W9FTAREVFzYPAloiapvHgU+QkrUZy8CW6jnoLP1FUAALvgwQh+5yJkShX+LtTix9RipGRVSRek+avl6OGmwM6M+kd9p/dQQ8ZSBiIiaiEMvkR0U6Ioouz4L8jfsQJlx36WlldlpUEURQiCAIMI7M8Htp8twN9FOqlNmIcC44LsEe6phEwQ0MtDifUnSlFQeW1iXXeVDNN7qBHlq2rV10VERNaFwZeIGqTZ/xVyv1+Kqot/1SwQZHCK/AfcxzwLuy79Ua41YNfFCuw4X47cipowayMDhvipcHeQPTo6Gf+aifJVIdLHFicLtCiqNMDlankDR3qJiKilMfgSUYMqzu5H1cW/INg6wHXYTLjFzIfSMwi55Xp8dUP9rqNSwOhOdhjd0Q4uKnm9+5QJAi9gIyKiVsfgS0SS6tzzKPh5FRz73guHbsMAAG53PQ252h2uIx+H3MEVfxdqsT21GH/cUL97d5A9hvqroJRz5JaIiCwTgy8RoeLsAeTvWAHNga8B0YDqnHQp+CrcA+A67gUcyKnCj0eM63d7eShw93X1u0RERJaMwZfISokGA0qPbEN+wkqUp/0mLXcIHQ230fMA4Gr9bqXJ+t1xQfbo5MRfIURE1Hbwfy0iK3XxvUkoPfJDzRO5As4DH4J7TCxUHcOQW67HNzfW7yqu1u92arh+l4iIyFIx+BJZCZ0mFzKVI2TKminDHCPGo/zv3+A68gm4Rc+Fws3fZP2un4Mc4zuzfpeIiNo+Bl+idq4qKw35Ce+geO9G+Ex9D64jHwMAOA+ZDqcBUwBbdU397l7W7xIRUfvG4EvUDomiiPK035G/Y8W1cgYAZaeSpOBbCQV2Zeux41w+63eJiMgq8H82onZGs/8r5G1/G5XnDkjL1L3vgfvYBbC/Ywhyy/XYcb6c9btERGR1GHyJ2pmiPetRee4ABIUtXIbMgFvMM7D1DcHpQi1+PKxBSnYVDFcLeP0c5Li7sz2GsX6XiIisAIMvURumLcxEwc734RY9Gwq3DgAAj/ELoQqKhNuopyCoPWvqd/cV4u9CrbQd63eJiMgaMfgStUGVGX8h/6d3UJy8CdBrAVEP7wfeAgDY3zEECBqEny5VYsdB4/rdwX4q3M36XSIislL834+ojRBFEWXHdiI/YSXKjv0sLbe/Yyjsu40EAOSW65FwvhyJrN8lIiKqg8GXqA0QRRHnlw5BRfq+mgWCDE6R/4D7mGdh16U/ThdqEZ9azPpdIiKiBjD4ElkofYUGMpUjBEGAIAiwC+qHyot/wnX4v+A2eh5sPAJxILsK21i/S0RE1CiCKIrizZtZJ41GA2dnZxQXF8PJycnc3SErUZ17HgU/r0Lh7k/RacFPsL9jMICaO68JchtUKZ2x61IlEs6V48rV+l25AAzxV2FckB0CnRTm7D4REVGrakpe44gvkYWoOLMf+QkroTnwNSDWBFrNga+k4Fto43a1fjffqH73rqv1u66s3yUiImoQgy+RGYkGA0qPbEP+jhUo//t3ablD6Gi4j3kWDqF31cy/e67cZP3uUH8VbFm/S0RE1CgMvkRmJSJ70zPQ5p4F5Ao4D3wI7jGxUAb0qqnfTS4yqt8NdVdgfGfW7xIREd0KBl+iVqTTXEHR7ni4jYmFTGELQSaH5z2voCo7DW7Rc6Fz8sWvFyuxY1c+63eJiIiaGYMvUSuoykpDfsI7KN67EaK2EjYuvnAZOgMA4DLsEeRV6LH5XDkSD7B+l4iIqKUw+BK1EFEUUZ72O/J3rEDpkR+k5aqgSNi4+AEA0ou02Ha2bv3uuCB7DOvA+l0iIqLmJDN3B6734YcfIjAwECqVClFRUdi/f3+9bT/55BMMHToUrq6ucHV1RXR0dJ32oihi0aJF8PX1hZ2dHaKjo3H69OmWfhlEMFSV4dxrUbgQN7wm9AoC1L3vQaeXfkOnRX/guMcwvLqvEC/vLURyVk3oDXVX4MVIZ6wc7oa7Otkx9BIRETUziwm+W7ZsQWxsLBYvXozU1FSEh4cjJiYGV65cMdk+KSkJU6ZMwa5du5CcnIyAgACMHj0aly9fltq89dZbWL16NdauXYuUlBQ4ODggJiYGlZWVrfWyyIqIumsXoclsHSC3d4agUMF15OPoEncSnnO+xW5lX8xPKsA7qRr8XaiFXACGd1Bh+VBXvDrAFb29bHnRGhERUQuxmBtYREVFITIyEh988AEAwGAwICAgAHPnzsWLL7540+31ej1cXV3xwQcfYNq0aRBFEX5+fnj22WexYMECAEBxcTG8vb2xfv16PPjggzfdJ29gQY2hLbiMgp2rUbRnPbq88SdsXHwAAFXZf0Nu74oihRt2nCtH4sVK1u8SERE1szZ3A4vq6mocOnQICxculJbJZDJER0cjOTm5UfsoLy+HVquFm5sbAODcuXPIzs5GdHS01MbZ2RlRUVFITk5uVPAlakhlxl/IT1iJ4j82AXodAKBo33/hMe45AMBFVRC2pZcjJTuf9btEREQWwCKCb15eHvR6Pby9vY2We3t749SpU43axwsvvAA/Pz8p6GZnZ0v7uHGftetuVFVVhaqqKum5RqNp9Gsg6yCKIsqO7UT+jhUoO75TWm4fMgzuYxfAPmwcUrIq8eO5CqRx/l0iIiKLYhHB93YtW7YMmzdvRlJSElQq1S3vJy4uDq+99loz9ozaG31ZAS6+NxGithIQZHDqfz/cxzwLdOyLXRcrsWN3IeffJSIislAWEXw9PDwgl8uRk5NjtDwnJwc+Pj4NbrtixQosW7YMv/zyC8LCwqTltdvl5OTA19fXaJ8REREm97Vw4ULExsZKzzUaDQICApr6cqgd0ZcVouTwD3AZMg0AYKN2h1v0HIh6LdxGz4NG3RH/O1+BXxKN59+N7mSHGNbvEhERWRSLCL5KpRJ9+/ZFYmIiJk2aBKDm4rbExETMmTOn3u3eeust/Pvf/8ZPP/2Efv36Ga0LCgqCj48PEhMTpaCr0WiQkpKCJ5980uT+bG1tYWtr2yyvidq26txzKPhpFQp/i4dYVQZbv+6w6xwJAPB+8G2kF2mx+Szrd4mIiNoSiwi+ABAbG4vp06ejX79+6N+/P1atWoWysjI88sgjAIBp06bB398fcXFxAIDly5dj0aJF2LRpEwIDA6W6XbVaDbVaDUEQMH/+fCxduhTBwcEICgrCq6++Cj8/PylcE92o4sx+5O9YAc3B/wFiTcmCbYdeMFSXwyCKOJBdZbJ+9+4ge0R4sX6XiIjIkllM8H3ggQeQm5uLRYsWITs7GxEREUhISJAuTsvIyIBMdm3a4TVr1qC6uhr/+Mc/jPazePFiLFmyBADw/PPPo6ysDI899hiKioowZMgQJCQk3FYdMLVP2oLLuLxmCsr//l1a5hA6Gu5jF0AWcid2X6rC9l35rN8lIiJqwyxmHl9LxHl82zdRFCFcHaEVdVqcfq4LdMXZcB74ENxjYlHq2RMJ5yuQmFGB8qv1u+qr8++yfpeIiMgytLl5fIlak05zBQWJ/0Hp4R8QtOgPCDYKCDYK+D/xOZReXXBB8MRH58rxxzHW7xIREbUnDL5kNaqy0pCf8A6K926AqK2Zr7kkdSuc+t8PgyjimFMkfjxZgbTCQmkb1u8SERG1Hwy+1K6JoojytN+Qv2MlSo/8IC1XBUXCY9xzsAmfhB3nyrH9XLlR/e5gPxXuDrJDoDPrd4mIiNoLBl9q16ouHcOFuBE1TwQBjr3vgduYZ1EeMBDfX6hEYlIR63eJiIisBIMvtSv6ihJUnN0Pdc9RAABVQC849IqB0iMIbmOewUVVED49V44/kgqk+l1fBznuZv0uERFRu8fgS+2CtuASCnauRmHSxxC1VQh+JwM2Tp4AgA6x23Ewpxprztat3x0XZI/erN8lIiKyCgy+1KZVZvyJ/B0rUZzyBaDXAQCUviHQ5mdAa++OpIuV2HG+HDnlrN8lIiKydgy+1CZVZaUh+79zUHb8F2mZfbfhcB+7AJXBMfhfRhUSE/ON63c72mF0oB3cWL9LRERklRh8qU2S2zmjPO03QCaHU+T9cB/7LC67hmPduXL8sbuQ9btERERUB4MvmY02PwO6krx619s4ekDh3hH6skIU/roWVZkn4P/4f2vWufjAb9YG2HYZiL/0PvjkbDlOnWD9LhEREdWPwZfMQpufgfQXQiBqK+ttI9jYwmnAQ9Ac+BJiVRkAwH3Ms1B1ikCFzoB93vdgx7Fy5JQXA2D9LhERETWMwZfMQleS12DoBQBRV4XiPZ8BAGwDwuA+5lmUuHXD1ydLkZhRwfpdIiIiahIGXzI7gyBDhscAlKi84ViZg455f0Am1szCoOo6EF6TX0O2/3BsPF+B5N81RvW744LsMMzfDiobljMQERFRwxh8yaxO+o9DQu+l0Nj7S8ucyi9jzOFXEJKZgMyJn2KDxhenLhRJ63u6K3A363eJiIioiRh8yWxO+o/Dl4Pi6yzX2Pniy0Hr4FCZi7KLHgC0rN8lIiKi28bgS2ZhEIGE3ktrnggy45VXn5fZecFepkdMkCPrd4mIiOi2MfiSWaSV20Jj73nTdk90KEBUN99W6BERERG1d7KbNyFqfsW6xn3rlRv4LUpERETNg6mCzMK++Hyj2rk5OrZsR4iIiMhqMPhSq9OXFsD5m8fhVH4ZEMV6WolwU4roFeRfz3oiIiKipmHwpVaX/fk8GAov4o6ig0C905EJmBHqzOnKiIiIqNnw4jZqde53v4CMglL86T8BEAF7G0G6CxsAuKtkmN5DjShflRl7SURERO0Ngy+1OrlvT3w7+FNoS/QI91TihX5OOFWoQ1GlAS4qGbq7KTjSS0RERM2OwZdahajXoerycag6hmNzWikulOjhqBTwZJgj5DIZerorzd1FIiIiaudY40utIu+HN3F2cV/s3f45tp2rAAA8EeYEV96UgoiIiFoJgy+1uPL0P5D73esot3HCRgwFANzV0Q79vG3N3DMiIiKyJgy+1KL0FSW4/NHDEA16JIzegiJRBT8HOf6vh9rcXSMiIiIrw+BLLSpn03xor5zBX6FzcNQ+HHIBmNvbCbZyXrxGRERErYvBl1qM5uA3KPptHfIdu2B7z5cBAA+GOKCzs8LMPSMiIiJrxOBLLUKnyUXWulnQCzb4fvQ3qBZlCHVXYHxne3N3jYiIiKwUpzOjFiF39IDnfW/gy/NAhtwHDgoBT4U7cX5eIiIiMhsGX2oRgiAgO+JfSKosAgA83ssR7nacuoyIiIjMh6UO1Kyqr5yFvkKDMq0BHxzRQAQwsoOKtx8mIiIis+OILzUbQ3UlLr43CfqqUmybtAv5lXbwsZdjRk9OXUZERETmxxFfajZXvn4JVZeO4rD7SKQU20F2deoylQ2/zYiIiMj8mEioWZQe24mCn95FoUMnbA9fCgD4R7ADurpw6jIiIiKyDAy+dNt0pfnI/GQGDIIc34/+HypFObq5KjC5K6cuIyIiIsvBGl+6LaIoIuuzx6ArysTeqDicVwTAzkbA7AhOXUZERESWhSO+dFuK9/4XJQe/wUXPAdjV6REAwL9CHeFlz6nLiIiIyLIw+NJtUYeNgU2f+/HdiP8HEQKG+NliiD+nLiMiIiLLw+BLt8XGyQs7h3+CfMERnnYyzAx1NHeXiIiIiExi8KVbUnnpOABgX2YlfrtcBQHAnAgn2Cv4LUVERESWiSmFmqw8PRlnXwnDsU/m4NOjJQCAe7vao5ub0sw9IyIiIqofgy81ib6iBJfXPgwDgE1uD6FMJ6Kriw3uDXYwd9eIiIiIGsTgS02S/f+ehjb3LP7o8zLOqLpCJRcwN8IJNjJOXUZERESWjcGXGk1z4GsU71mPTLcI/NplNgDgkZ5q+DhwOmgiIiKyfAy+1CjagsvI/OwxVMvtsfXOzdBDwABfWwzvwKnLiIiIqG3gUB3dlCiKyPz0ERjKCvHLiPXIlbnCXSXDrFBHCLw7GxEREbURDL50U4IgwH3sszgidsABr7EQADwV7gS1kh8YEBERUdvB4EuNUh0cje979wG0IiZ0tkeoB6cuIyIioraFQ3ZUL0N1JarzLsAgivjPnxqUakUEOdnggRBOXUZERERtD0d8qV5XvnoRRb9/hmNTEnC0tAuUMmBub05dRkRERG0TR3zJpNKjP6Pg5/eQaRuAb0uDAADTejjCX82/lYiIiKhtYvClOnQlecj8ZDq0cjtsvXMLdJChn7cS0R05dRkRERG1XRy+IyOiKCJr3SzoirOROGQtcmw84WIrw+O9nDh1GREREbVpHPElI0W741GSuhV/+8cgxW8yAODJcEc42fJbhYiIiNo2phmSVOecQfbn81Bq64ltgz8CAIwLskOEp62Ze0ZERER0+1jqQBIbV3+43PkENgujUQI7dHSUY0qI2tzdIiIiImoWDL4kkSlV+GvgGzh1vBQKGTC3tzOUctb1EhERUfvAUgdCde55iHodLpbo8N+TpQCAqd3U6OjIv4uIiIio/WCysXL6Cg0uLL8TcOmAjwd/Ca1BhghPJcYE2pm7a0RERETNiiO+Vi77/z0Nbe457HCfjIxyGZyUAp4Mc+TUZURERNTuWFTw/fDDDxEYGAiVSoWoqCjs37+/3rbHjx/Hfffdh8DAQAiCgFWrVtVps2TJEgiCYPTo1q1bC76CtkWz/ysU79mAM94jsLfjVADAE2FOcFHJzdwzIiIiouZnMaUOW7ZsQWxsLNauXYuoqCisWrUKMTExSEtLg5eXV5325eXl6Ny5M+6//34888wz9e63Z8+e+OWXX6TnNjYW85LNSltwCZnrH0e50g3fD4sHANzV0Q59vTl1GRERtS69Xg+tVmvubpAFUyqVkMluf7zWYlLgO++8g1mzZuGRRx4BAKxduxY//vgj1q1bhxdffLFO+8jISERGRgKAyfW1bGxs4OPj0zKdbqNEgwGXP5kOfVkhfrxrKzSCGn4OcvxfD05dRkRErUcURWRnZ6OoqMjcXSELJ5PJEBQUBKVSeVv7sYjgW11djUOHDmHhwoXSMplMhujoaCQnJ9/Wvk+fPg0/Pz+oVCoMHDgQcXFx6Nixo8m2VVVVqKqqkp5rNJrbOralKvh5FcpP/IrDwY/ihOtAyAXg6d5OsOXUZURE1IpqQ6+Xlxfs7e15fQmZZDAYkJmZiaysLHTs2PG2vk8sIvjm5eVBr9fD29vbaLm3tzdOnTp1y/uNiorC+vXrERISgqysLLz22msYOnQojh07BkdHxzrt4+Li8Nprr93y8doKda8xSD+0Cz9FvAEAmBKiRpCzwsy9IiIia6LX66XQ6+7ubu7ukIXz9PREZmYmdDodFIpbzywWdXFbcxs7dizuv/9+hIWFISYmBtu3b0dRURG+/PJLk+0XLlyI4uJi6XHx4sVW7nHrkPt2x9Zh61ENG4S6K3B3Z05dRkRErau2ptfe3t7MPaG2oLbEQa/X39Z+LGLE18PDA3K5HDk5OUbLc3JymrU+18XFBXfccQfS09NNrre1tYWtbfu9uKs67wKUHp3w5d9lOKvRw0EhYHaEE2T8aImIiMyE5Q3UGM31fWIRI75KpRJ9+/ZFYmKitMxgMCAxMREDBw5stuOUlpbizJkz8PX1bbZ9thWlfyUg/bku2Lv1Y3x/phwA8HgvR7hx6jIiIiKyEhYx4gsAsbGxmD59Ovr164f+/ftj1apVKCsrk2Z5mDZtGvz9/REXFweg5oK4EydOSF9fvnwZR44cgVqtRteuXQEACxYswIQJE9CpUydkZmZi8eLFkMvlmDJlinlepJnoNLm4/OkMVMjV2Ci/EyKAkQEqRPmqzN01IiIiolZjMcH3gQceQG5uLhYtWoTs7GxEREQgISFBuuAtIyPDaP62zMxM9O7dW3q+YsUKrFixAsOHD0dSUhIA4NKlS5gyZQry8/Ph6emJIUOG4I8//oCnp2ervjZzEkURmev+BV1xDnbcuRlFMif42Msxg1OXERERkZURRFEUzd0JS6XRaODs7Izi4mI4OTmZuzu3pHDXx8ha/zj+7DwFW/utglwAXh/kiq4unMWBiIjMp7KyEufOnUNQUBBUqqZ/AqnNz4CuJK/e9TaOHlC4m56+9HaNGDECERERJu8aaw5N7U9ycjIGDRqEcePG4ccffzTZZvjw4fjtt9+MlslkMhQXF0Otbv3Bs4a+X5qS1yxmxJeaX1X238je9AwKHDphR7+3AQD33+HA0EtERG2aNj8D6S+EQNRW1ttGUKjQdXlai4Xf21VdXX3bN2O4VfHx8ZgyZQq++eYbZGZmws/Pz2i9KIo4fPgwVqxYgalTp0rLZTKZWUJvc7KIi9uo+Yk6LS6vfRh6bRW+G/k5qqBANzcFJnbhtDFERNS26UryGgy9ACBqKxscEb5VM2bMwO7du/Hee+9BEAQIgoDz588jISEBQ4YMgYuLC9zd3TF+/HicOXNG2m7EiBGYM2cO5s+fDw8PD8TExAAASkpKMHXqVDg4OMDX1xfvvvsuRowYgfnz50vbGgwGxMXFISgoCHZ2dggPD8fXX3/dYH/qU1paii1btmD+/PkYOXIk1q9fX6fN6dOnUVJSgmHDhsHHx0d6eHl5Nct7aE4Mvu2V3Aauw2diT9/FyLAPhr2NgDnhnLqMiIgsm6GqrP5HdcNh91b221TvvfceBg4ciFmzZiErKwtZWVkICAhAWVkZYmNjcfDgQSQmJkImk2Hy5MkwGAzSths2bIBSqcTevXuxdu1aADUX9+/duxfff/89du7cid9//x2pqalGx4yLi8PGjRuxdu1aHD9+HM888wwefvhhKfCa6k99vvzyS/j4+KB///6YOnUq1q1bhxurXg8dOgQbGxuEhYU1+f2xdCx1aKcEQcCV8EeRVF4IAPhXL0d42nPqMiIismynHqv/o3R1+Dh0jDVdk3ozp58NhN7ECHCPDU271MnZ2RlKpRL29vZG9xq47777jNqtW7cOnp6eOHHiBEJDQwEAwcHBeOutt6Q2JSUl2LBhAzZt2oRRo0YBAD777DOj0oOqqiq8+eab+OWXX6QpXjt37ow9e/bgo48+wqZNm0z2pz7x8fFS+cKkSZPw+OOPY/fu3RgxYoTUJjU1FXq93uiOeuHh4di7dy8AYNu2bXj22WdhMBjwwgsv4F//+led4wQGBsLJyQkymQyurq7YtWvXTfvWGhh82xl9hQYQRVQpHPH+kWKIAIb622KwH6cuIyIiaimnT5/GokWLkJKSgry8PGmkNyMjQwq+ffv2Ndrm7Nmz0Gq16N+/v7TM2dkZISEh0vP09HSUl5fjrrvuMtq2urraaHarxkhLS8O+ffuk8ga1Wo2JEyciPj6+TvCdMmUKXnvtNWmZg4MDAECn0yE2Nha7du2Cs7Mz+vbti8mTJ5u87fS+ffssriaYwbedyd44G+V//47tkxKRW+EITzsZHu3paO5uERERNUq3j0vrXync+ieXwSvP3/K2jVF734BPPvkEfn5+MBgMCA0NRXV1tdSmNjw2RWlpzfvx448/wt/f32hdU+82Gx8fj8jISAQHB0vLpk6divvvvx8ffPABnJ2dAdQE37i4OOm+CNfbv38/evbsKfVl7Nix+Pnnn9vMPRJY49uOFP+xGcX7/h9S7ftin8YRAoA5EU6wV/A0ExFR2yCzdaj/obz1Ty/r2+etUCqV0Ov10vP8/HykpaXhlVdewahRo9C9e3cUFhbedD+dO3eGQqHAgQMHpGXFxcX4+++/pec9evSAra0tMjIy0LVrV6NHbS3vjf0xRafTYePGjXjooYeMlo8ePRr29vb44osvANSMQhcVFdU7mpyZmWkUwP39/XH58uU67QRBwPDhwxEZGYnPP//8Ju9E6+GIbzuhzc9A1oYnUGTfATuiVgEA7g22Rzc380yVQkRE1F4FBgYiJSUF58+fh1qthpubG9zd3fHxxx/D19cXGRkZePHFF2+6H0dHR0yfPh3PPfcc3Nzc4OXlhcWLF0Mmk0G4ejG6o6MjFixYgGeeeQYGgwFDhgxBcXEx9u7dCycnJ0yfPt1kf66/6RdQU5ebk5OD0NBQHDt2zGjdsGHDEB8fjyeeeAKHDh2CXC6/7Qvb9uzZA39/f2RlZSE6Ohq9evWyiIvlGHzbAdGgx+WPp0FXUYLvxnyDCsEWwS42uK/rrf0lS0REZMlsHD0gKFQ3ncfXxtGjRY6/YMECTJ8+HT169EBFRQXOnTuHzZs34+mnn0ZoaChCQkKwevVqo7rZ+rzzzjt44oknMH78eDg5OeH555/HxYsXjW7S8MYbb8DT0xNxcXE4e/YsXFxc0KdPH7z00kv19icwMNDoOPHx8QBQp1b4en/99RdSU1PRrVs32NnZmWzj5+dnNMJ7+fJloxrlWrWjwr6+vhg3bhxSU1MtIvjyzm0NaCt3bsv78S1c+fIF/B76HH7tsQAquYDlQ13h48C/a4iIyDK15Tu3taSysjL4+/tj5cqVmDlzprm7U4dOp0P37t2RlJQkXdy2b98+o4vbysrKYDAY4OjoiNLSUgwfPhxr165FZGTkLR+Xd24jAEDFhcO48r9XcNmtN5J6xAIAHg1VM/QSEVG7pnDv2CaD7Y0OHz6MU6dOoX///iguLsbrr78OAJg4caKZe2aajY0NVq5ciZEjR8JgMOD555+XQm9ERASOHDmCnJwcTJ48GQCg1+sxa9as2wq9zYnpqI1TuAVAGXEvtga8DANkGOhri2H+nLqMiIiorVixYgXS0tKgVCrRt29f/P777/DwaJkyjeZwzz334J577qmz/MiRIwBqLtr7888/W7lXjcPg28bZOHogccTHyLtYCXeVDP/q5SgVxBMREZFl6927Nw4dOmTublgNznPVRmmLsgAAKVmV+PViJQQAsyOcoObUZUREREQmccS3DdJpruDsqxHQ9boXH3equavKPV3s0dOdU5cRERER1YfDg22MKIrIjJ8JnSYXnzvcg1Id0NnZBv+8g1OXERERETWEwbeNKdz1EUqPbMMf3WbjjEtf2MqBuRFOsJGxrpeIiIioIQy+bUhVVhpyvohFtktPJIbVTFo9vYcj/NSsWCEiIiK6GQbfNkLUVePy2qmo1gPfDvsv9JCjn7cSdwZw6jIiIiKixuBQYRtx5dslqDx/CL/0fxdXVP5wsZXh8TAnTl1GRERE1EgMvm2EQ487sf/UWewPfAgA8FS4I5yUHLAnIiIiaiwG3zZC12Ukvo8KA7TA3UF2CPe0NXeXiIiIiNoUDhlaOJ3mCkRRxJq/SqDRAh0dbfBgiNrc3SIiIrIIBlHE8fxq7L1cieP51TCIYqscNzs7G/PmzUPXrl2hUqng7e2NwYMHY82aNSgvLwcAzJgxA4IgYNmyZUbbbt26tU6pYlPa3rjdpEmTGlwvCAIEQYBSqUTXrl3x+uuvQ6fT3fQ1JicnQxAE3H333SbXDx8+XNp37UMul6O0tPSm+zYXjvhasOLkL5C1/nGceOAHHKnoDoUMeLq3E5Ry1vUSERGlZFVi/YlSFFQapGVuKhlm9FAjyrflLv4+e/YsBg8eDBcXF7z55pvo1asXbG1tcfToUXz88cfw9/fHPffcAwBQqVRYvnw5Hn/8cbi6uja436a0bYoxY8bgs88+Q1VVFbZv347Zs2dDoVBg4cKFDW4XHx+PKVOm4JtvvkFmZib8/PykdaIo4vDhw1ixYgWmTp0qLZfJZFCrLXeAjiO+Fqo67wKyNj6JbKU//lceDAB4uLsaAY78W4WIiCglqxLvpGqMQi8AFFQa8E6qBilZlS127Keeego2NjY4ePAg/vnPf6J79+7o3LkzJk6ciB9//BETJkyQ2kZHR8PHxwdxcXE33W9T2jaFra0tfHx80KlTJzz55JOIjo7G999/3+A2paWl2LJlC+bPn4+RI0di/fr1RutPnz6NkpISDBs2DD4+PtLDy8urWfve3Jot+B47dgwbN27EoUOHmmuXVkObn4GK86nSo/zsAVxaPRnVlZX4duhn0Ak26O2pREwnO3N3lYiIqEWIoohKXeMe5VoDPjve8Mfp60+UolxraNT+xCaUR+Tn5+Pnn3/G7Nmz4eBg+q6p15cmyOVyvPnmm3j//fdx6dKlBvfdlLa3w87ODtXV1Q22+fLLL+Hj44P+/ftj6tSpWLdundH7dOjQIdjY2CAsLKzF+tkSbnn4cNSoUUhMTAQAbNq0CW+++SbGjRuH1atXY8aMGZgzZ06zdbI90+ZnIP2FEIjaun+Z/hLxOrIdOsO+Kg//ClBDEFxav4NEREStoEoPTP8pt9n2V1BpwCM/5zWq7YYYT6gamYjS09MhiiJCQkKMlnt4eKCysub/8tmzZ2P58uXSusmTJyMiIgKLFy9GfHx8g/tvStumEkURiYmJ+OmnnzB37twG28bHx0slDJMmTcLjjz+O3bt3Y8SIEQCA1NRU6PV6uLu7S9uEh4dj7969AIBt27bh2WefhcFgwAsvvIB//etfJo8TGBgIJycnyGQyuLq6YteuXc3wSut3y8G3oKBA+vq9997DL7/8Ah8fH5SWlmLQoEEMvo2kK8mTQq9BkCHDYwBKVN7Q2Pkg5Y7HAQAT98+DQ5/lADqasadERERUn/3798NgMGDq1Kmoqqqqs3758uW48847sWDBgpvuqyltG2Pbtm1Qq9XQarUwGAx46KGHsGTJknrbp6WlYd++fVJ5g1qtxsSJExEfH28UfKdMmYLXXntN2q52BFyn0yE2Nha7du2Cs7Mz+vbti8mTJxuF5Ovt27ev1eqCbzn4iqKIiooKGAwGGAwG+Pj4AKh5c+RyebN10Fqc9B+HhN5LobH3N1reJfMX3JH1i5l6RURE1Dps5TUjr41xsqAayw4U37Tdi5HO6O6mbNSxG6tr164QBAFpaWlGyzt37gygpozAlGHDhiEmJgYLFy7EjBkzGjxGU9o2xsiRI7FmzRoolUr4+fnBxqbh+BcfH4/IyEgEBwdLy6ZOnYr7778fH3zwAZydnZGamoq4uDh07dq1zvb79+9Hz5494e9fk2nGjh2Ln3/+GVOmTLnt13K7bjn4FhUVoWfPnhBFEYIgICsrC76+vigtLW1SrQzVhN4vB5n4OEMUccb3Tpz0H4eg1u8WERFRqxEEodHlBuGeSripZHUubLueu0qGcE8lZM18h1N3d3fcdddd+OCDDzB37tx663xNWbZsGSIiIuqUSdxu25txcHAwGVBN0el02LhxI1588UWj5aNHj4a9vT2++OILjB49GkVFRejdu7fJfWRmZkqhFwD8/f1x+fJlk20FQcDw4cMhk8kwf/58oxkiWsItB9/z58+bXC6TyfDtt9/e6m6tjkEEEnovrXki3HCtoSAAooiE3ksxhn9MEBERAQBkgoAZPdR4J1VTb5vpPdTNHnpr/ec//8HgwYPRr18/LFmyBGFhYZDJZDhw4ABOnTqFvn37mtyuV69emDp1KlavXn3TYzSlbXFxMY4cOWK0zN3dHQEBAY16Pdfbtm0bcnJyEBoaimPHjhmtGzZsGOLj4+Hu7g65XN4sF7bt2bMH/v7+yMrKQnR0NHr16tWiF8w1+9xY9vb2CAri+GRjpZXbQmPfwEc7ggwae3+kleeiT+t1i4iIyKJF+aoQ2wd15vF1V8kwvYXn8e3SpQsOHz6MN998EwsXLsSlS5dga2uLHj16YMGCBXjqqafq3fb111/Hli1bGnWcxrZNSkqqM/o6c+ZMfPrpp406zvVqL6i766676m2zadMmdOvWrd6yDj8/P6MR3suXL6N///4m29aODPv6+mLcuHFITU1t0eAriC1Ql9CnTx+kpqY2925bnUajgbOzM4qLi+Hk5NQix9j150msvWS62Pt6T3TIx8jw7i3SByIiotZWWVmJc+fOISgoCCrVrYdUgyjiZIEWRZUGuKhk6O6maLGRXmocnU6H7t27IykpSbq4bd++fXUubisrK4PBYICjoyNKS0sxfPhwrF27FpGRkXX22dD3S1PyWovcDaE9hN7W4mxTf33SrbQjIiKyJjJBQE/3m1/ARq3HxsYGK1euxMiRI2EwGPD8888bhd6IiAgcOXIEOTk5mDx5MgBAr9dj1qxZJkNvs/btdjYuLCzEzz//LA1n+/n5ISYmpllvs9fe9fR2hNOJTGjsfOrW+AKAaIBTRTZ6eru1fueIiIiIbsE999wj3bb5RrX1yJ07d8aff/7Zir26jTu3xcfHY+DAgUhJSZGmNEtJScGgQYOafcLl9szWoyMeifCouZANN1adiIAg4JEID9h6cA5fIiIiottxyyO+b731FlJTU+tM4/HGG2+gT58+mDlz5m13zloMCu4AubrSRIG+vMUL9ImIiIisxS0HX0EQUFJSUif4lpSUGN2jmhonyleFSB9bFugTERERtZBbDr4rVqzA8OHDERoaKk1FcenSJRw/fhwrV65stg5aExboExEREbWcRgff8vJy2NvbS8/Hjx+PsWPHYv/+/cjMzARQc3Fb//79ectiIiIiIrI4jQ6+zs7OyMrKgoeHh7RMLpdj4MCBLdIxIiIiIqLm1OhZHfR6PQyGaxdeDR06FDk5OS3SKSIiIiKi5nbL05kdOXIEZWVlzdkXIiIiIqIWc8vBl4iIiIioLWlS8N20aRNSU1Oh1Wpbqj9ERERERC2i0Re3DR06FIsXL0ZJSQkUCgV0Oh0WL16MoUOHonfv3ggLC4OtrW1L9pWIiIioXRgxYgQiIiKwatUqc3fFqjQ6+O7evRsAcPr0aRw6dAipqalITU3Fiy++iKKiItjY2KB79+6tfs9lIiIiorbmm2++gUKhaJVjJScnY9CgQRg3bhx+/PHHOuuHDx+O3377zWiZTCZDcXEx1Gp1q/SxtTT5BhbBwcEIDg7Ggw8+KC07d+4cDh06hC+++KJZO0dERERkir7sIgyVufWul6m8IHfo0Io9aho3N7dWO1Z8fDymTJmCb775BpmZmfDz85PWiaKIw4cPY8WKFZg6daq0XCaTtbvQCzTDxW0lJSXYuXMn3nrrLWzdurUZukRERERUP1FfhbwfIpH3Q98GHpEQ9VUtcvwRI0bg6aefxvPPPw83Nzf4+PhgyZIl0vqEhAQMGTIELi4ucHd3x/jx43HmzJk6+5g/fz4A4OOPP4afn5/RtLEAMHHiRDz66KMAAIPBgLi4OAQFBcHOzg7h4eH4+uuvb9rX0tJSbNmyBfPnz8fIkSOxfv16o/WnT59GSUkJhg0bBh8fH+nh5eXV9DemDbjl4Pvbb79h+vTp8PHxwSuvvIKAgIDm7BcRERGRaTIl5A4dUX+MkUHuEADIlC3WhQ0bNsDBwQEpKSl466238Prrr2Pnzp0AgLKyMsTGxuLgwYNITEyETCbD5MmT6wTbWvfffz/y8/Oxa9cuaVlBQQESEhKkUdi4uDhs3LgRa9euxfHjx/HMM8/g4YcflkpR6/Pll1/Cx8cH/fv3x9SpU7Fu3TqIoiitP3ToEGxsbBAWFna7b0mb0KTgm52djWXLliE4OBjjxo2DTqfDV199hczMTLz22mst1UciIiJq50RRhEFb1qiHqCuHQ9jLAEwHScAAh7CXIerKG7e/64JgY4WFhWHx4sUIDg7GtGnT0K9fPyQmJgIA7rvvPtx7773o2rUrIiIisG7dOhw9ehQnTpwwuS9XV1eMHTsWmzZtkpZ9/fXX8PDwwMiRI1FVVYU333wT69atQ0xMDDp37owZM2bg4YcfxkcffdRgP+Pj46XwPGnSJGRlZRmF5dTUVOj1eri7u0OtVkOtVmPw4MHS+m3btiEkJATBwcH49NNP6z1OYGAgwsLCEBERgZEjR978DTSTRtf4TpgwAYmJiRg5ciSWLFmCSZMmwcHBQVovCEKLdJCIiIjaP1FXjpzPm6+mtGjXpEa39Z5aCkHhcPOG17lxhNTX1xdXrlwBUFM+sGjRIqSkpCAvL08a6c3IyEBoaKjJ/U2dOhWzZs3Cf/7zH9ja2uLzzz/Hgw8+CJlMhvT0dJSXl+Ouu+4y2qa6uhq9e/eut49paWnYt2+fVN6gVqsxceJExMfHY8SIEQBqgu+UKVOMBjBr851Op0NsbCx27doFZ2dn9O3bF5MnT4a7u7vJ4+3bt8/i64IbHXx//PFHPPTQQ5g/fz769evXkn0iIiIismg3zsggCIIUcCdMmIBOnTrhk08+kWp3Q0NDUV1dXe/+JkyYAFEU8eOPPyIyMhK///473n33XQA1dbpATRbz9/c32q6hqWTj4+MRGRmJ4OBgadnUqVNx//3344MPPoCzszNSU1MRFxeHrl271tl+//796Nmzp3TMsWPH4ueff8aUKVMaemssWqOD7759+xAfH48777wTvr6+mDp1KqZOnYouXbq0ZP+IiIjICgg29vCeWtqkbURRREHCcOgK/wREPSDIYeMaDrcxu5v0SbRgY9/U7tYrPz8faWlp+OSTTzB06FAAwJ49e266nUqlwr333ovPP/8c6enpCAkJQZ8+fQAAPXr0gK2tLTIyMjB8+PBG9UOn02Hjxo148cUXjZaPHj0a9vb2+OKLLzB69GgUFRXVO2qcmZlpFLT9/f1x+fJlk20FQcDw4cMhk8kwf/58oxkiLEmjg++AAQMwYMAArFq1Clu2bMG6devw2muvITIyElOnTkXPnj1bsp9ERETUjgmC0ORyAwBw6vsmCnaOqXki6uHU903Ileb7uN3V1RXu7u74+OOP4evri4yMjDrhsz5Tp07F+PHjcfz4cTz88MPSckdHRyxYsADPPPMMDAYDhgwZguLiYuzduxdOTk6YPn16nX1t27YNOTk5CA0NxbFjx4zWDRs2DPHx8XB3d4dcLm+WC9v27NkDf39/ZGVlITo6Gr169bLIC+aaPI+vg4MDHn30UTz66KNIS0tDfHw83nzzTeTk5LDOl4iIiFqV0m80FO6R0OYfgMI9Ekq/0Wbtj0wmw+bNm/H0008jNDQUISEhWL16tVRT25A777wTbm5uSEtLw0MPPWS07o033oCnpyfi4uJw9uxZuLi4oE+fPnjppZdM7is+Ph4A6tQFX2/Tpk3o1q0b7OzsTK738/MzGuG9fPky+vfvb7Jt7ciwr68vxo0bh9TUVIsMvoJ4K5cy3kCv1+OHH37AunXr8P333zdHvyyCRqOBs7MziouL4eTkZO7uEBERtRuVlZU4d+4cgoKCoFKpbmtfVZm/oDjlaThHrYatX3Qz9ZB0Oh26d++OpKQk6eK2ffv21bm4raysDAaDAY6OjigtLcXw4cOxdu1aREZGNltfGvp+aUpea/KIrylyuRyTJk3CpEmTmmN3RERERI1m6xcNr8mmpwqjW2djY4OVK1di5MiRMBgMeP7556XQGxERgSNHjgAAcnJyMHnyZAA1g6GzZs1q1tDbnJplxLe94ogvERFRy2jOEV9q/5prxPe2b1lMRERERNQWMPgSERERkVWwqOD74YcfIjAwECqVClFRUdi/f3+9bY8fP4777rsPgYGBEAQBq1atuu19EhEREVH7ZTHBd8uWLYiNjcXixYuRmpqK8PBwxMTESLf/u1F5eTk6d+6MZcuWwcfHp1n2SURERETtl8UE33feeQezZs3CI488gh49emDt2rWwt7fHunXrTLaPjIzE22+/jQcffLDe2/U1dZ9ERERE1H5ZRPCtrq7GoUOHEB19be49mUyG6OhoJCcnt9o+q6qqoNFojB5ERERE1D5YRPDNy8uDXq+Ht7e30XJvb29kZ2e32j7j4uLg7OwsPQICAm7p2ERERERkeSwi+FqKhQsXori4WHpcvHjR3F0iIiIiombSLHduu10eHh6Qy+XIyckxWp6Tk1PvhWstsU9bW9t664WJiIiIqG2ziBFfpVKJvn37IjExUVpmMBiQmJiIgQMHWsw+iYiIiEaMGIH58+eb7fgzZszApEmTLKY/bYlFjPgCQGxsLKZPn45+/fqhf//+WLVqFcrKyvDII48AAKZNmwZ/f3/ExcUBqLl47cSJE9LXly9fxpEjR6BWq9G1a9dG7ZOIiIiorfvmm2+gUCjM3Y02wWKC7wMPPIDc3FwsWrQI2dnZiIiIQEJCgnRxWkZGBmSyawPUmZmZ6N27t/R8xYoVWLFiBYYPH46kpKRG7ZOIiIiorXNzczN3F9oMiyh1qDVnzhxcuHABVVVVSElJQVRUlLQuKSkJ69evl54HBgZCFMU6j9rQ25h9EhEREd0KnU6HOXPmwNnZGR4eHnj11VchiiIA4L///S/69esHR0dH+Pj44KGHHjK6eVZhYSGmTp0KT09P2NnZITg4GJ999pm0/uLFi/jnP/8JFxcXuLm5YeLEiTh//ny9fbmx1CEwMBBvvvkmHn30UTg6OqJjx474+OOPjbZp6jHaC4sKvkRERGTdDAZDg4/mbnurNmzYABsbG+zfvx/vvfce3nnnHXz66acAAK1WizfeeAN//vkntm7divPnz2PGjBnStq+++ipOnDiBHTt24OTJk1izZg08PDykbWNiYuDo6Ijff/8de/fuhVqtxpgxY1BdXd3o/q1cuRL9+vXD4cOH8dRTT+HJJ59EWlpasx6jLbKYUgciIiKi2ut3TFGr1QgMDJSenzx5UhplvZG9vT06d+4sPU9LS4Ner6/TLjQ09Jb6GRAQgHfffReCICAkJARHjx7Fu+++i1mzZuHRRx+V2nXu3BmrV69GZGQkSktLoVarkZGRgd69e6Nfv34AYPSatmzZAoPBgE8//RSCIAAAPvvsM7i4uCApKQmjR49uVP/GjRuHp556CgDwwgsv4N1338WuXbsQEhLSbMdoizjiS0RERNREAwYMkEIjAAwcOBCnT5+GXq/HoUOHMGHCBHTs2BGOjo4YPnw4gJrrlQDgySefxObNmxEREYHnn38e+/btk/bz559/Ij09HY6OjlCr1VCr1XBzc0NlZSXOnDnT6P6FhYVJXwuCAB8fH6ncormO0RZxxJeIiIgsRo8ePRrdtnv37o1uGxIScivdabLKykrExMQgJiYGn3/+OTw9PZGRkYGYmBipjGDs2LG4cOECtm/fjp07d2LUqFGYPXs2VqxYgdLSUvTt2xeff/55nX17eno2uh83zvIgCIJU2tFcx2iLGHyJiIjIYlw/g5O52jZGSkqK0fM//vgDwcHBOHXqFPLz87Fs2TIEBAQAAA4ePFhne09PT0yfPh3Tp0/H0KFD8dxzz2HFihXo06cPtmzZAi8vLzg5OTVrn2u1xjEsFUsdiIiIiJooIyMDsbGxSEtLwxdffIH3338f8+bNQ8eOHaFUKvH+++/j7Nmz+P777/HGG28Ybbto0SJ89913SE9Px/Hjx7Ft2zZp9Hrq1Knw8PDAxIkT8fvvv+PcuXNISkrC008/jUuXLjVL31vjGJaKwZeIiIioiaZNm4aKigr0798fs2fPxrx58/DYY4/B09MT69evx1dffYUePXpg2bJlWLFihdG2SqUSCxcuRFhYGIYNGwa5XI7NmzcDqLko77fffkPHjh1x7733onv37pg5cyYqKyubbXS2NY5hqQSxvsshCRqNBs7OziguLm733whEREStqbKyEufOnUNQUBBUKpW5u0MWrqHvl6bkNY74EhEREZFVYPAlIiIiIqvA4EtEREREVoHBl4iIiIisAoMvEREREVkFBl8iIiIyG04uRY3RXN8nDL5ERETU6mpvqVteXm7mnlBbUHu7Z7lcflv74S2LiYiIqNXJ5XK4uLjgypUrAGpuqiAIgpl7RZbIYDAgNzcX9vb2sLG5vejK4EtERERm4ePjAwBS+CWqj0wmQ8eOHW/7jyMGXyIiIjILQRDg6+sLLy8vaLVac3eHLJhSqYRMdvsVugy+REREZFZyufy2azeJGoMXtxERERGRVWDwJSIiIiKrwOBLRERERFaBwZeIiIiIrAKDLxERERFZBQZfIiIiIrIKDL5EREREZBUYfImIiIjIKjD4EhEREZFVYPAlIiIiIqvA4EtEREREVoHBl4iIiIisAoMvEREREVkFBl8iIiIisgoMvkRERERkFRh8iYiIiMgqMPgSERERkVVg8CUiIiIiq8DgS0RERERWgcGXiIiIiKwCgy8RERERWQUGXyIiIiKyCgy+RERERGQVGHyJiIiIyCow+BIRERGRVWDwJSIiIiKrwOBLRERERFaBwZeIiIiIrAKDLxERERFZBQZfIiIiIrIKDL5EREREZBUYfImIiIjIKjD4EhEREZFVYPAlIiIiIqvA4EtEREREVoHBl4iIiIisAoMvEREREVkFBl8iIiIisgoMvkRERERkFRh8iYiIiMgqMPgSERERkVVg8CUiIiIiq8DgS0RERERWwaKC74cffojAwECoVCpERUVh//79Dbb/6quv0K1bN6hUKvTq1Qvbt283Wj9jxgwIgmD0GDNmTEu+BCIiIiKyUBYTfLds2YLY2FgsXrwYqampCA8PR0xMDK5cuWKy/b59+zBlyhTMnDkThw8fxqRJkzBp0iQcO3bMqN2YMWOQlZUlPb744ovWeDlEREREZGEEURRFc3cCAKKiohAZGYkPPvgAAGAwGBAQEIC5c+fixRdfrNP+gQceQFlZGbZt2yYtGzBgACIiIrB27VoANSO+RUVF2Lp16y31SaPRwNnZGcXFxXBycrqlfRARERFRy2lKXrOIEd/q6mocOnQI0dHR0jKZTIbo6GgkJyeb3CY5OdmoPQDExMTUaZ+UlAQvLy+EhITgySefRH5+fr39qKqqgkajMXoQERERUftgEcE3Ly8Per0e3t7eRsu9vb2RnZ1tcpvs7Oybth8zZgw2btyIxMRELF++HLt378bYsWOh1+tN7jMuLg7Ozs7SIyAg4DZfGRERERFZChtzd6AlPfjgg9LXvXr1QlhYGLp06YKkpCSMGjWqTvuFCxciNjZWeq7RaBh+iYiIiNoJixjx9fDwgFwuR05OjtHynJwc+Pj4mNzGx8enSe0BoHPnzvDw8EB6errJ9ba2tnBycjJ6EBEREVH7YBHBV6lUom/fvkhMTJSWGQwGJCYmYuDAgSa3GThwoFF7ANi5c2e97QHg0qVLyM/Ph6+vb/N0nIiIiIjaDIsIvgAQGxuLTz75BBs2bMDJkyfx5JNPoqysDI888ggAYNq0aVi4cKHUft68eUhISMDKlStx6tQpLFmyBAcPHsScOXMAAKWlpXjuuefwxx9/4Pz580hMTMTEiRPRtWtXxMTEmOU1EhEREZH5WEyN7wMPPIDc3FwsWrQI2dnZiIiIQEJCgnQBW0ZGBmSyazl90KBB2LRpE1555RW89NJLCA4OxtatWxEaGgoAkMvl+Ouvv7BhwwYUFRXBz88Po0ePxhtvvAFbW1uzvEYiIiIiMh+LmcfXEnEeXyIiIiLL1ubm8SUiIiIiamkMvkRERERkFRh8iYiIiMgqMPgSERERkVVg8CUiIiIiq8DgS0RERERWgcGXiIiIiKwCgy8RERERWQUGXyIiIiKyCgy+RERERGQVGHyJiIiIyCow+BIRERGRVWDwJSIiIiKrwOBLRERERFaBwZeIiIiIrAKDLxERERFZBQZfIiIiIrIKDL5EREREZBUYfImIiIjIKjD4EhEREZFVYPAlIiIiIqvA4EtEREREVoHBl4iIiIisAoMvEREREVkFG3N3gAB92UUYKnPrXS9TeUHu0KEVe0RERETU/jD4mpmor0LeD5EwVObU20am8oHX/echyG1bsWdERERE7QtLHcxNpoTcoSPqPxUyyB0CAJmyNXtFRERE1O5wxLcRDAYDDAaDyXUymcyoXUPqa+vQeykKf510raFogCBW1baEQ++lEEURoii2WB8sta0gCBAEod23bej8tsW2wLXz3J7bAm3j56g9t7W0n2X+juDviOvbAm3j56itt71Z++sx+DbCqVOnoFar6yxXq9UIDAyUnp88ebLeHwZ7e3t07txZep6Wlga9Xn/1mR/QZ7+0Tl52DC4np9Q8EeQ4l6uEoeiEyf3a2toiODhYen7mzBlUVVWZbKtQKBASEiI9P3fuHCoqKky2lcvl6N69u/T8/PnzKC8vN9lWEAT07NlTep6RkYHS0lKTbQEgNDRU+vrSpUvQaDT1tu3Ro4f0izQzMxNFRUX1tu3WrRtsbGq+pbOzs1FQUFBv2zvuuANKZc0o+pUrV5CXl1dv265du0KlUgEAcnNzkZtbfz12586dYW9vDwDIz89HTk79JSyBgYHS91VBQQGysrLqbdupUyc4OjoCAIqKinD58uV62wYEBMDZ2RkAoNFocPHixXrb+vv7w9XVFQBQWlqKCxcu1NvW19cX7u7uAICysjKcP3++3rbe3t7w9PQEAFRUVODs2bP1tvX09IS3tzcAoKqqCunp6fW29fDwgI+PDwBAq9Xi77//rretm5sb/Pz8AAB6vR6nTp2qt62Liws6dKipoxdFESdOmP55AwAnJyd07NhRet5Q2+b7HWHMzs4OXbp0kZ6fPn0aWq3WZFv+jriGvyNq8HdEDf6OqNEefkdkZmbW2/ZGDL6WTtQDBtPfgACg05xBQeJzsHHqArljF4iGgQAUrdc/IiIiojZCEG82Xm/FNBoNnJ2dUVhYCCcnJ5NtmmsYXxRF5O8YDl3hEcCggwAdbFzD4TzoU2hLzkFfeg76krPQlZ6DoeQs9OWXANEAQIRgqLy2H5kKgAAIMsjtAyBz7AwbdRDkTl1gow6Cwqkz5I6dIVOoLeYjioZY2seN/BiTH2Ne3xawnI/6rLWtpf0s83cEf0dc3xZoGz9Hbb2tRqOBq6sriouL681rtRh8G1AbfBvzRjaHqss/oWDnGOm5210JsPWPMdlW1FdDX3YBOs0Z6EtqHrqS9Kv/ngX0lSa3qyVTedeEYceakWK547WvZSpP6ZcXERERkSVrSl5jqYMFUfqNhsI9Etr8A1C4R0LpN7retoJcCRunYNg4BddZJ4oGGMqzoCu5PhRf+1esKoChMgeGyhxor+yru2+Fo1EQvj4cyx0CIMjkzfJ6OX8xERERtSYGXwsiCAIc+76J4pSn4dj3zVsedRUEGeQO/pA7+AM+w+qsN1QV3TBCfAZ6Tc2/hvJLELUl0BUcga7gSN2dyxSQqwONR4qdutb8qw6CYGPXqD5y/mIiIiJqbQy+FsbWLxpek+u/ArQ5yGxdILPtC4VH3zrrRF0ldKXn6o4Ua85AX3oWMGih15yGXnPa9L7t/euWT1wtqZDZul3XsGb+4poRX1O1PJy/mIiIiJoXgy8ZEWxUULh0h8Kle511okEPffmla6FYYxyORa0GhvLLqC6/DOT8VnffSperobgr5E5doPAeAm3+gXp6YoBjnzdYa0xERETNhsGXGk2QyWGj7gQbdSfA906jdaIoQqzKN6ol1mvSpeeGimyI1UXQ5h+CNv/QzY4EuXMPKHzuvEk7IiIiosZj8KVmIQgCBJUHlCoPwDOqznqDtgz60rNSLXFtONYVHoWh4sZJ2UXoi48jZ5MzFO59ofTsD4VHFBSeUZA7dOQoMBEREd0SBl9qFTKFA2SuvaBw7WW0XBRF5G2LhK7g8NV5iQUICjVEEYCuBNore6C9sufaflReUHhGQeHRH0rPKCjcIyGzdWnV10JERERtE4MvmZUgCHDq8+/r5i8W4TriKyj97oK++G9U56VAm7cf2twUaAv+hKHyCqou/oCqiz9I+5A7hVw3KtwfCtdwCHJeFEdERETGeAOLBrT2DSyslSiKyN8WJc1f7D4+xWQ5g6irgLbgCLS5KajO2w9tXgr0JSbu7y6zhcK997VRYY/+kDt2YYkEERFRO9SUvMbg2wAG39ZTlfkLilOehnPUatj6RTd6O0NlXk0Izq0JwtV5+yFWFdRpJ9i6QelxbVRY6dEfMpVHc74EIiIiMgMG32bC4Nv2iKIIfcmZa6PCuSnQFhwGDNV12sodO0PhESWNCivcekOwUZmh10RERHSrGHybCYNv+yDqq6Et/PPaqHDufug1aXUbCjZQuIVfu3jOIwpy5zsgCLLW7zQRERE1CoNvM2Hwbb8MVYXQ5h9EdW5KzahwXsrVu8gZExTOUHhEXi2PuDqlmp23GXpMREREpjD4NhMGX+shiiL0pRdqZpC4OiqszT8E6CvqtJU7dDSaUs3GrQ9kCgcz9JqIiIgYfJsJg691Ew1a6AqP10ypllszrZqu6ASAG35kBDlsXEKv1Qp7RsHGuTsEmdws/SYiIrImDL7NhMGXbmSo1tTcdvm6KdUM5Zl12gk2aig8+klBWOnRH3KHDmboMRERUfvG4NtMGHypMfRll6+OCtcEYW3eQYi60jrtZPZ+0kVzNaUS/SBTOJqhx0RERO0Hg28zYfClWyEa9NAVn5TKI6pzU6ArOgaI+htaCrBx6WE0Kmzj2guCjDdUJCIiaiwG32bC4EvNxaAtg64gteaiuaujw/qyC3Ubyu2gcO9ztV746l3n1J141zkiIqJ6MPg2EwZfakn68mxo8w/U1AvnpkCbdwCitrhOO5nKy2hUWOHRHzJbl9bvMBERkQVi8G0mDL7UmkTRAH3x39JFc9rc/dAWHAFEXZ22cqc7ro0Ke/aHwjUcglzZ+p0mIiIyMwbfZsLgS+Ym6iqhLThSM7fw1Vsw60vO1G0oU0Lh1vvaqLBnFOSOXVgiQURE7R6DbzNh8CVLZKjMQ3XegWsXz+WlQKwqqNNOsHW7WhoRdfXOc/0hU3mYocdEREQth8G3mTD4UlsgiiL0JWev3XEuLwXa/MOAoapOW7ljZyg8ro0KK9wiINjYmaHXREREzYPBt5kw+FJbJeqroS38y2hKNb0mrW5DwQYKt/DrLp6Lgtz5DgiCrPU7TUREdAsYfJsJgy+1J4aqImjzD1w3pVoKDJVX6rQTFM5QeEReLY+4OqWavY8ZekxERHRzDL7NhMGX2jNRFKEvy4D26kVz1bkp0OYfAvQVddrKHTpKo8IKj/5QuPeFTOFghl4TEREZY/BtJgy+ZG1Egw66wmPSRXPa3P3QFR0HcMOvCUEOG5fQa6PCnv1h49wDgkze4P71ZRdhqMytd71M5QW5Q4dmeCVkSXjeiaglMfg2EwZfIsCgLYE279DVi+dqaoYN5ZfrtBNsHKBw7yeNCis9o4zCjKivwpWvOsFQmVPvsWQqH3jdfx6C3LZFXgu1Pp53ImppTclrFnUFy4cffojAwECoVCpERUVh//79Dbb/6quv0K1bN6hUKvTq1Qvbt283Wi+KIhYtWgRfX1/Y2dkhOjoap0+fbsmXQNTuyBSOsPUdAXWvF+B25zfw/ucleN1/Ca4jv4FD6AtQ+oyEYKOGqCtDdc5ulB17C0VJ/8CVrwKQ86U/Cn6djNK/lqEqZw9k9n6o/9eODHKHAEDGG3G0KzIl5A4dwfNORJbAYkZ8t2zZgmnTpmHt2rWIiorCqlWr8NVXXyEtLQ1eXl512u/btw/Dhg1DXFwcxo8fj02bNmH58uVITU1FaGgoAGD58uWIi4vDhg0bEBQUhFdffRVHjx7FiRMnoFKpbtonjvgSNY5o0ENXfEq641x1Xgp0hUcBUd+k/TgN/AhKr8HXFtz0Bhz1rxcaWHfzfTew7jb6ZIn7bfB9aqY+VWUnofj3h+tt6Rq9HaoOY29yLCIi09pkqUNUVBQiIyPxwQcfAAAMBgMCAgIwd+5cvPjii3XaP/DAAygrK8O2bdukZQMGDEBERATWrl0LURTh5+eHZ599FgsWLAAAFBcXw9vbG+vXr8eDDz540z4x+BLdOlFXDm1+qlQeoc1Ngb7sgrm7RZZKkAMyJQSZApApIEhfGy8z/ve69XIlBOFqO7kSEK7+K1NAuP5rU9s28hj1b9twbTsRtaym5DWbVupTg6qrq3Ho0CEsXLhQWiaTyRAdHY3k5GST2yQnJyM2NtZoWUxMDLZu3QoAOHfuHLKzsxEdHS2td3Z2RlRUFJKTk00G36qqKlRVXZv0X6PR3M7LIrJqgo09lN5DoPQeIi3TV+SgPG0tSo8sqdte4VwTTmrd5G9y8cYL7m5YW/+qm/2t30LbmmG/Nx/XuNU+3cJ+RUPDnwCIekBfAfHqrCIWMSLTaEKTg7kgXF0mUxiH9FsI5g0etzHreWtzagZt5SJWiwi+eXl50Ov18Pb2Nlru7e2NU6dOmdwmOzvbZPvs7Gxpfe2y+trcKC4uDq+99totvQYiujm5nTfU4YtQdfFHaAtSa8KOIIfCrQ/cx6fwP+B2TBRF5G+LMjrvNm7hcBudCIhawKCFaNAChuqaf/XVEGuX66sB8eq/Bm3Ncn21cXuDFqLB9DLjf69fb7rdzbaFaLjx1V3dV3Xts7ZFsDEeZb/J6LkU2OsJ140L3LcR5q/vH0fbLYKor0LeD5Ft4iJWiwi+lmLhwoVGo8gajQYBAQFm7BFR+yMIAhz7vIGCnWNqFoh6OPZ5g6G3nTN13p36vAm5rYtZ+3UrRIP+WgC/STBvfOC+/UBe77a1/bu6vO4L0gF6HUR9RRsM7bJGB/OmjYTfRqhvZNkNBJv283vv6kWsNSO+N/5hCFjSRawWEXw9PDwgl8uRk2P8l0JOTg58fEzfMcrHx6fB9rX/5uTkwNfX16hNRESEyX3a2trC1pbT6RC1NKXfaCjcI6HNPwCFeySUfqPN3SVqBe3lvAsyOSCTQ8DNL5K2NKIoAqKuUaPn9YXr2mXNNsrehNH9OuPpogEQqyAaasoU21xwv8V6c9Oh/hbCfG0wv5XR/etubV/nD9s6DBYzwGERwVepVKJv375ITEzEpEmTANRc3JaYmIg5c+aY3GbgwIFITEzE/PnzpWU7d+7EwIEDAQBBQUHw8fFBYmKiFHQ1Gg1SUlLw5JNPtuTLIaKbEAQBjn3fRHHK03Ds+6ZF/DKklsfzbn6CIEhBpy2SRttNhOtbKWtpVKi/xbKbG0fqIerqvqDa46O87YV2QW4UrmtKZpRX/0Axbqdw62Mxf+haRPAFgNjYWEyfPh39+vVD//79sWrVKpSVleGRRx4BAEybNg3+/v6Ii4sDAMybNw/Dhw/HypUrcffdd2Pz5s04ePAgPv74YwA1P9zz58/H0qVLERwcLE1n5ufnJ4VrIjIfW79oeE0+Ye5uUCvjeafb0bZH2w2AQXdLwbzBUN/IYN74kXpT/ao28YL0gF4PUV95kxduWeVsFhN8H3jgAeTm5mLRokXIzs5GREQEEhISpIvTMjIyIJNdG1YfNGgQNm3ahFdeeQUvvfQSgoODsXXrVmkOXwB4/vnnUVZWhsceewxFRUUYMmQIEhISGjWHLxEREVFzEQRZTemA3Px1rk1VUyJz42h73UBu0FejeM806Iv/BmCwuNFewILm8bVEnMeXiIiIqPGqLv9kVOvrdlcCbP1jWvSYbfaWxURERETUdtVexArAIi9iZfAlIiIiomZRexGr3Lm7RV7EajE1vkRERETU9lnyRawc8SUiIiIiq8DgS0RERERWgcGXiIiIiKwCgy8RERERWQUGXyIiIiKyCgy+RERERGQVGHyJiIiIyCow+BIRERGRVWDwJSIiIiKrwOBLRERERFaBwZeIiIiIrIKNuTtgyURRBABoNBoz94SIiIiITKnNabW5rSEMvg0oKSkBAAQEBJi5J0RERETUkJKSEjg7OzfYRhAbE4+tlMFgQGZmJhwdHSEIgrm7065pNBoEBATg4sWLcHJyMnd3qBXwnFsnnnfrw3NufVr7nIuiiJKSEvj5+UEma7iKlyO+DZDJZOjQoYO5u2FVnJyc+IvRyvCcWyeed+vDc259WvOc32yktxYvbiMiIiIiq8DgS0RERERWgcGXLIKtrS0WL14MW1tbc3eFWgnPuXXiebc+POfWx5LPOS9uIyIiIiKrwBFfIiIiIrIKDL5EREREZBUYfImIiIjIKjD4EhEREZFVYPClVvXbb79hwoQJ8PPzgyAI2Lp1q9F6URSxaNEi+Pr6ws7ODtHR0Th9+rR5OkvNIi4uDpGRkXB0dISXlxcmTZqEtLQ0ozaVlZWYPXs23N3doVarcd999yEnJ8dMPabbtWbNGoSFhUmT1w8cOBA7duyQ1vN8t3/Lli2DIAiYP3++tIznvf1ZsmQJBEEwenTr1k1ab4nnnMGXWlVZWRnCw8Px4Ycfmlz/1ltvYfXq1Vi7di1SUlLg4OCAmJgYVFZWtnJPqbns3r0bs2fPxh9//IGdO3dCq9Vi9OjRKCsrk9o888wz+OGHH/DVV19h9+7dyMzMxL333mvGXtPt6NChA5YtW4ZDhw7h4MGDuPPOOzFx4kQcP34cAM93e3fgwAF89NFHCAsLM1rO894+9ezZE1lZWdJjz5490jqLPOcikZkAEL/99lvpucFgEH18fMS3335bWlZUVCTa2tqKX3zxhRl6SC3hypUrIgBx9+7doijWnGOFQiF+9dVXUpuTJ0+KAMTk5GRzdZOamaurq/jpp5/yfLdzJSUlYnBwsLhz505x+PDh4rx580RR5M95e7V48WIxPDzc5DpLPecc8SWLce7cOWRnZyM6Olpa5uzsjKioKCQnJ5uxZ9SciouLAQBubm4AgEOHDkGr1Rqd927duqFjx4487+2AXq/H5s2bUVZWhoEDB/J8t3OzZ8/G3XffbXR+Af6ct2enT5+Gn58fOnfujKlTpyIjIwOA5Z5zG7MdmegG2dnZAABvb2+j5d7e3tI6atsMBgPmz5+PwYMHIzQ0FEDNeVcqlXBxcTFqy/Peth09ehQDBw5EZWUl1Go1vv32W/To0QNHjhzh+W6nNm/ejNTUVBw4cKDOOv6ct09RUVFYv349QkJCkJWVhddeew1Dhw7FsWPHLPacM/gSUauZPXs2jh07ZlQDRu1TSEgIjhw5guLiYnz99deYPn06du/ebe5uUQu5ePEi5s2bh507d0KlUpm7O9RKxo4dK30dFhaGqKgodOrUCV9++SXs7OzM2LP6sdSBLIaPjw8A1LniMycnR1pHbdecOXOwbds27Nq1Cx06dJCW+/j4oLq6GkVFRUbted7bNqVSia5du6Jv376Ii4tDeHg43nvvPZ7vdurQoUO4cuUK+vTpAxsbG9jY2GD37t1YvXo1bGxs4O3tzfNuBVxcXHDHHXcgPT3dYn/WGXzJYgQFBcHHxweJiYnSMo1Gg5SUFAwcONCMPaPbIYoi5syZg2+//Ra//vorgoKCjNb37dsXCoXC6LynpaUhIyOD570dMRgMqKqq4vlup0aNGoWjR4/iyJEj0qNfv36YOnWq9DXPe/tXWlqKM2fOwNfX12J/1lnqQK2qtLQU6enp0vNz587hyJEjcHNzQ8eOHTF//nwsXboUwcHBCAoKwquvvgo/Pz9MmjTJfJ2m2zJ79mxs2rQJ3333HRwdHaXaLmdnZ9jZ2cHZ2RkzZ85EbGws3Nzc4OTkhLlz52LgwIEYMGCAmXtPt2LhwoUYO3YsOnbsiJKSEmzatAlJSUn46aefeL7bKUdHR6luv5aDgwPc3d2l5Tzv7c+CBQswYcIEdOrUCZmZmVi8eDHkcjmmTJliuT/rZptPgqzSrl27RAB1HtOnTxdFsWZKs1dffVX09vYWbW1txVGjRolpaWnm7TTdFlPnG4D42WefSW0qKirEp556SnR1dRXt7e3FyZMni1lZWebrNN2WRx99VOzUqZOoVCpFT09PcdSoUeLPP/8sref5tg7XT2cmijzv7dEDDzwg+vr6ikqlUvT39xcfeOABMT09XVpviedcEEVRNFPmJiIiIiJqNazxJSIiIiKrwOBLRERERFaBwZeIiIiIrAKDLxERERFZBQZfIiIiIrIKDL5EREREZBUYfImIiIjIKjD4ElG7FRgYiFWrVjW6/fnz5yEIAo4cOVJvm/Xr18PFxeW2+2bKkiVLEBER0SL7vpkRI0Zg/vz5rXKsG9/npKQkCIKAoqKiFj1ufHw8Ro8e3aLHWLt2LSZMmNCixyCiW8fgS0QWZcaMGRAEAcuWLTNavnXrVgiC0KR9HThwAI899lhzdo9awKBBg5CVlQVnZ+cWO0ZlZSVeffVVLF68uMWOAQCPPvooUlNT8fvvv7focYjo1jD4EpHFUalUWL58OQoLC29rP56enrC3t2+mXrUsrVZr7i40SXV1dbPtS6lUwsfHp8l/2DTF119/DScnJwwePLjFjgHUvJaHHnoIq1evbtHjENGtYfAlIosTHR0NHx8fxMXFNdhuz549GDp0KOzs7BAQEICnn34aZWVl0vobSx1OnTqFIUOGQKVSoUePHvjll18gCAK2bt1qtN+zZ89i5MiRsLe3R3h4OJKTk+sce+vWrQgODoZKpUJMTAwuXrxotH7NmjXo0qULlEolQkJC8N///tdovSAIWLNmDe655x44ODjg3//+t7Tuv//9LwIDA+Hs7IwHH3wQJSUl0rqqqio8/fTT8PLygkqlwpAhQ3DgwAGjfe/evRv9+/eHra0tfH198eKLL0Kn00nry8rKMG3aNKjVavj6+mLlypUNvs/AtTKMTz/9FEFBQVCpVACAhIQEDBkyBC4uLnB3d8f48eNx5swZo23379+P3r17Q6VSoV+/fjh8+LDR+htLHUyVfKxatQqBgYFG2/Tv3x8ODg5wcXHB4MGDceHChXr7v3nzZpMlCOvWrUPPnj2l92rOnDnSOkEQ8NFHH2H8+PGwt7dH9+7dkZycjPT0dIwYMQIODg4YNGhQndc7YcIEfP/996ioqKi3P0RkHgy+RGRx5HI53nzzTbz//vu4dOmSyTZnzpzBmDFjcN999+Gvv/7Cli1bsGfPHqPgcj29Xo9JkybB3t4eKSkp+Pjjj/Hyyy+bbPvyyy9jwYIFOHLkCO644w5MmTLFKDiWl5fj3//+NzZu3Ii9e/eiqKgIDz74oLT+22+/xbx58/Dss8/i2LFjePzxx/HII49g165dRsdZsmQJJk+ejKNHj+LRRx+VXtfWrVuxbds2bNu2Dbt37zYq+3j++efxv//9Dxs2bEBqaiq6du2KmJgYFBQUAAAuX76McePGITIyEn/++SfWrFmD+Ph4LF26VNrHc889h927d+O7777Dzz//jKSkJKSmpjZ0SgAA6enp+N///odvvvlGqs8tKytDbGwsDh48iMTERMhkMkyePBkGgwEAUFpaivHjx6NHjx44dOgQlixZggULFtz0WA3R6XSYNGkShg8fjr/++gvJycl47LHHGhwx3rNnD/r162e0bM2aNZg9ezYee+wxHD16FN9//z26du1q1OaNN97AtGnTcOTIEXTr1g0PPfQQHn/8cSxcuBAHDx6EKIp1vuf69esHnU6HlJSU23qdRNQCRCIiCzJ9+nRx4sSJoiiK4oABA8RHH31UFEVR/Pbbb8Xrf2XNnDlTfOyxx4y2/f3330WZTCZWVFSIoiiKnTp1Et99911RFEVxx44doo2NjZiVlSW137lzpwhA/Pbbb0VRFMVz586JAMRPP/1UanP8+HERgHjy5ElRFEXxs88+EwGIf/zxh9Tm5MmTIgAxJSVFFEVRHDRokDhr1iyjvt1///3iuHHjpOcAxPnz5xu1Wbx4sWhvby9qNBpp2XPPPSdGRUWJoiiKpaWlokKhED///HNpfXV1tejn5ye+9dZboiiK4ksvvSSGhISIBoNBavPhhx+KarVa1Ov1YklJiahUKsUvv/xSWp+fny/a2dmJ8+bNE+uzePFiUaFQiFeuXKm3jSiKYm5urghAPHr0qCiKovjRRx+J7u7u0jkRRVFcs2aNCEA8fPiwKIqiuGvXLhGAWFhYKB0rPDzcaL/vvvuu2KlTJ6m/AMSkpKQG+1KrsLBQBCD+9ttvRsv9/PzEl19+ud7tAIivvPKK9Dw5OVkEIMbHx0vLvvjiC1GlUtXZ1tXVVVy/fn2j+kdErYcjvkRksZYvX44NGzbg5MmTddb9+eefWL9+PdRqtfSIiYmBwWDAuXPn6rRPS0tDQEAAfHx8pGX9+/c3edywsDDpa19fXwDAlStXpGU2NjaIjIyUnnfr1g0uLi5SP0+ePFmnlnTw4MF1XseNI5BATXmGo6Oj0fFrj33mzBlotVqjfSsUCvTv39/o2AMHDjQa/Rw8eDBKS0tx6dIlnDlzBtXV1YiKipLWu7m5ISQkxOR7cb1OnTrB09PTaNnp06cxZcoUdO7cGU5OTlI5QkZGhtSfsLAwqTQCAAYOHHjTYzXEzc0NM2bMQExMDCZMmID33nsPWVlZ9bavLTm4vg9XrlxBZmYmRo0a1eCxrv9e8Pb2BgD06tXLaFllZSU0Go3RdnZ2digvL2/8iyKiVsHgS0QWa9iwYYiJicHChQvrrCstLcXjjz+OI0eOSI8///wTp0+fRpcuXW7ruAqFQvq6NkDWfnTfnBwcHBo8du3xW+LYt8JUfydMmICCggJ88sknSElJkT7ev52L32QyGURRNFp248V/n332GZKTkzFo0CBs2bIFd9xxB/744w+T+3N3d4cgCEYXS9rZ2TWqL6a+Fxrz/VFQUFDnjwQiMj8GXyKyaMuWLcMPP/xQ5wKzPn364MSJE+jatWudh1KprLOfkJAQXLx4ETk5OdKyGy8KayydToeDBw9Kz9PS0lBUVITu3bsDALp37469e/cabbN371706NHjlo5Xq/Ziuev3rdVqceDAAWnftRdgXR8c9+7dC0dHR3To0AFdunSBQqEwqj8tLCzE33//3eT+5OfnIy0tDa+88gpGjRqF7t2715mJo3v37vjrr79QWVkpLasvoNby9PREdna20WswNbdy7969sXDhQuzbtw+hoaHYtGmTyf0plUr06NEDJ06ckJY5OjoiMDAQiYmJjXmpTXLmzBlUVlaid+/ezb5vIro9DL5EZNF69eqFqVOn1pke6oUXXsC+ffswZ84cHDlyBKdPn8Z3331X78Vtd911F7p06YLp06fjr7/+wt69e/HKK68AQJOn0VIoFJg7dy5SUlJw6NAhzJgxAwMGDJBKJ5577jmsX78ea9aswenTp/HOO+/gm2++ue2LuhwcHPDkk0/iueeeQ0JCAk6cOIFZs2ahvLwcM2fOBAA89dRTuHjxIubOnYtTp07hu+++w+LFixEbGwuZTAa1Wo2ZM2fiueeew6+//opjx45hxowZkMma/t+Bq6sr3N3d8fHHHyM9PR2//vorYmNjjdo89NBDEAQBs2bNwokTJ7B9+3asWLGiwf2OGDECubm5eOutt3DmzBl8+OGH2LFjh7T+3LlzWLhwIZKTk3HhwgX8/PPPOH36tPSHhykxMTHYs2eP0bIlS5Zg5cqVWL16NU6fPo3U1FS8//77TX4fbvT777+jc+fOt/3JAxE1PwZfIrJ4r7/+ep2PksPCwrB79278/fffGDp0KHr37o1FixbBz8/P5D7kcjm2bt2K0tJSREZG4l//+pc0q8P1tZ+NYW9vjxdeeAEPPfQQBg8eDLVajS1btkjrJ02ahPfeew8rVqxAz5498dFHH+Gzzz7DiBEjmvbCTVi2bBnuu+8+/N///R/69OmD9PR0/PTTT3B1dQUA+Pv7Y/v27di/fz/Cw8PxxBNPYObMmVLIB4C3334bQ4cOxYQJExAdHY0hQ4agb9++Te6LTCbD5s2bcejQIYSGhuKZZ57B22+/bdRGrVbjhx9+wNGjR9G7d2+8/PLLWL58eYP77d69O/7zn//gww8/RHh4OPbv32/0R4O9vT1OnTqF++67D3fccQcee+wxzJ49G48//ni9+5w5cya2b9+O4uJiadn06dOxatUq/Oc//0HPnj0xfvx4nD59usnvw42++OILzJo167b3Q0TNTxBvLKQiIrISe/fuxZAhQ5Cens7ROStw//33o0+fPiZrxpvL8ePHceedd+Lvv/9u0TvREdGtsTF3B4iIWsu3334LtVqN4OBgpKenY968eRg8eDBDr5V4++238cMPP7ToMbKysrBx40aGXiILxRFfIrIaGzduxNKlS5GRkQEPDw9ER0dj5cqVcHd3N3fXiIioFTD4EhEREZFV4MVtRERERGQVGHyJiIiIyCow+BIRERGRVWDwJSIiIiKrwOBLRERERFaBwZeIiIiIrAKDLxERERFZBQZfIiIiIrIKDL5EREREZBX+P0U98o8iwvjEAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 800x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(8, 5))\n",
    "\n",
    "ax = fig.add_subplot(111)\n",
    "\n",
    "ax.plot(r_ranges[:5], (np.array(target_F) - np.array(clue3D_F))[:5], '--s', label=\"target $\\Delta F_{0.5}$\", c=\"#D55E00\")\n",
    "ax.plot(r_ranges[:5], (np.array(model_reco_F) - np.array(clue3D_F))[:5], '-o', label=\"GNN LP $\\Delta F_{0.5}$\", c=\"#56B4E9\")\n",
    "ax.plot(r_ranges[:5], (np.array(naive_reco_F) - np.array(clue3D_F))[:5], '-v', label=\"naive $\\Delta F_{0.5}$\", c=\"#E69F00\")\n",
    "\n",
    "ax.axhline(max(np.array(naive_reco_F) - np.array(clue3D_F)), label=\"baseline\", c=\"lightgray\", linestyle=\"--\")\n",
    "\n",
    "ax.legend()\n",
    "ax.set_xlabel(\"Neighborhood radius (cm)\")\n",
    "ax.set_ylabel(\"$\\Delta F_{0.5}$\")\n",
    "# ax.set_ylim((-0.1, 0.1))\n",
    "\n",
    "fig.savefig(f'export/{ds_name}-e10-graph-lp-to50-de-radius-plot.png', dpi=300, bbox_inches='tight', transparent=True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "id": "931b57e9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Reco 0.23555861490015229\n",
      "Naive 0.0252482447684228\n",
      "Target 0.24060987025656067\n"
     ]
    }
   ],
   "source": [
    "print(\"Reco\", max(np.array(model_reco_F) - np.array(clue3D_F)))\n",
    "print(\"Naive\", max(np.array(naive_reco_F) - np.array(clue3D_F)))\n",
    "print(\"Target\", max(np.array(target_F) - np.array(clue3D_F)))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "9a309e4f",
   "metadata": {},
   "source": [
    "## Setup comparison:\n",
    "\n",
    "Pairwise:\n",
    "- mean clue3d_to_sim:\tP: 0.994 R: 0.301 F: 0.652\n",
    "- mean target_to_sim:\tP: 0.992 R: 0.689 F: 0.892\n",
    "- mean reco_to_sim:\t    P: 0.659 R: 0.699 F: 0.642\n",
    "\n",
    "Graph (node-pred):\n",
    "- mean clue3d_to_sim:\tP: 0.994 R: 0.301 F: 0.652\n",
    "- mean target_to_sim:\tP: 0.992 R: 0.688 F: 0.892\n",
    "- mean reco_to_sim:\t    P: 0.659 R: 0.699 F: 0.642\n",
    "\n",
    "Graph (link-pred):\n",
    "- mean clue3d_to_sim:\tP: 0.994 R: 0.301 F: 0.652\n",
    "- mean target_to_sim:\tP: 0.992 R: 0.689 F: 0.892\n",
    "- mean reco_to_sim:\tP: 0.915 R: 0.329 F: 0.637\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "328866f1-c862-4dfb-a201-91d2baa53d9f",
   "metadata": {},
   "source": [
    "## Results\n",
    "\n",
    "Using alpha=balance seems to converge faster (100 epochs rather than 200)\n",
    "- maybe just sticking to 0.25? (works just fine)\n",
    "\n",
    "Hyperparams\n",
    "- (64, 128, noskip), alpha=balance, roc_auc = 0.9733\n",
    "- (64, 128, 256fc, noskip, 200e), alpha=1-balance, roc_auc = 0.9771\n",
    "- (64, 128, 256fc, noskip, 200e), alpha=balance, roc_auc = 0.981\n",
    "- (64, 128, 256fc, noskip, 100e), alpha=0.25, roc_auc = 0.9796"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.9.6 ('ve': venv)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.6"
  },
  "vscode": {
   "interpreter": {
    "hash": "37c238d9819f69c2c770157eac01081978c120e64661e10d7fd52c4caf977dc9"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}