metricas copy.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 795,
   "metadata": {},
   "outputs": [],
   "source": [
    "from logging import raiseExceptions\n",
    "from tokenize import Double\n",
    "import networkx as nx\n",
    "import matplotlib.pyplot as plt\n",
    "import torch\n",
    "from torch import nn\n",
    "from torch.utils.data import Dataset, DataLoader\n",
    "import os\n",
    "import numpy as np\n",
    "import torch\n",
    "torch.manual_seed(125)\n",
    "import random\n",
    "random.seed(125)\n",
    "import torch_f as torch_f\n",
    "from modelovae import Node, GRASSEncoder, GRASSDecoder, deserialize\n",
    "import seaborn as sns\n",
    "from matplotlib.ticker import MaxNLocator\n",
    "import statistics\n",
    "from scipy.stats import wasserstein_distance\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 796,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy.stats import norm\n",
    "\n",
    "# Define the x range\n",
    "x = np.linspace(-4, 4, 1000)\n",
    "\n",
    "# Define the two Gaussian distributions\n",
    "mu_s, sigma_s = 1, 0.8    # Approximate posterior: p(z_s(x))\n",
    "mu_prior, sigma_prior = 0, 1  # Prior: p(z)\n",
    "\n",
    "# Calculate the PDFs\n",
    "p_z_s = norm.pdf(x, mu_s, sigma_s)      # p(z_s(x))\n",
    "p_z = norm.pdf(x, mu_prior, sigma_prior)  # p(z)\n",
    "\n",
    "# Custom color palette\n",
    "custom_palette = {\"Real\": \"#9CCFFC\", \"Synthetic\": \"#FF85B1\"}\n",
    "\n",
    "# Plot the distributions\n",
    "plt.figure(figsize=(8, 5))\n",
    "plt.fill_between(x, p_z_s, color=custom_palette[\"Synthetic\"], alpha=0.7, label='$p(z_s(x))$')\n",
    "plt.fill_between(x, p_z, color=custom_palette[\"Real\"], alpha=0.7, label='$p(z)$')\n",
    "\n",
    "# Add KL divergence annotation\n",
    "#plt.text(-3, 0.3, '$L_{KL} = D_{KL}(p(z_s(x)) \\\\, || \\\\, p(z))$', fontsize=12)\n",
    "\n",
    "# Add labels and legend\n",
    "plt.xlabel('$z$')\n",
    "plt.ylabel('Density')\n",
    "plt.title('KL Divergence Between $p(z_s(x))$ and $p(z)$')\n",
    "plt.legend()\n",
    "plt.grid(False)\n",
    "\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 797,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1600x400 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "# Function to plot vectors and annotate angles\n",
    "def plot_cosine_similarity(ax, theta, cosine_value, label):\n",
    "    # Vector A (fixed) and B (rotated)\n",
    "    A = np.array([1, 0])\n",
    "    B = np.array([np.cos(np.radians(theta)), np.sin(np.radians(theta))])\n",
    "\n",
    "    # Plot vectors\n",
    "    ax.quiver(0, 0, *A, angles='xy', scale_units='xy', scale=1, color='blue', label='A')\n",
    "    ax.quiver(0, 0, *B, angles='xy', scale_units='xy', scale=1, color='royalblue', label='B')\n",
    "\n",
    "    # Set axis limits\n",
    "    ax.set_xlim(-1.5, 1.5)\n",
    "    ax.set_ylim(-1.5, 1.5)\n",
    "\n",
    "    # Plot angle and cosine value\n",
    "    ax.text(0.5, -0.2, f\"cos(θ) = {cosine_value}\", fontsize=10)\n",
    "    ax.text(0.7, 0.8, f\"θ = {theta}°\", fontsize=10)\n",
    "\n",
    "    # Add labels\n",
    "    ax.axhline(0, color='black',linewidth=0.5)\n",
    "    ax.axvline(0, color='black',linewidth=0.5)\n",
    "    ax.grid(False)\n",
    "    ax.set_aspect('equal')\n",
    "    ax.set_xticks([])\n",
    "    ax.set_yticks([])\n",
    "\n",
    "# Create subplots\n",
    "fig, axes = plt.subplots(1, 4, figsize=(16, 4))\n",
    "\n",
    "# Plot examples for cos(θ)\n",
    "plot_cosine_similarity(axes[0], 0, 1, \"Most similar vectors\")\n",
    "plot_cosine_similarity(axes[1], 30, 0.86, \"Small angle\")\n",
    "plot_cosine_similarity(axes[2], 90, 0, \"Orthogonal vectors\")\n",
    "plot_cosine_similarity(axes[3], 180, -1, \"Opposite vectors\")\n",
    "\n",
    "# Add titles\n",
    "axes[0].set_title(\"cos(θ) = 1, most similar vectors\")\n",
    "axes[1].set_title(\"cos(θ) ≈ 0.86\")\n",
    "axes[2].set_title(\"cos(θ) = 0\")\n",
    "axes[3].set_title(\"cos(θ) = -1, opposite vectors\")\n",
    "\n",
    "# Adjust layout and show the plot\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 798,
   "metadata": {},
   "outputs": [],
   "source": [
    "use_gpu = True\n",
    "device = torch.device(\"cuda:0\" if use_gpu and torch.cuda.is_available() else \"cpu\")\n",
    "\n",
    "def count_fn(f):\n",
    "    def wrapper(*args, **kwargs):\n",
    "        wrapper.count += 1\n",
    "        return f(*args, **kwargs)\n",
    "    wrapper.count = 0\n",
    "    return wrapper"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 799,
   "metadata": {},
   "outputs": [],
   "source": [
    "def traverseleaves(root, leaves):\n",
    "    if root is not None:\n",
    "        traverseleaves(root.left, leaves)\n",
    "        if root.is_leaf():\n",
    "            leaves.append(root)\n",
    "        traverseleaves(root.right, leaves)\n",
    "        return\n",
    "\n",
    "def traverse(root):\n",
    "        if root is not None:\n",
    "            traverse(root.left)\n",
    "            root.radius=root.radius.reshape(4)\n",
    "            traverse(root.right)\n",
    "            return\n",
    "\n",
    "def traverseradius(root, rlist):\n",
    "    if root is not None:\n",
    "        traverseradius(root.left, rlist)\n",
    "        rlist.append(root.radius[3])\n",
    "        print(\"radius\", root.radius[3])\n",
    "        traverseradius(root.right, rlist)\n",
    "        return \n",
    "\n",
    "def traversefeatures(root, features):\n",
    "    if root is not None:\n",
    "        traversefeatures(root.left, features)\n",
    "        features.append(root.radius[3].cpu().item())\n",
    "        traversefeatures(root.right, features)\n",
    "        return features\n",
    "\n",
    "def traverse_branch_features(root, branch_features, current_branch=None):\n",
    "    if root is None:\n",
    "        return\n",
    "\n",
    "    # Initialize current branch if starting a new one\n",
    "    if current_branch is None:\n",
    "        current_branch = []\n",
    "\n",
    "    # Add the current node's radius to the current branch\n",
    "    current_branch.append(root.radius[3].cpu().item())\n",
    "\n",
    "    # Check if it's a leaf node or bifurcation (has two children)\n",
    "    is_leaf = root.left is None and root.right is None\n",
    "    is_bifurcation = root.left is not None and root.right is not None\n",
    "\n",
    "    if is_leaf or is_bifurcation:\n",
    "        # Add the current branch to the list of branches\n",
    "        branch_features.append(current_branch)\n",
    "        # If it's a bifurcation, continue with new branches for left and right\n",
    "        if is_bifurcation:\n",
    "            traverse_branch_features(root.left, branch_features)\n",
    "            traverse_branch_features(root.right, branch_features)\n",
    "    else:\n",
    "        # Continue down the current branch\n",
    "        if root.left is not None:\n",
    "            traverse_branch_features(root.left, branch_features, current_branch.copy())\n",
    "        if root.right is not None:\n",
    "            traverse_branch_features(root.right, branch_features, current_branch.copy())\n",
    "\n",
    "# Usage example:\n",
    "# branch_features = []\n",
    "# traverse_branch_features(root, branch_features)\n",
    "# print(branch_features)\n",
    "\n",
    "\n",
    "def traversefeatures2(root, features):\n",
    "    if root is not None:\n",
    "        traversefeatures2(root.left, features)\n",
    "        features.append(root.radius[0][3].cpu().item())\n",
    "        traversefeatures2(root.right, features)\n",
    "        return features          \n",
    "\n",
    "def traverse_conexiones(root, tree):\n",
    "        \"\"\"\n",
    "        traverse function will print all the node in the tree.\n",
    "        \"\"\"\n",
    "        if root is not None:\n",
    "            traverse_conexiones(root.left, tree)\n",
    "            if root.right is not None:\n",
    "                tree.append((root.data, root.right.data))\n",
    "            if root.left is not None:\n",
    "                tree.append((root.data, root.left.data))\n",
    "            traverse_conexiones(root.right, tree)\n",
    "            return tree\n",
    "\n",
    "@count_fn\n",
    "def createNode(data, radius,left = None, right = None):\n",
    "        \"\"\"\n",
    "        Utility function to create a node.\n",
    "        \"\"\"\n",
    "        return Node(data, radius, left, right)\n",
    " \n",
    "\n",
    "\n",
    "def read_tree(filename, dir):\n",
    "    with open('./' +dir +'/' +filename, \"r\") as f:\n",
    "        byte = f.read() \n",
    "        return byte\n",
    "\n",
    "def numerar_nodos(root, count):\n",
    "    if root is not None:\n",
    "        numerar_nodos(root.left, count)\n",
    "        root.data = len(count)\n",
    "        count.append(1)\n",
    "        numerar_nodos(root.right, count)\n",
    "        return \n",
    "\n",
    "def number_nodes(data_loader, batch_size):\n",
    "    n_no = []\n",
    "    qzero = 0\n",
    "    qOne = 0\n",
    "    qtwo = 0\n",
    "    for batch in data_loader:\n",
    "        for tree in batch:\n",
    "            tree = tree[0]\n",
    "            count = []\n",
    "            n = tree.count_nodes(tree, count)\n",
    "            n_no.append(len(n))\n",
    "            li = []\n",
    "            tree.traverseInorderChilds(tree, li)\n",
    "            zero = [a for a in li if a == 0]\n",
    "            one = [a for a in li if a == 1]\n",
    "            two = [a for a in li if a == 2]\n",
    "            qzero += len(zero)\n",
    "            qOne += len(one)\n",
    "            qtwo += len(two)\n",
    "\n",
    "    nprom = np.mean(n_no)\n",
    "    qzero /= len(data_loader)*batch_size\n",
    "    qOne /= len(data_loader)*batch_size\n",
    "    qtwo /= len(data_loader)*batch_size\n",
    "    if round(qzero) == 0:\n",
    "        qzero = 1\n",
    "    if round(qOne) == 0:\n",
    "        qOne = 1\n",
    "    if round(qtwo) == 0:\n",
    "        qtwo = 1\n",
    "    mult = torch.tensor([1/round(qzero),1/round(qOne),1/round(qtwo)], device = device)\n",
    "    return mult\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 800,
   "metadata": {},
   "outputs": [],
   "source": [
    "def decode_testing(v, max, decoder, mult, min):\n",
    "    def decode_node(v, max, decoder, mult, min):\n",
    "\n",
    "        cl = decoder.nodeClassifier(v)\n",
    "        _, label = torch.max(cl, 1)\n",
    "        label = label.data\n",
    "        \n",
    "        \n",
    "        if label == 1 and createNode.count <= max:\n",
    "            right, radius = decoder.internalDecoder(v)\n",
    "                \n",
    "            d = createNode(1, radius) \n",
    "            \n",
    "            d.right = decode_node(right, max, decoder, mult, min)\n",
    "            return d\n",
    "\n",
    "        elif label == 2 and createNode.count <= max:\n",
    "            left, right, radius = decoder.bifurcationDecoder(v)\n",
    "                \n",
    "            d = createNode(1, radius)\n",
    "            \n",
    "            d.right = decode_node(right, max, decoder, mult, min)\n",
    "            d.left = decode_node(left, max, decoder, mult, min)\n",
    "        \n",
    "            return d\n",
    "\n",
    "       \n",
    "        elif label == 0 : ##output del classifier\n",
    "            radio = decoder.featureDecoder(v)\n",
    "            return createNode(1,radio)  \n",
    "        \n",
    "\n",
    "    createNode.count = 0\n",
    "    v = decoder.sample_decoder(v)\n",
    "    dec = decode_node (v, max, decoder, mult, min)\n",
    "\n",
    "    return dec"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 801,
   "metadata": {},
   "outputs": [],
   "source": [
    "def my_collate(batch):\n",
    "    return batch\n",
    "\n",
    "\n",
    "class tDataset(Dataset):\n",
    "    def __init__(self, l, dir, transform=None):\n",
    "        self.names = l\n",
    "        self.transform = transform\n",
    "        self.data = [] #lista con las strings de todos los arboles\n",
    "        for file in self.names:\n",
    "            self.data.append(read_tree(file, dir))\n",
    "        self.trees = []\n",
    "\n",
    "        for tree in self.data:\n",
    "            deserial = deserialize(tree)\n",
    "            self.trees.append(deserial)\n",
    "            \n",
    "\n",
    "    def __len__(self):\n",
    "        return len(self.names)\n",
    "\n",
    "    def __getitem__(self, idx):\n",
    "        #file = self.names[idx]\n",
    "        #string = read_tree(file)\n",
    "        tree = self.trees[idx]\n",
    "        name = self.names[idx]\n",
    "        return tree\n",
    "\n",
    "batch_size = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 802,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "d data/paper/AneuriskP15eps01\n",
      "100\n"
     ]
    }
   ],
   "source": [
    "dataset_name = \"Aneurisk\"\n",
    "p = str(15)\n",
    "eps = str(0)+str(1)\n",
    "d = \"data/paper/\" + dataset_name + \"P\" +p + \"eps\" + eps\n",
    "print(\"d\", d)\n",
    "t_list = os.listdir(d)[:100]\n",
    "print(len(t_list))\n",
    "dataset = tDataset(t_list, d )\n",
    "data_loader = DataLoader(dataset, batch_size = batch_size, shuffle=True, collate_fn=my_collate)\n",
    "    \n",
    "i = 0\n",
    "d = None\n",
    "d_nodos = []\n",
    "n_bifo = []\n",
    "for batch in data_loader:\n",
    "    for input in batch:\n",
    "        count = []\n",
    "        numerar_nodos(input, count)\n",
    "        d_nodos.append(len(count))\n",
    "        li = []\n",
    "        input.traverseInorderChilds(input, li)\n",
    "        two = [a for a in li if a == 2]\n",
    "        n_bifo.append(len(two))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 803,
   "metadata": {},
   "outputs": [],
   "source": [
    "def distance(node1, node2):\n",
    "    \"\"\"\n",
    "    Calculates the Euclidean distance between two nodes\n",
    "    \"\"\"\n",
    "    if node1 is not None and node2 is not None:\n",
    "        dist = ((node1.radius[0]-node2.radius[0])**2 + \n",
    "                (node1.radius[1]-node2.radius[1])**2 + \n",
    "                (node1.radius[2]-node2.radius[2])**2)**0.5\n",
    "    else:\n",
    "        dist = 0\n",
    "    return dist"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 804,
   "metadata": {},
   "outputs": [],
   "source": [
    "def tree_length(node):\n",
    "    '''calcula el largo total del arbol'''\n",
    "    if node is None:\n",
    "        return 0\n",
    "\n",
    "    stack = [(node, None)]\n",
    "    length = 0\n",
    "\n",
    "    while stack:\n",
    "        curr, prev = stack.pop()\n",
    "\n",
    "        if curr.left:\n",
    "            stack.append((curr.left, curr))\n",
    "\n",
    "        if curr.right:\n",
    "            stack.append((curr.right, curr))\n",
    "\n",
    "        if prev:\n",
    "            length += distance(prev, curr)\n",
    "\n",
    "    return length"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 805,
   "metadata": {},
   "outputs": [],
   "source": [
    "def total_length(root, node, d):\n",
    "    '''\n",
    "    calcula la distancia del primer a ultimo nodo, arbol sin bifurcaciones\n",
    "    '''\n",
    "    if node.left:\n",
    "        total_length(root, node.left, d)\n",
    "    if node.isLeaf():\n",
    "        #d = distance(root, node)\n",
    "        d.append(distance(root, node))\n",
    "        return \n",
    "    if node.right:\n",
    "        total_length(root, node.right, d)\n",
    "\n",
    "def total_path_length(node):\n",
    "    \"\"\"\n",
    "    Calculates the total path length of the tree\n",
    "    \"\"\"\n",
    "    if node is None:\n",
    "        return 0\n",
    "    else:\n",
    "        return distance(node, node.left) + total_path_length(node.left) + distance(node, node.right) + total_path_length(node.right)\n",
    "\n",
    "def tortuosity(root):\n",
    "    \"\"\"\n",
    "    Calculates the tortuosity of the tree\n",
    "    \"\"\"\n",
    "    total_path_length_value = total_path_length(root)\n",
    "    d = []\n",
    "    max_path_length_value = total_length(root,root, d)\n",
    "    return total_path_length_value / d[0]"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "calculo tortuosidad, numero de bifurcaciones, largo total para el dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 806,
   "metadata": {},
   "outputs": [],
   "source": [
    "def collect_radius(node, radius_list):\n",
    "    \"\"\"\n",
    "    Collects the radius of each node in the branch.\n",
    "    \"\"\"\n",
    "    if node is not None:\n",
    "        radius_list.append(node.radius[3])\n",
    "        for child in node.children:\n",
    "            print(\"////\", node.children)\n",
    "            collect_radius(child, radius_list)\n",
    "\n",
    "import math\n",
    "\n",
    "def collect_radius(node, radius_list):\n",
    "    \"\"\"\n",
    "    Collects the radius of each node in the branch.\n",
    "    \"\"\"\n",
    "    if node is not None:\n",
    "        radius_list.append(node.radius[3])  # Append the current node's radius\n",
    "        if node.left is not None:\n",
    "            collect_radius(node.left, radius_list)\n",
    "        if node.right is not None:\n",
    "            collect_radius(node.right, radius_list)\n",
    "\n",
    "# Usage example:\n",
    "# branch_radii = []\n",
    "# for each subtree (branch):\n",
    "#     radius_list = []\n",
    "#     collect_radius(nodoRaiz, radius_list)\n",
    "#     branch_radii.append(radius_list)\n",
    "\n",
    "\n",
    "def collect_length(node, length_list):\n",
    "    \"\"\"\n",
    "    Collects the length of each segment in the branch.\n",
    "    \"\"\"\n",
    "    if node is not None:\n",
    "        if node.left is not None:\n",
    "            # Calculate the Euclidean distance between the node and its left child\n",
    "            length = euclidean_distance(node.radius[:3], node.left.radius[:3])\n",
    "            length_list.append(length)\n",
    "            collect_length(node.left, length_list)\n",
    "        \n",
    "        if node.right is not None:\n",
    "            # Calculate the Euclidean distance between the node and its right child\n",
    "            length = euclidean_distance(node.radius[:3], node.right.radius[:3])\n",
    "            length_list.append(length)\n",
    "            collect_length(node.right, length_list)\n",
    "\n",
    "def euclidean_distance(coord1, coord2):\n",
    "    \"\"\"\n",
    "    Calculates the Euclidean distance between two 3D points.\n",
    "    \"\"\"\n",
    "    return math.sqrt(sum((c1 - c2) ** 2 for c1, c2 in zip(coord1, coord2)))\n",
    "\n",
    "# Usage example:\n",
    "# branch_lengths = []\n",
    "# for each subtree (branch):\n",
    "#     length_list = []\n",
    "#     collect_length(nodoRaiz, length_list)\n",
    "#     total_branch_length = sum(length_list)\n",
    "#     branch_lengths.append(total_branch_length)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 807,
   "metadata": {},
   "outputs": [],
   "source": [
    "tortuosidad = []\n",
    "n_bif = []\n",
    "largo = []\n",
    "n_nodes = []\n",
    "ratio_radios = []\n",
    "ratio_radios_por_rama = []\n",
    "ratio_largos_por_rama = []\n",
    "for batch in data_loader:\n",
    "    for input in batch:\n",
    "\n",
    "        graph = nx.Graph()\n",
    "        input.toGraph(graph, 0 , False, flag = 0)\n",
    "        degree1 = []\n",
    "\n",
    "        for node in graph.nodes():\n",
    "            if graph.degree(node) == 1:\n",
    "                degree1.append(node)\n",
    "       \n",
    "        caminos = [ nx.shortest_path( graph, degree1[0], t ) for t in degree1[1:] ]\n",
    "    \n",
    "        subarboles = []\n",
    "        for g in caminos:\n",
    "            grafo = nx.Graph()\n",
    "            grafo.add_nodes_from(g)\n",
    "            et = []\n",
    "            for node in grafo.nodes:\n",
    "                edge = graph[node]\n",
    "                edges = [a for a in list(edge.keys()) if a in g]\n",
    "                edge_tuple = [(a,b) for a,b in zip([node]*len(edges), edges) ]\n",
    "                for tuple in edge_tuple:\n",
    "                    et.append(tuple)\n",
    "            grafo.add_edges_from(et)\n",
    "    \n",
    "            aRecorrer = []\n",
    "            numeroNodoInicial = degree1[0]\n",
    "            r = graph.nodes[numeroNodoInicial]['radio']\n",
    "            rad = list(graph.nodes[numeroNodoInicial]['posicion'])\n",
    "            rad.append(r)\n",
    "\n",
    "            nodoRaiz = Node( numeroNodoInicial, radius =  rad )\n",
    "\n",
    "            for vecino in grafo.neighbors( numeroNodoInicial ):\n",
    "                if vecino != numeroNodoInicial:\n",
    "                    aRecorrer.append( (vecino, numeroNodoInicial,nodoRaiz ) )\n",
    "            \n",
    "            while len(aRecorrer) != 0:\n",
    "                nodoAAgregar, numeroNodoPadre,nodoPadre = aRecorrer.pop(0)\n",
    "                r = graph.nodes[nodoAAgregar]['radio']\n",
    "                rad = list(graph.nodes[nodoAAgregar]['posicion'])\n",
    "                rad.append(r)\n",
    "                nodoActual = Node( nodoAAgregar, radius =  rad)\n",
    "                nodoPadre.agregarHijo( nodoActual )\n",
    "                for vecino in grafo.neighbors( nodoAAgregar ):\n",
    "                    if vecino != numeroNodoPadre:\n",
    "                        aRecorrer.append( (vecino, nodoAAgregar,nodoActual) )\n",
    "            subarboles.append(nodoRaiz)\n",
    "            tortuosidad.append(tortuosity(nodoRaiz))\n",
    "            radius_list = []\n",
    "            collect_radius(nodoRaiz, radius_list)\n",
    "            average_radius = sum(radius_list) / len(radius_list)\n",
    "            ratio_radios_por_rama.append(average_radius)\n",
    "            length_list = []\n",
    "            collect_length(nodoRaiz, length_list)\n",
    "            average_length = sum(length_list) / len(length_list)\n",
    "            ratio_largos_por_rama.append(average_length)\n",
    "        li = []\n",
    "        input.traverseInorderChilds(input, li)\n",
    "        two = [a for a in li if a == 2]\n",
    "        n_bif.append(len(two))\n",
    "        total_path_length_value = tree_length(input)\n",
    "        largo.append(total_path_length_value.cpu().item())\n",
    "        count = []\n",
    "        n = input.countNodes(input, count)\n",
    "        n_nodes.append(len(n))\n",
    "        features = []\n",
    "        traversefeatures(input, features)\n",
    "       \n",
    "        #ratio_radios_por_rama.append([sum(lst) / len(lst) for lst in branch_features])\n",
    "    \n",
    "        ratio_radios.append(sum(features)/len(features))\n",
    "        #print(\"ratio radios\", ratio_radios)\n",
    "        #print(\"ratio radios por ramas\", ratio_radios_por_rama)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 808,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "381"
      ]
     },
     "execution_count": 808,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(ratio_radios_por_rama)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 809,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "381"
      ]
     },
     "execution_count": 809,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(ratio_largos_por_rama)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 810,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.033153825802924135,\n",
       " 0.02714068285297668,\n",
       " 0.032120310799880245,\n",
       " 0.031034931084150324,\n",
       " 0.029027056904002792,\n",
       " 0.027628920810015754,\n",
       " 0.02814963044432826,\n",
       " 0.03556139500286595,\n",
       " 0.03821482429261847,\n",
       " 0.04576364610776237,\n",
       " 0.047694635808665724,\n",
       " 0.046544713561189666,\n",
       " 0.049446592090873924,\n",
       " 0.04716104959864289,\n",
       " 0.046216822617029946,\n",
       " 0.047419080889343365,\n",
       " 0.04727922453764609,\n",
       " 0.036003636977061396,\n",
       " 0.03522225201907866,\n",
       " 0.04819458414419353,\n",
       " 0.042193213004295654,\n",
       " 0.042181016390784434,\n",
       " 0.04477326150151795,\n",
       " 0.02384047678195542,\n",
       " 0.024236179718119564,\n",
       " 0.036936474086838054,\n",
       " 0.0364962282540967,\n",
       " 0.03030259439887593,\n",
       " 0.03392175242480975,\n",
       " 0.039833501489635825,\n",
       " 0.03694592973323344,\n",
       " 0.04964925425889261,\n",
       " 0.05198634481424171,\n",
       " 0.05739125750292184,\n",
       " 0.05293109965186083,\n",
       " 0.044079173157142555,\n",
       " 0.05094545200438931,\n",
       " 0.039536077718460974,\n",
       " 0.04817083363960487,\n",
       " 0.06075875102443075,\n",
       " 0.029404779139634005,\n",
       " 0.03139758906953522,\n",
       " 0.045583983930645715,\n",
       " 0.04470177511061821,\n",
       " 0.039779253362942114,\n",
       " 0.03953214442307487,\n",
       " 0.05038887793924755,\n",
       " 0.05310823574208443,\n",
       " 0.047082560056813404,\n",
       " 0.043339374715818293,\n",
       " 0.04267823426597918,\n",
       " 0.04935275754599655,\n",
       " 0.05179261077253942,\n",
       " 0.05245536147647611,\n",
       " 0.04894837943052516,\n",
       " 0.04953140919679359,\n",
       " 0.04949562895466163,\n",
       " 0.046911220827938645,\n",
       " 0.04953983244266745,\n",
       " 0.049407332015040537,\n",
       " 0.03671926291421815,\n",
       " 0.03825528326548663,\n",
       " 0.03656054007452519,\n",
       " 0.03909583358432999,\n",
       " 0.03705518577229433,\n",
       " 0.03876265360702051,\n",
       " 0.042865511099224166,\n",
       " 0.035050897889662924,\n",
       " 0.04209540869597072,\n",
       " 0.04023418652476898,\n",
       " 0.040789223796630275,\n",
       " 0.05561938479391235,\n",
       " 0.06294373486589405,\n",
       " 0.05225820350846488,\n",
       " 0.061170656596626814,\n",
       " 0.0610608177197564,\n",
       " 0.028459366022422893,\n",
       " 0.03531721669619387,\n",
       " 0.04019523020332876,\n",
       " 0.04272251611780121,\n",
       " 0.03581811236647164,\n",
       " 0.027681811782700076,\n",
       " 0.027559200240044386,\n",
       " 0.032369173283619705,\n",
       " 0.0312920627472786,\n",
       " 0.03136072501075411,\n",
       " 0.04625960108059592,\n",
       " 0.05485903939171931,\n",
       " 0.051901870906218145,\n",
       " 0.07237371067260108,\n",
       " 0.06131473601800478,\n",
       " 0.060288554659260786,\n",
       " 0.06426786243811762,\n",
       " 0.060574099043682085,\n",
       " 0.07002437180633182,\n",
       " 0.06357597706831744,\n",
       " 0.04917423957948762,\n",
       " 0.04886731071648743,\n",
       " 0.04602065100263807,\n",
       " 0.04576609735567259,\n",
       " 0.030724692460259835,\n",
       " 0.030307355766868188,\n",
       " 0.03109240249361307,\n",
       " 0.02583013551807954,\n",
       " 0.027731344636223988,\n",
       " 0.024816858738898925,\n",
       " 0.023687373801298997,\n",
       " 0.031595721158066484,\n",
       " 0.031110213970160845,\n",
       " 0.045583983930645715,\n",
       " 0.04470177511061821,\n",
       " 0.031055293562508392,\n",
       " 0.039671212693218,\n",
       " 0.0398467486795106,\n",
       " 0.03913392858057298,\n",
       " 0.03790653481904943,\n",
       " 0.05030627207115324,\n",
       " 0.04324444169171112,\n",
       " 0.05882131767826159,\n",
       " 0.06756768484138741,\n",
       " 0.04695096438816493,\n",
       " 0.044358321450706624,\n",
       " 0.044768746323999666,\n",
       " 0.04941874695886202,\n",
       " 0.049451397506784955,\n",
       " 0.03415573893135252,\n",
       " 0.038755859096970546,\n",
       " 0.022298533676296754,\n",
       " 0.021961522161522775,\n",
       " 0.02174160790993042,\n",
       " 0.03787036803649791,\n",
       " 0.043502879781449094,\n",
       " 0.04080526857701491,\n",
       " 0.04102291343172,\n",
       " 0.039107322694196645,\n",
       " 0.04106946516195625,\n",
       " 0.03657923298024248,\n",
       " 0.029482976063265505,\n",
       " 0.04643240645945771,\n",
       " 0.03366844504413658,\n",
       " 0.03804919359331753,\n",
       " 0.0444799207078694,\n",
       " 0.03880161961150165,\n",
       " 0.047238053553107444,\n",
       " 0.042619145830616705,\n",
       " 0.031982327733175345,\n",
       " 0.03543293664064822,\n",
       " 0.04103941068548646,\n",
       " 0.03441397581957603,\n",
       " 0.04303134527158895,\n",
       " 0.040549020654048765,\n",
       " 0.04305853279006823,\n",
       " 0.04197271751454058,\n",
       " 0.052334774621430984,\n",
       " 0.045833720965709486,\n",
       " 0.05040045458610269,\n",
       " 0.050521229719336015,\n",
       " 0.043862180262198444,\n",
       " 0.04354524946571642,\n",
       " 0.036470294515196726,\n",
       " 0.036682710965753616,\n",
       " 0.05119552427244528,\n",
       " 0.05100061152162558,\n",
       " 0.04904236098849296,\n",
       " 0.04809559594355671,\n",
       " 0.05045823584127052,\n",
       " 0.028015382652863564,\n",
       " 0.02946307095873546,\n",
       " 0.03319367473755779,\n",
       " 0.03749150152671198,\n",
       " 0.040326900553268985,\n",
       " 0.0365683217388823,\n",
       " 0.056203215164935205,\n",
       " 0.06961413940676169,\n",
       " 0.0647936972321732,\n",
       " 0.07231031236452383,\n",
       " 0.059466041877919075,\n",
       " 0.07432470199313494,\n",
       " 0.08519067705391635,\n",
       " 0.027374246939641796,\n",
       " 0.035451139344722715,\n",
       " 0.0366698437866198,\n",
       " 0.03579013530042444,\n",
       " 0.03880631261428236,\n",
       " 0.04016388253592631,\n",
       " 0.03667773735492771,\n",
       " 0.04648668559563522,\n",
       " 0.05482902711527415,\n",
       " 0.05939741748292075,\n",
       " 0.03961149565166346,\n",
       " 0.039503440652391274,\n",
       " 0.025636657145658118,\n",
       " 0.026355681070681976,\n",
       " 0.025372449919879782,\n",
       " 0.0301974539980789,\n",
       " 0.028160376616747494,\n",
       " 0.04346098408148561,\n",
       " 0.04138922302285054,\n",
       " 0.04209936451894255,\n",
       " 0.03245544954886094,\n",
       " 0.035742738265249804,\n",
       " 0.034799137149961426,\n",
       " 0.035789533285070674,\n",
       " 0.03864813328940666,\n",
       " 0.0369307966426716,\n",
       " 0.036951731663281195,\n",
       " 0.04460308096550436,\n",
       " 0.050027962477617795,\n",
       " 0.04813629264440894,\n",
       " 0.03393823701489668,\n",
       " 0.04211083574417049,\n",
       " 0.0384028551938497,\n",
       " 0.03894729741403096,\n",
       " 0.04180949150070212,\n",
       " 0.03955300149060764,\n",
       " 0.04066466433746077,\n",
       " 0.041999254670492994,\n",
       " 0.046122773414545834,\n",
       " 0.043905926319680676,\n",
       " 0.039076818126997634,\n",
       " 0.045982732457101584,\n",
       " 0.04556504756872837,\n",
       " 0.044500197463158905,\n",
       " 0.04708302592598061,\n",
       " 0.05033849219260153,\n",
       " 0.05015312949556436,\n",
       " 0.04919879542272887,\n",
       " 0.024845252484749197,\n",
       " 0.025033181772610238,\n",
       " 0.02199463860505756,\n",
       " 0.02450636067590512,\n",
       " 0.031965486093971186,\n",
       " 0.030838922804828334,\n",
       " 0.033392928323331184,\n",
       " 0.030257131307307214,\n",
       " 0.029044381253059897,\n",
       " 0.05015850050773679,\n",
       " 0.04925472388447732,\n",
       " 0.028035183228690093,\n",
       " 0.030966828933307974,\n",
       " 0.03126976045173943,\n",
       " 0.03751659843689499,\n",
       " 0.033962502321616024,\n",
       " 0.03690394463819406,\n",
       " 0.032036894801961285,\n",
       " 0.03053326534673261,\n",
       " 0.03475114962877877,\n",
       " 0.030309903287992348,\n",
       " 0.03120651154614969,\n",
       " 0.03343351501241362,\n",
       " 0.03162946829743964,\n",
       " 0.05169074802845271,\n",
       " 0.05347388783989561,\n",
       " 0.047739412986552196,\n",
       " 0.04919955940429827,\n",
       " 0.049722267222726546,\n",
       " 0.04117011532701464,\n",
       " 0.04798008185995048,\n",
       " 0.05287334363582188,\n",
       " 0.04808408146277155,\n",
       " 0.026917728672108465,\n",
       " 0.02990146922156393,\n",
       " 0.03170885532289215,\n",
       " 0.032117031923825624,\n",
       " 0.03435087140717362,\n",
       " 0.030884950897718033,\n",
       " 0.03209083384546892,\n",
       " 0.0310352211506498,\n",
       " 0.03032516865321012,\n",
       " 0.03310141574879119,\n",
       " 0.021981724982218447,\n",
       " 0.023394043054678834,\n",
       " 0.026878403898175227,\n",
       " 0.027834578690030816,\n",
       " 0.02293873836549804,\n",
       " 0.03796279347177216,\n",
       " 0.031076215439620504,\n",
       " 0.04492317824952988,\n",
       " 0.044216351230388945,\n",
       " 0.044971920891560144,\n",
       " 0.04032474281675661,\n",
       " 0.03994960438001733,\n",
       " 0.041017882264216224,\n",
       " 0.04379807671634136,\n",
       " 0.04535894218913808,\n",
       " 0.04403593062171226,\n",
       " 0.04676899125772029,\n",
       " 0.04186800353087243,\n",
       " 0.04469985137110734,\n",
       " 0.03678231024476842,\n",
       " 0.04218508651030367,\n",
       " 0.03680935154898119,\n",
       " 0.03902920904312473,\n",
       " 0.03994934967846885,\n",
       " 0.03726703470224631,\n",
       " 0.02746659511877386,\n",
       " 0.02999549918956052,\n",
       " 0.03006917014990554,\n",
       " 0.03041920515515592,\n",
       " 0.02344408311469281,\n",
       " 0.021530375475998474,\n",
       " 0.02082872734265908,\n",
       " 0.03618762704827856,\n",
       " 0.040472252825197945,\n",
       " 0.039141538044270084,\n",
       " 0.038802814407448726,\n",
       " 0.04379807671634136,\n",
       " 0.04535894218913808,\n",
       " 0.04403593062171226,\n",
       " 0.04676899125772029,\n",
       " 0.04186800353087243,\n",
       " 0.04469985137110734,\n",
       " 0.04568881084709651,\n",
       " 0.04576337408059668,\n",
       " 0.05331723857628618,\n",
       " 0.0495554626139895,\n",
       " 0.05093524003162608,\n",
       " 0.058725524590538626,\n",
       " 0.05777449719273791,\n",
       " 0.060173193741700416,\n",
       " 0.07869586615500207,\n",
       " 0.06035315844336591,\n",
       " 0.05700730925240338,\n",
       " 0.03654902686036044,\n",
       " 0.03989911003867632,\n",
       " 0.033127805515679946,\n",
       " 0.035958108673024604,\n",
       " 0.08892624726375843,\n",
       " 0.10424523993284211,\n",
       " 0.08597706527223817,\n",
       " 0.1286839637034813,\n",
       " 0.16004817998446907,\n",
       " 0.026751704379580087,\n",
       " 0.024191787936394628,\n",
       " 0.03058133341959188,\n",
       " 0.03955300149060764,\n",
       " 0.04066466433746077,\n",
       " 0.041999254670492994,\n",
       " 0.046122773414545834,\n",
       " 0.043905926319680676,\n",
       " 0.059712996816591396,\n",
       " 0.04989032386415945,\n",
       " 0.05862085065109081,\n",
       " 0.059877091701695515,\n",
       " 0.06185350019060945,\n",
       " 0.04199125100620871,\n",
       " 0.04946904241670649,\n",
       " 0.027896903217195103,\n",
       " 0.029478466721093426,\n",
       " 0.02853208747074124,\n",
       " 0.027596546295815955,\n",
       " 0.029256654909526027,\n",
       " 0.029478466721093426,\n",
       " 0.02853208747074124,\n",
       " 0.027596546295815955,\n",
       " 0.029256654909526027,\n",
       " 0.018566232750817074,\n",
       " 0.02595681589231536,\n",
       " 0.025820326890032514,\n",
       " 0.027366932626521522,\n",
       " 0.02922601232073779,\n",
       " 0.030327188257600467,\n",
       " 0.04341612960519535,\n",
       " 0.04134049490662611,\n",
       " 0.057896558996954374,\n",
       " 0.061072148635966626,\n",
       " 0.06409009478674711,\n",
       " 0.037970987758535614,\n",
       " 0.037239993312515624,\n",
       " 0.0373142417284272,\n",
       " 0.04299549991392099,\n",
       " 0.053730049275766086,\n",
       " 0.03960903636215043,\n",
       " 0.028767780556142635,\n",
       " 0.028235754286381465,\n",
       " 0.030938762422507572,\n",
       " 0.04696499964969346,\n",
       " 0.04716796800050841,\n",
       " 0.049943597240574886,\n",
       " 0.04689048265974417,\n",
       " 0.045298513015764055]"
      ]
     },
     "execution_count": 810,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ratio_largos_por_rama"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 811,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.021653898060321808,\n",
       " 0.036887064847784735,\n",
       " 0.04330315151862029,\n",
       " 0.04304551859867984,\n",
       " 0.02973809253809781,\n",
       " 0.036315720592593324,\n",
       " 0.0363799223740553,\n",
       " 0.029986338689923286,\n",
       " 0.034376556585942,\n",
       " 0.04221309348940849,\n",
       " 0.029124620258808135,\n",
       " 0.03148658946156502,\n",
       " 0.0314789379388094,\n",
       " 0.03747069071336039,\n",
       " 0.039735320334633194,\n",
       " 0.034793133564692526,\n",
       " 0.0447144164735901,\n",
       " 0.021237906933362995,\n",
       " 0.02181175204792193,\n",
       " 0.022226870913679402,\n",
       " 0.021956746830888416,\n",
       " 0.031658416225885354,\n",
       " 0.028450043639168143,\n",
       " 0.029308349986014694,\n",
       " 0.027882604827654773,\n",
       " 0.04213260271554363,\n",
       " 0.04239278228889252,\n",
       " 0.04243478988265169,\n",
       " 0.0401199726620689,\n",
       " 0.036234031676461824,\n",
       " 0.027785900588098326,\n",
       " 0.0578367734560743,\n",
       " 0.06919632049213195,\n",
       " 0.07078991175211709,\n",
       " 0.07355792190019901,\n",
       " 0.02649685957779487,\n",
       " 0.04828955171008905,\n",
       " 0.04958104491233826,\n",
       " 0.039583728593938494,\n",
       " 0.030414149397984147,\n",
       " 0.01626379363005981,\n",
       " 0.018148201459553093,\n",
       " 0.03980308927987751,\n",
       " 0.032927999763112316,\n",
       " 0.021657812675195082,\n",
       " 0.021732468911934467,\n",
       " 0.025093558776591505,\n",
       " 0.025011884270324594,\n",
       " 0.0454355925321579,\n",
       " 0.04585774133688417,\n",
       " 0.04400093185490575,\n",
       " 0.03453350174976023,\n",
       " 0.036055027654296475,\n",
       " 0.04273574112448841,\n",
       " 0.02666760940104723,\n",
       " 0.035678110117542336,\n",
       " 0.029780595434893822,\n",
       " 0.024273126450335157,\n",
       " 0.027876481060580962,\n",
       " 0.02832459741882209,\n",
       " 0.03385265024068455,\n",
       " 0.03437574611355861,\n",
       " 0.036125302159537874,\n",
       " 0.02763496587673823,\n",
       " 0.02776805233831207,\n",
       " 0.028195423462117713,\n",
       " 0.030704499369389134,\n",
       " 0.01892822289741353,\n",
       " 0.027799493094023905,\n",
       " 0.02909436935765876,\n",
       " 0.02446342916472962,\n",
       " 0.05839169102118296,\n",
       " 0.053797402791678904,\n",
       " 0.0364749853539726,\n",
       " 0.0396007996221835,\n",
       " 0.0378809523485277,\n",
       " 0.031751593382194125,\n",
       " 0.03352472290862352,\n",
       " 0.03235757794102718,\n",
       " 0.0338365925026351,\n",
       " 0.015048590333511433,\n",
       " 0.013260016543790698,\n",
       " 0.012345860627564516,\n",
       " 0.017463241492144087,\n",
       " 0.018959128954024476,\n",
       " 0.020212587307799946,\n",
       " 0.025737358113894097,\n",
       " 0.027009856576720875,\n",
       " 0.02962566612081395,\n",
       " 0.036379084564172305,\n",
       " 0.05677226267755032,\n",
       " 0.05065405979752541,\n",
       " 0.04472320802509785,\n",
       " 0.043319945633411405,\n",
       " 0.035927986204624174,\n",
       " 0.0334112835675478,\n",
       " 0.02381505196293195,\n",
       " 0.024968007020652293,\n",
       " 0.02617432816574971,\n",
       " 0.026577984370912116,\n",
       " 0.016843839016343867,\n",
       " 0.017292690906851066,\n",
       " 0.01655028179465305,\n",
       " 0.014607664837967604,\n",
       " 0.02287273180177983,\n",
       " 0.016756960871414497,\n",
       " 0.019074580698673214,\n",
       " 0.020714034592466696,\n",
       " 0.0203446549407783,\n",
       " 0.03980308927987751,\n",
       " 0.032927999763112316,\n",
       " 0.0459845424469175,\n",
       " 0.03475531054815898,\n",
       " 0.024889061134308575,\n",
       " 0.033238282155555986,\n",
       " 0.036609583728325866,\n",
       " 0.03063057054980443,\n",
       " 0.03797932962576548,\n",
       " 0.035004936655362445,\n",
       " 0.0460728890572985,\n",
       " 0.02754710813363393,\n",
       " 0.02971161642160855,\n",
       " 0.03533392952864661,\n",
       " 0.027898078539261694,\n",
       " 0.02885450633536828,\n",
       " 0.044555838444623456,\n",
       " 0.04495928121794914,\n",
       " 0.021151145920157433,\n",
       " 0.021290768323273493,\n",
       " 0.02186876834769334,\n",
       " 0.02089176799303719,\n",
       " 0.021532513234498247,\n",
       " 0.026338375666562247,\n",
       " 0.03049036016797318,\n",
       " 0.02684169780049059,\n",
       " 0.028006728117664654,\n",
       " 0.018395524006336927,\n",
       " 0.01946538547053933,\n",
       " 0.020212848903611302,\n",
       " 0.024701490779133403,\n",
       " 0.026320989627171967,\n",
       " 0.02492403465190104,\n",
       " 0.04756802587459485,\n",
       " 0.02848232458703793,\n",
       " 0.03130921591073275,\n",
       " 0.02091915812343359,\n",
       " 0.023627405799925328,\n",
       " 0.029736432805657388,\n",
       " 0.050869309392414595,\n",
       " 0.02934404177700772,\n",
       " 0.03605787397215241,\n",
       " 0.0511440201417396,\n",
       " 0.050768013259297924,\n",
       " 0.04533109323758828,\n",
       " 0.030400745570659637,\n",
       " 0.05307646789070633,\n",
       " 0.04531450205970378,\n",
       " 0.02797149813600949,\n",
       " 0.02813919136921565,\n",
       " 0.0154509828115503,\n",
       " 0.02127457046034661,\n",
       " 0.023295109601397263,\n",
       " 0.024348086285355845,\n",
       " 0.038242916772632224,\n",
       " 0.04000080175893871,\n",
       " 0.03696832526475191,\n",
       " 0.027545810118317603,\n",
       " 0.024741552118211985,\n",
       " 0.02836984691434893,\n",
       " 0.03423153962297686,\n",
       " 0.03790917594371171,\n",
       " 0.04064432126951629,\n",
       " 0.06554054003208876,\n",
       " 0.08036496629938483,\n",
       " 0.0578804203708257,\n",
       " 0.059856134025674114,\n",
       " 0.054884893989021126,\n",
       " 0.03894327317054073,\n",
       " 0.04983405058737844,\n",
       " 0.03699461903808446,\n",
       " 0.04861785298972218,\n",
       " 0.04779680381561148,\n",
       " 0.04615669536950259,\n",
       " 0.04179787462384536,\n",
       " 0.032409091126816024,\n",
       " 0.03426864788193127,\n",
       " 0.050131224805400484,\n",
       " 0.03186695521267561,\n",
       " 0.03874756917357445,\n",
       " 0.041846879083534766,\n",
       " 0.04224037102841098,\n",
       " 0.09676610468886793,\n",
       " 0.09563796292059124,\n",
       " 0.09634174755774438,\n",
       " 0.03254379017342781,\n",
       " 0.03367104150097946,\n",
       " 0.04427471826784313,\n",
       " 0.04344230209445131,\n",
       " 0.04368768548914071,\n",
       " 0.03090148780714063,\n",
       " 0.028141927173627273,\n",
       " 0.03099190496972629,\n",
       " 0.024816994500510833,\n",
       " 0.027515367550008437,\n",
       " 0.023149417283443306,\n",
       " 0.021549043618142603,\n",
       " 0.02888257447630167,\n",
       " 0.03176739842941364,\n",
       " 0.046243567091341205,\n",
       " 0.034538709041145116,\n",
       " 0.028032865518560775,\n",
       " 0.031168956102596387,\n",
       " 0.030642134171945078,\n",
       " 0.029244954171388046,\n",
       " 0.041486350545550096,\n",
       " 0.026633830467129454,\n",
       " 0.030228254275725168,\n",
       " 0.025820447679828194,\n",
       " 0.02837081642492729,\n",
       " 0.026509049534797668,\n",
       " 0.02291642263531685,\n",
       " 0.02317539483308792,\n",
       " 0.04424344885506128,\n",
       " 0.027442504268358734,\n",
       " 0.03418563012229769,\n",
       " 0.03427496179938316,\n",
       " 0.03528220814309622,\n",
       " 0.02683958687402051,\n",
       " 0.027118466280657668,\n",
       " 0.015222682462384304,\n",
       " 0.01775338643175714,\n",
       " 0.018270017077716496,\n",
       " 0.02213893225416541,\n",
       " 0.020737260329322174,\n",
       " 0.019852088083728,\n",
       " 0.023129481157021865,\n",
       " 0.02020014515146613,\n",
       " 0.02308939266103235,\n",
       " 0.02253996995701031,\n",
       " 0.02261062854086049,\n",
       " 0.019946554379368372,\n",
       " 0.035935365047086686,\n",
       " 0.022944488562643528,\n",
       " 0.02429649272166631,\n",
       " 0.025985521846450865,\n",
       " 0.023355067358352244,\n",
       " 0.012512983261224101,\n",
       " 0.017052351153994862,\n",
       " 0.016483259436331298,\n",
       " 0.04769514970205448,\n",
       " 0.04589778807913435,\n",
       " 0.0496071736017863,\n",
       " 0.0466402892023325,\n",
       " 0.05924973835547765,\n",
       " 0.054465731233358385,\n",
       " 0.0588804562886556,\n",
       " 0.0538578774034977,\n",
       " 0.061000274921146534,\n",
       " 0.06190902572767488,\n",
       " 0.06083721153695008,\n",
       " 0.038877309845953154,\n",
       " 0.04435949888208817,\n",
       " 0.041312234668896115,\n",
       " 0.04446141794323921,\n",
       " 0.041385515689336023,\n",
       " 0.0400148459613837,\n",
       " 0.03093388520314225,\n",
       " 0.036384742064722655,\n",
       " 0.03558552438585923,\n",
       " 0.03670610534027219,\n",
       " 0.016745816295345623,\n",
       " 0.019194106213175334,\n",
       " 0.013610753808812848,\n",
       " 0.026206343053755434,\n",
       " 0.017708987446018943,\n",
       " 0.012457487018158039,\n",
       " 0.01864078565267846,\n",
       " 0.027119624145005062,\n",
       " 0.029441028833389282,\n",
       " 0.05198099104494884,\n",
       " 0.06243104667023376,\n",
       " 0.05312729202981653,\n",
       " 0.05263544998035349,\n",
       " 0.041050550367297794,\n",
       " 0.044327413141727445,\n",
       " 0.04619473251032418,\n",
       " 0.04505295030258853,\n",
       " 0.043838181372346546,\n",
       " 0.0511528615338298,\n",
       " 0.03696115022259099,\n",
       " 0.027562126073808896,\n",
       " 0.02401034285624822,\n",
       " 0.042252042073884916,\n",
       " 0.05252218220768304,\n",
       " 0.05426821691383209,\n",
       " 0.024778093628842254,\n",
       " 0.030942502849060912,\n",
       " 0.026863266103740395,\n",
       " 0.026452286716769722,\n",
       " 0.03679899378296207,\n",
       " 0.015342767084283488,\n",
       " 0.017411715515396175,\n",
       " 0.04473025403502914,\n",
       " 0.02789330542744959,\n",
       " 0.029953452861971326,\n",
       " 0.033489750801689096,\n",
       " 0.041050550367297794,\n",
       " 0.044327413141727445,\n",
       " 0.04619473251032418,\n",
       " 0.04505295030258853,\n",
       " 0.043838181372346546,\n",
       " 0.0511528615338298,\n",
       " 0.02446601196954196,\n",
       " 0.034705501049757004,\n",
       " 0.024817807515236465,\n",
       " 0.020933749486825297,\n",
       " 0.04203299122552077,\n",
       " 0.03765833253661791,\n",
       " 0.036218246883329225,\n",
       " 0.033694439153704375,\n",
       " 0.031152482299755018,\n",
       " 0.0360830278446277,\n",
       " 0.03002485227105873,\n",
       " 0.03442848688954937,\n",
       " 0.03719611157631052,\n",
       " 0.027834907623714412,\n",
       " 0.03102156571273146,\n",
       " 0.0873756490361232,\n",
       " 0.11767101244857678,\n",
       " 0.16151584296797714,\n",
       " 0.06895931926555932,\n",
       " 0.06260396423749626,\n",
       " 0.012207723719378312,\n",
       " 0.012701614294201136,\n",
       " 0.012501446095605691,\n",
       " 0.041486350545550096,\n",
       " 0.026633830467129454,\n",
       " 0.030228254275725168,\n",
       " 0.025820447679828194,\n",
       " 0.02837081642492729,\n",
       " 0.02363535612821579,\n",
       " 0.030461080571902648,\n",
       " 0.0356306446523502,\n",
       " 0.039769674409870746,\n",
       " 0.03324295490704201,\n",
       " 0.03913761764440848,\n",
       " 0.025311381290001527,\n",
       " 0.033043887466192245,\n",
       " 0.033210547429255464,\n",
       " 0.029854791560049714,\n",
       " 0.03062706723295409,\n",
       " 0.03265791665762663,\n",
       " 0.033210547429255464,\n",
       " 0.029854791560049714,\n",
       " 0.03062706723295409,\n",
       " 0.03265791665762663,\n",
       " 0.013317240740765224,\n",
       " 0.014838718070552266,\n",
       " 0.015080638294076097,\n",
       " 0.017224225857905274,\n",
       " 0.0217219565436244,\n",
       " 0.024642349028124892,\n",
       " 0.024998009416528723,\n",
       " 0.02657436973161318,\n",
       " 0.06637146028465238,\n",
       " 0.06530796685095491,\n",
       " 0.06801833719015121,\n",
       " 0.04729497124408853,\n",
       " 0.035860574804246426,\n",
       " 0.034833325818181035,\n",
       " 0.037454847142928176,\n",
       " 0.0284816674888134,\n",
       " 0.04587883958593011,\n",
       " 0.029296871563740844,\n",
       " 0.029846626558694345,\n",
       " 0.02919251809362322,\n",
       " 0.034739363261244514,\n",
       " 0.034673681144009934,\n",
       " 0.03598237545652823,\n",
       " 0.03405523096973246,\n",
       " 0.0359123091806065]"
      ]
     },
     "execution_count": 811,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ratio_radios_por_rama"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "calculo tortuosidad, numero de bifurcaciones para los generados"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 812,
   "metadata": {},
   "outputs": [],
   "source": [
    "a = [1.,1.,1.]\n",
    "mult = torch.Tensor(a)\n",
    "latent_size = 64\n",
    "Grassdecoder = GRASSDecoder(latent_size=latent_size, hidden_size=256, mult = mult.to(device))\n",
    "Grassdecoder = Grassdecoder.to(device)\n",
    "Grassdecoder.eval()\n",
    "checkpoint = torch.load(\"output/\" + dataset_name + \"P\" +p + \"eps\" + eps+ \"-best-nuevoweight.pth\")\n",
    "#checkpoint = torch.load(\"output/\" + dataset_name + eps+ \"-best.pth\")\n",
    "\n",
    "\n",
    "\n",
    "Grassdecoder.load_state_dict(checkpoint['decoder_state_dict'])\n",
    "epoch = checkpoint['epoch']\n",
    "\n",
    "i = 0\n",
    "n = 100\n",
    "tortuosidad_generados = []\n",
    "n_bif_generados = []\n",
    "n_nodes_generados = []\n",
    "largo_generados = []\n",
    "ratio_radios_generados = []\n",
    "ratio_radios_por_rama_generados = []\n",
    "ratio_largos_por_rama_generados = []\n",
    "with torch.no_grad():\n",
    "    for i in range(n):\n",
    "        noise = torch.randn(batch_size, latent_size).to(device)\n",
    "        generated_images = decode_testing(noise, 100, Grassdecoder, mult, 1)\n",
    "        count = []\n",
    "        numerar_nodos(generated_images, count)\n",
    "        graph = nx.Graph()\n",
    "        generated_images.toGraph(graph, 0 , True, flag = 0)\n",
    "        degree1 = []\n",
    "\n",
    "        for node in graph.nodes():\n",
    "            if graph.degree(node) == 1:\n",
    "                degree1.append(node)\n",
    "        caminos = [ nx.shortest_path( graph, degree1[0], t ) for t in degree1[1:] ]\n",
    "        subarboles = []\n",
    "        for g in caminos:\n",
    "            grafo = nx.Graph()\n",
    "            grafo.add_nodes_from(g)\n",
    "            et = []\n",
    "            for node in grafo.nodes:\n",
    "                edge = graph[node]\n",
    "                edges = [a for a in list(edge.keys()) if a in g]\n",
    "                edge_tuple = [(a,b) for a,b in zip([node]*len(edges), edges) ]\n",
    "                for tuple in edge_tuple:\n",
    "                    et.append(tuple)\n",
    "            grafo.add_edges_from(et)\n",
    "    \n",
    "            aRecorrer = []\n",
    "            numeroNodoInicial = degree1[0]\n",
    "            r = graph.nodes[numeroNodoInicial]['radio']\n",
    "            rad = list(graph.nodes[numeroNodoInicial]['posicion'])\n",
    "            rad.append(r)\n",
    "\n",
    "            nodoRaiz = Node( numeroNodoInicial, radius =  rad )\n",
    "\n",
    "            for vecino in grafo.neighbors( numeroNodoInicial ):\n",
    "                if vecino != numeroNodoInicial:\n",
    "                    aRecorrer.append( (vecino, numeroNodoInicial,nodoRaiz ) )\n",
    "            \n",
    "            while len(aRecorrer) != 0:\n",
    "                nodoAAgregar, numeroNodoPadre,nodoPadre = aRecorrer.pop(0)\n",
    "                r = graph.nodes[nodoAAgregar]['radio']\n",
    "                rad = list(graph.nodes[nodoAAgregar]['posicion'])\n",
    "                rad.append(r)\n",
    "                nodoActual = Node( nodoAAgregar, radius =  rad)\n",
    "                nodoPadre.agregarHijo( nodoActual )\n",
    "                for vecino in grafo.neighbors( nodoAAgregar ):\n",
    "                    if vecino != numeroNodoPadre:\n",
    "                        aRecorrer.append( (vecino, nodoAAgregar,nodoActual) )\n",
    "            subarboles.append(nodoRaiz)\n",
    "            tortuosidad_generados.append(tortuosity(nodoRaiz))\n",
    "            radius_list = []\n",
    "            collect_radius(nodoRaiz, radius_list)\n",
    "            average_radius = sum(radius_list) / len(radius_list)\n",
    "            ratio_radios_por_rama_generados.append(average_radius)\n",
    "            length_list = []\n",
    "            collect_length(nodoRaiz, length_list)\n",
    "            average_length = sum(length_list) / len(length_list)\n",
    "            ratio_largos_por_rama_generados.append(average_length)\n",
    "        li = []\n",
    "        generated_images.traverseInorderChilds(generated_images, li)\n",
    "        two = [a for a in li if a == 2]\n",
    "        n_bif_generados.append(len(two))\n",
    "        count = []\n",
    "        n = generated_images.countNodes(generated_images, count)\n",
    "        n_nodes_generados.append(len(n))\n",
    "        traverse(generated_images)\n",
    "        \n",
    "        total_path_length_value = tree_length(generated_images)\n",
    "        largo_generados.append(total_path_length_value.cpu().item())\n",
    "        features = []\n",
    "        \n",
    "        traversefeatures(generated_images, features)\n",
    "        ratio_radios_generados.append(sum(features)/len(features))\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 813,
   "metadata": {},
   "outputs": [],
   "source": [
    "def tr(root):\n",
    "        if root is not None:\n",
    "            tr(root.left)\n",
    "            root.radius[0][3]=root.radius[0][3]/10\n",
    "            if root.radius[0][3]<0:\n",
    "                root.radius[0][3]=0\n",
    "            tr(root.right)\n",
    "            return"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 814,
   "metadata": {},
   "outputs": [],
   "source": [
    "enmax_palette = [\"#FE7871\", \"#B50603\"]\n",
    "\n",
    "color_codes_wanted = ['pink', 'blue']\n",
    "col = lambda x: enmax_palette[color_codes_wanted.index(x)]\n",
    "#sns.set_palette(palette=enmax_palette)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "HISTOGRAMAS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 815,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.03864957364878553"
      ]
     },
     "execution_count": 815,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.average(ratio_radios_por_rama_generados)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 816,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0348992936776117"
      ]
     },
     "execution_count": 816,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.average(ratio_radios_por_rama)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 817,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "110.74600536566068"
      ]
     },
     "execution_count": 817,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.average(ratio_radios_por_rama_generados)*100/np.average(ratio_radios_por_rama)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 818,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plt.figure(figsize=(10,20))\n",
    "##RADIO por rama\n",
    "sns.displot([ratio_radios_por_rama, ratio_radios_por_rama_generados], multiple=\"dodge\", height=10, aspect=1, legend=False, palette = enmax_palette)\n",
    "ax = plt.gca()\n",
    "\n",
    "#ax.set_title(\"Radius P{} {}\".format(p, dataset_name), fontsize=35)\n",
    "\n",
    "#ax.set_xlabel(\"Radius\", fontsize = 25)\n",
    "ax.set_ylabel(\"  \", fontsize = 25)\n",
    "plt.xticks(fontsize=15)\n",
    "plt.yticks(fontsize=15)\n",
    "\n",
    "ax.xaxis.set_major_locator(MaxNLocator(nbins=1))\n",
    "ax.yaxis.set_major_locator(MaxNLocator(nbins=1))\n",
    "# get the patches for each dataset\n",
    "patches = ax.patches\n",
    "l = int(len(patches)/2)\n",
    "patches_radius = patches[:l]\n",
    "patches_radius_generados = patches[l:]\n",
    "\n",
    "# get the counts for ratio_radios\n",
    "counts_radius = [patch.get_height() for patch in patches_radius]\n",
    "\n",
    "# get the counts for ratio_radios_generados\n",
    "counts_radius_generados = [patch.get_height() for patch in patches_radius_generados]\n",
    "\n",
    "#ax.legend(labels=[\"Synthetic samples\", \"Real samples\"], fontsize = 25)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 819,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.857417031235328"
      ]
     },
     "execution_count": 819,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.dot(counts_radius, counts_radius_generados)/(np.linalg.norm(counts_radius)*np.linalg.norm(counts_radius_generados))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 820,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.8676460532613777"
      ]
     },
     "execution_count": 820,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "emd_radios = wasserstein_distance(range(len(counts_radius)), range(len(counts_radius_generados)), counts_radius, counts_radius_generados)\n",
    "emd_radios"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 821,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plt.figure(figsize=(10,20))\n",
    "##RLARGO por rama\n",
    "sns.displot([ratio_largos_por_rama, ratio_largos_por_rama_generados], multiple=\"dodge\", height=10, aspect=1, legend=False, palette = enmax_palette)\n",
    "ax = plt.gca()\n",
    "\n",
    "#ax.set_title(\"Radius P{} {}\".format(p, dataset_name), fontsize=35)\n",
    "\n",
    "#ax.set_xlabel(\"Radius\", fontsize = 25)\n",
    "ax.set_ylabel(\"  \", fontsize = 25)\n",
    "plt.xticks(fontsize=15)\n",
    "plt.yticks(fontsize=15)\n",
    "\n",
    "ax.xaxis.set_major_locator(MaxNLocator(nbins=1))\n",
    "ax.yaxis.set_major_locator(MaxNLocator(nbins=1))\n",
    "# get the patches for each dataset\n",
    "patches = ax.patches\n",
    "l = int(len(patches)/2)\n",
    "patches_radius = patches[:l]\n",
    "patches_radius_generados = patches[l:]\n",
    "\n",
    "# get the counts for ratio_radios\n",
    "counts_radius = [patch.get_height() for patch in patches_radius]\n",
    "\n",
    "# get the counts for ratio_radios_generados\n",
    "counts_radius_generados = [patch.get_height() for patch in patches_radius_generados]\n",
    "\n",
    "#ax.legend(labels=[\"Synthetic samples\", \"Real samples\"], fontsize = 25)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 822,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8223253582290164"
      ]
     },
     "execution_count": 822,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.dot(counts_radius, counts_radius_generados)/(np.linalg.norm(counts_radius)*np.linalg.norm(counts_radius_generados))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 823,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.3018109710234116"
      ]
     },
     "execution_count": 823,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "emd_radios = wasserstein_distance(range(len(counts_radius)), range(len(counts_radius_generados)), counts_radius, counts_radius_generados)\n",
    "emd_radios"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 824,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plt.figure(figsize=(10,20))\n",
    "##RADIO\n",
    "sns.displot([ratio_radios, ratio_radios_generados], multiple=\"dodge\", height=10, aspect=1, legend=False, palette = enmax_palette)\n",
    "ax = plt.gca()\n",
    "\n",
    "#ax.set_title(\"Radius P{} {}\".format(p, dataset_name), fontsize=35)\n",
    "\n",
    "#ax.set_xlabel(\"Radius\", fontsize = 25)\n",
    "ax.set_ylabel(\"  \", fontsize = 25)\n",
    "plt.xticks(fontsize=15)\n",
    "plt.yticks(fontsize=15)\n",
    "\n",
    "ax.xaxis.set_major_locator(MaxNLocator(nbins=1))\n",
    "ax.yaxis.set_major_locator(MaxNLocator(nbins=1))\n",
    "# get the patches for each dataset\n",
    "patches = ax.patches\n",
    "l = int(len(patches)/2)\n",
    "patches_radius = patches[:l]\n",
    "patches_radius_generados = patches[l:]\n",
    "\n",
    "# get the counts for ratio_radios\n",
    "counts_radius = [patch.get_height() for patch in patches_radius]\n",
    "\n",
    "# get the counts for ratio_radios_generados\n",
    "counts_radius_generados = [patch.get_height() for patch in patches_radius_generados]\n",
    "\n",
    "#ax.legend(labels=[\"Synthetic samples\", \"Real samples\"], fontsize = 25)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 825,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9057538948140841"
      ]
     },
     "execution_count": 825,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.dot(counts_radius, counts_radius_generados)/(np.linalg.norm(counts_radius)*np.linalg.norm(counts_radius_generados))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 826,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean: 0.9248455743132006\n",
      "Standard Deviation: 0.0309613557942636\n"
     ]
    }
   ],
   "source": [
    "values = [0.946306381992226, 0.8893526816459894, 0.9388776593013862]\n",
    "\n",
    "# Mean\n",
    "mean = statistics.mean(values)\n",
    "print(\"Mean:\", mean)\n",
    "\n",
    "# Standard Deviation\n",
    "std_dev = statistics.stdev(values)\n",
    "print(\"Standard Deviation:\", std_dev)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 827,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "t-statistic: nan\n",
      "p-value: nan\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/eiarussi/.local/lib/python3.10/site-packages/scipy/stats/_axis_nan_policy.py:502: RuntimeWarning: Precision loss occurred in moment calculation due to catastrophic cancellation. This occurs when the data are nearly identical. Results may be unreliable.\n",
      "  res = hypotest_fun_out(*samples, **kwds)\n",
      "/home/eiarussi/.local/lib/python3.10/site-packages/scipy/stats/_stats_py.py:1214: RuntimeWarning: divide by zero encountered in divide\n",
      "  var *= np.divide(n, n-ddof)  # to avoid error on division by zero\n",
      "/home/eiarussi/.local/lib/python3.10/site-packages/scipy/stats/_stats_py.py:1214: RuntimeWarning: invalid value encountered in scalar multiply\n",
      "  var *= np.divide(n, n-ddof)  # to avoid error on division by zero\n"
     ]
    }
   ],
   "source": [
    "from scipy import stats\n",
    "\n",
    "# Example cosine similarity score (replace with your actual score)\n",
    "cosine_similarity_score = 0.946306381992226\n",
    "\n",
    "# Define null hypothesis value (e.g., expected similarity score)\n",
    "null_hypothesis_value = 1  # Adjust this value according to your hypothesis\n",
    "\n",
    "# Perform one-sample t-test\n",
    "t_statistic, p_value = stats.ttest_1samp(cosine_similarity_score, null_hypothesis_value)\n",
    "print(\"t-statistic:\", t_statistic)\n",
    "print(\"p-value:\", p_value)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 828,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Observed cosine similarity score is consistent with the variability expected under bootstrapping.\n",
      "Bootstrap Mean: 0.9252566311006762\n",
      "Bootstrap Standard Deviation: 0.014603613740931007\n",
      "95% Confidence Interval: [0.90586101 0.94383014]\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "\n",
    "# Example cosine similarity scores (replace with your actual scores)\n",
    "cosine_similarity_scores = [0.946306381992226, 0.8893526816459894, 0.9388776593013862]\n",
    "\n",
    "# Number of bootstrap samples\n",
    "num_bootstrap_samples = 1000\n",
    "\n",
    "# Generate bootstrap samples\n",
    "bootstrap_scores = []\n",
    "for _ in range(num_bootstrap_samples):\n",
    "    bootstrap_sample = np.random.choice(cosine_similarity_scores, size=len(cosine_similarity_scores), replace=True)\n",
    "    bootstrap_scores.append(np.mean(bootstrap_sample))\n",
    "\n",
    "# Calculate statistics of interest\n",
    "bootstrap_mean = np.mean(bootstrap_scores)\n",
    "bootstrap_std = np.std(bootstrap_scores)\n",
    "confidence_interval = np.percentile(bootstrap_scores, [5, 95])\n",
    "\n",
    "# Assess fidelity\n",
    "if confidence_interval[0] <= np.mean(cosine_similarity_scores) <= confidence_interval[1]:\n",
    "    print(\"Observed cosine similarity score is consistent with the variability expected under bootstrapping.\")\n",
    "else:\n",
    "    print(\"Observed cosine similarity score is not consistent with the variability expected under bootstrapping.\")\n",
    "\n",
    "print(\"Bootstrap Mean:\", bootstrap_mean)\n",
    "print(\"Bootstrap Standard Deviation:\", bootstrap_std)\n",
    "print(\"95% Confidence Interval:\", confidence_interval)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "tortuosidad"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 829,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA+MAAAPdCAYAAADh9LkNAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8hTgPZAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAx30lEQVR4nO3dfZBV9X348c8islCEJYthHyYgG2OKig8Rdd1gMxp3RJJYGUkzZHCKlpFOClYkPjERH4iGalLDoAg1k2odxSZ2KkY7pWNRYTSICrFTIyGYMoHW3aVRd1dIWQm7vz/y8yYbWUBZPpe7+3rN3Bnv+Z575nPNzXHec8/eU9bV1dUVAAAAQJoBxR4AAAAA+hsxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+P70dXVFe3t7eFW7AAAAPQmMb4f7777blRUVMS7775b7FEAAADoQ8Q4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJBhZ7AA7d5845J3Y0N+93n1HV1bH2xReTJgIAAGB/xHgfsKO5OTZ947r97nPiHd9OmgYAAIADcZk6AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJBhZ7AErDxAlnREtTU4/rVTU18cKGjYkTAQAAlC4xzkFpaWqKh8ef1OP6Za+9njgNAABAaXOZOgAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJDsiY3zt2rVx8cUXR21tbZSVlcXKlSsLa3v27IkbbrghTjnllBg6dGjU1tbGn//5n8ebb77Z7Rhvv/12TJ8+PYYPHx4jRoyImTNnxs6dO5PfCQAAAHzQERnju3btitNOOy2WLl36gbVf//rXsXHjxliwYEFs3Lgx/vmf/zk2b94cf/qnf9ptv+nTp8dPf/rTePrpp+Opp56KtWvXxqxZs7LeAgAAAPRoYLEH2JfJkyfH5MmT97lWUVERTz/9dLdt9957b5x99tmxbdu2GDNmTGzatClWrVoVL7/8cpx55pkREXHPPffEF77whfjOd74TtbW1+zx2R0dHdHR0FJ63t7f30jsCAACA3zkivxn/sNra2qKsrCxGjBgRERHr1q2LESNGFEI8IqKxsTEGDBgQ69ev7/E4ixYtioqKisJj9OjRh3t0AAAA+qGSj/Hdu3fHDTfcEF/96ldj+PDhERHR3Nwco0aN6rbfwIEDo7KyMpqbm3s81vz586Otra3w2L59+2GdHQAAgP7piLxM/WDt2bMnvvKVr0RXV1csW7bskI9XXl4e5eXlvTAZAAAA9KxkY/z9EP/lL38ZzzzzTOFb8YiI6urq2LFjR7f9f/Ob38Tbb78d1dXV2aMCAABANyV5mfr7Ib5ly5b493//9xg5cmS39YaGhmhtbY0NGzYUtj3zzDPR2dkZ9fX12eMCAABAN0fkN+M7d+6MN954o/B869at8eqrr0ZlZWXU1NTEl7/85di4cWM89dRTsXfv3sLfgVdWVsagQYPixBNPjIsuuiiuvPLKWL58eezZsyfmzJkT06ZN6/GX1AEAACDLERnjr7zySpx//vmF5/PmzYuIiBkzZsStt94aP/rRjyIi4vTTT+/2umeffTbOO++8iIh45JFHYs6cOXHBBRfEgAEDYurUqbFkyZKU+QEAAGB/jsgYP++886Krq6vH9f2tva+ysjJWrFjRm2MBAABAryjJvxkHAACAUibGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAINnAYg/AkeFz55wTO5qbe1x/p60tcRoAAIC+TYwTERE7mptj0zeu63H92KuvTZwGAACgb3OZOgAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJDsiY3zt2rVx8cUXR21tbZSVlcXKlSu7rXd1dcXNN98cNTU1MWTIkGhsbIwtW7Z02+ftt9+O6dOnx/Dhw2PEiBExc+bM2LlzZ+K7AAAAgH07ImN8165dcdppp8XSpUv3uX7XXXfFkiVLYvny5bF+/foYOnRoTJo0KXbv3l3YZ/r06fHTn/40nn766Xjqqadi7dq1MWvWrKy3AAAAAD0aWOwB9mXy5MkxefLkfa51dXXF4sWL46abbopLLrkkIiIeeuihqKqqipUrV8a0adNi06ZNsWrVqnj55ZfjzDPPjIiIe+65J77whS/Ed77znaitrd3nsTs6OqKjo6PwvL29vZffGQAAAByh34zvz9atW6O5uTkaGxsL2yoqKqK+vj7WrVsXERHr1q2LESNGFEI8IqKxsTEGDBgQ69ev7/HYixYtioqKisJj9OjRh++NAAAA0G+VXIw3NzdHRERVVVW37VVVVYW15ubmGDVqVLf1gQMHRmVlZWGffZk/f360tbUVHtu3b+/l6QEAAOAIvUy9WMrLy6O8vLzYYwAAANDHldw349XV1RER0dLS0m17S0tLYa26ujp27NjRbf03v/lNvP3224V9AAAAoFhKLsbr6uqiuro6Vq9eXdjW3t4e69evj4aGhoiIaGhoiNbW1tiwYUNhn2eeeSY6Ozujvr4+fWYAAAD4fUfkZeo7d+6MN954o/B869at8eqrr0ZlZWWMGTMm5s6dG7fffnuccMIJUVdXFwsWLIja2tqYMmVKRESceOKJcdFFF8WVV14Zy5cvjz179sScOXNi2rRpPf6SOgAAAGQ5ImP8lVdeifPPP7/wfN68eRERMWPGjHjwwQfj+uuvj127dsWsWbOitbU1zj333Fi1alUMHjy48JpHHnkk5syZExdccEEMGDAgpk6dGkuWLEl/LwAAAPCHjsgYP++886Krq6vH9bKysli4cGEsXLiwx30qKytjxYoVh2M8AAAAOCRHZIzTN02ccEa0NDX1uF5VUxMvbNiYOBEAAEBxiHHStDQ1xcPjT+px/bLXXk+cBgAAoHhK7tfUAQAAoNSJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIFlJxvjevXtjwYIFUVdXF0OGDInjjz8+vvnNb0ZXV1dhn66urrj55pujpqYmhgwZEo2NjbFly5YiTg0AAAC/VZIxfuedd8ayZcvi3nvvjU2bNsWdd94Zd911V9xzzz2Ffe66665YsmRJLF++PNavXx9Dhw6NSZMmxe7du4s4OQAAAEQMLPYAH8WPf/zjuOSSS+KLX/xiRESMHTs2Hn300XjppZci4rffii9evDhuuummuOSSSyIi4qGHHoqqqqpYuXJlTJs2bZ/H7ejoiI6OjsLz9vb2w/xOAAAA6I9K8pvxz372s7F69er4+c9/HhER//Ef/xHPP/98TJ48OSIitm7dGs3NzdHY2Fh4TUVFRdTX18e6det6PO6iRYuioqKi8Bg9evThfSMAAAD0SyX5zfiNN94Y7e3tMW7cuDjqqKNi7969cccdd8T06dMjIqK5uTkiIqqqqrq9rqqqqrC2L/Pnz4958+YVnre3twtyAAAAel1JxvgPf/jDeOSRR2LFihVx8sknx6uvvhpz586N2tramDFjxkc+bnl5eZSXl/fipAAAAPBBJRnj1113Xdx4442Fv/0+5ZRT4pe//GUsWrQoZsyYEdXV1RER0dLSEjU1NYXXtbS0xOmnn16MkQEAAKCgJP9m/Ne//nUMGNB99KOOOio6OzsjIqKuri6qq6tj9erVhfX29vZYv359NDQ0pM4KAAAAf6gkvxm/+OKL44477ogxY8bEySefHD/5yU/i7rvvjr/4i7+IiIiysrKYO3du3H777XHCCSdEXV1dLFiwIGpra2PKlCnFHR4AAIB+ryRj/J577okFCxbEX/3VX8WOHTuitrY2/vIv/zJuvvnmwj7XX3997Nq1K2bNmhWtra1x7rnnxqpVq2Lw4MFFnBwAAABKNMaHDRsWixcvjsWLF/e4T1lZWSxcuDAWLlyYNxgAAAAchJL8m3EAAAAoZWIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEhWkvcZp3+aOOGMaGlq2u8+VTU18cKGjUkTAQAAfDRinJLR0tQUD48/ab/7XPba60nTAAAAfHQuUwcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgWcnG+P/8z//EZZddFiNHjowhQ4bEKaecEq+88kphvaurK26++eaoqamJIUOGRGNjY2zZsqWIEwMAAMBvlWSMv/POOzFx4sQ4+uij41//9V/j9ddfj7/927+Nj33sY4V97rrrrliyZEksX7481q9fH0OHDo1JkybF7t27izg5AAAARAws9gAfxZ133hmjR4+OBx54oLCtrq6u8M9dXV2xePHiuOmmm+KSSy6JiIiHHnooqqqqYuXKlTFt2rT0mQEAAOB9JfnN+I9+9KM488wz48/+7M9i1KhR8ZnPfCa+973vFda3bt0azc3N0djYWNhWUVER9fX1sW7duh6P29HREe3t7d0eAAAA0NtKMsb/67/+K5YtWxYnnHBC/Nu//Vt87Wtfi7/+67+Of/iHf4iIiObm5oiIqKqq6va6qqqqwtq+LFq0KCoqKgqP0aNHH743AQAAQL9VkjHe2dkZZ5xxRnzrW9+Kz3zmMzFr1qy48sorY/ny5Yd03Pnz50dbW1vhsX379l6aGAAAAH6nJGO8pqYmTjrppG7bTjzxxNi2bVtERFRXV0dEREtLS7d9WlpaCmv7Ul5eHsOHD+/2AAAAgN5WkjE+ceLE2Lx5c7dtP//5z+O4446LiN/+mFt1dXWsXr26sN7e3h7r16+PhoaG1FkBAADgD5Xkr6lfc8018dnPfja+9a1vxVe+8pV46aWX4v7774/7778/IiLKyspi7ty5cfvtt8cJJ5wQdXV1sWDBgqitrY0pU6YUd3gAAAD6vZKM8bPOOisef/zxmD9/fixcuDDq6upi8eLFMX369MI+119/fezatStmzZoVra2tce6558aqVati8ODBRZwcAAAASjTGIyK+9KUvxZe+9KUe18vKymLhwoWxcOHCxKkAAADgwEryb8YBAACglIlxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABINrDYA9A3vNPWFuPGjj3gPgAAAIhxektXV2z6xnX73eXYq69NGgYAAODI5jJ1AAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkg0s9gCQaeKEM6KlqanH9aqamnhhw8bEiQAAgP5IjNOvtDQ1xcPjT+px/bLXXk+cBgAA6K9cpg4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAECygcUeAN73TltbjBs7dr/rAAAAfYEY58jR1RWbvnFdj8vHXn1t4jAAAACHj8vUAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgWZ+I8b/5m7+JsrKymDt3bmHb7t27Y/bs2TFy5Mg45phjYurUqdHS0lK8IQEAAOD/K/kYf/nll+Pv/u7v4tRTT+22/Zprroknn3wyHnvssVizZk28+eabcemllxZpSgAAAPidko7xnTt3xvTp0+N73/tefOxjHytsb2tri+9///tx9913x+c///mYMGFCPPDAA/HjH/84XnzxxSJODAAAACUe47Nnz44vfvGL0djY2G37hg0bYs+ePd22jxs3LsaMGRPr1q3r8XgdHR3R3t7e7QEAAAC9bWCxB/io/vEf/zE2btwYL7/88gfWmpubY9CgQTFixIhu26uqqqK5ubnHYy5atChuu+223h4VAAAAuinJb8a3b98eV199dTzyyCMxePDgXjvu/Pnzo62trfDYvn17rx0bAAAA3leSMb5hw4bYsWNHnHHGGTFw4MAYOHBgrFmzJpYsWRIDBw6MqqqqeO+996K1tbXb61paWqK6urrH45aXl8fw4cO7PQAAAKC3leRl6hdccEH853/+Z7dtV1xxRYwbNy5uuOGGGD16dBx99NGxevXqmDp1akREbN68ObZt2xYNDQ3FGBkAAAAKSjLGhw0bFuPHj++2bejQoTFy5MjC9pkzZ8a8efOisrIyhg8fHldddVU0NDTEOeecU4yRAQAAoKAkY/xgfPe7340BAwbE1KlTo6OjIyZNmhT33XdfsccCAACAvhPjzz33XLfngwcPjqVLl8bSpUuLMxAAAAD0oM/EOGSZOOGMaGlq6nG9qqYmXtiwMXEiAACg1Ihx+JBampri4fEn9bh+2WuvJ04DAACUopK8tRkAAACUMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDK/pt5PvPWrX8Wnamt6XG9rfzdxGgAAgP5NjPcTnZ1793s7ri8+/+PEaQAAAPo3l6kDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBsYLEHgP5m4oQzoqWpab/7VNXUxAsbNiZNBAAAZBPjkKylqSkeHn/Sfve57LXXk6YBAACKwWXqAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJSjLGFy1aFGeddVYMGzYsRo0aFVOmTInNmzd322f37t0xe/bsGDlyZBxzzDExderUaGlpKdLEAAAA8DslGeNr1qyJ2bNnx4svvhhPP/107NmzJy688MLYtWtXYZ9rrrkmnnzyyXjsscdizZo18eabb8all15axKkBAADgtwYWe4CPYtWqVd2eP/jggzFq1KjYsGFDfO5zn4u2trb4/ve/HytWrIjPf/7zERHxwAMPxIknnhgvvvhinHPOOfs8bkdHR3R0dBSet7e3H743AQAAQL9Vkt+M/6G2traIiKisrIyIiA0bNsSePXuisbGxsM+4ceNizJgxsW7duh6Ps2jRoqioqCg8Ro8efXgHBwAAoF8q+Rjv7OyMuXPnxsSJE2P8+PEREdHc3ByDBg2KESNGdNu3qqoqmpubezzW/Pnzo62trfDYvn374RwdAACAfqokL1P/fbNnz47XXnstnn/++UM+Vnl5eZSXl/fCVAAAANCzkv5mfM6cOfHUU0/Fs88+G5/4xCcK26urq+O9996L1tbWbvu3tLREdXV18pQAAADQXUnGeFdXV8yZMycef/zxeOaZZ6Kurq7b+oQJE+Loo4+O1atXF7Zt3rw5tm3bFg0NDdnjAgAAQDcleZn67NmzY8WKFfHEE0/EsGHDCn8HXlFREUOGDImKioqYOXNmzJs3LyorK2P48OFx1VVXRUNDQ4+/pA4AAABZSjLGly1bFhER5513XrftDzzwQFx++eUREfHd7343BgwYEFOnTo2Ojo6YNGlS3HfffcmTAgAAwAeVZIx3dXUdcJ/BgwfH0qVLY+nSpQkTAQAAwMEryb8ZBwAAgFImxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACDZwGIPAL3pnba2GDd27H7XAQAAik2M07d0dcWmb1zX4/KxV1+bOAwAAMC+uUwdAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZAOLPQDw4U2ccEa0NDX1uF5VUxMvbNiYOBEAAPBhiHEoQS1NTfHw+JN6XL/stdcTpwEAAD4sl6kDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJBPjAAAAkEyMAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJBhZ7ADiSvNPWFuPGjj3gPgAAAIdCjMPv6+qKTd+4br+7HHv1tUnDAAAAfZXL1AEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACDZwGIPABTHxAlnREtTU4/rVTU18cKGjYkTAQBA/yHGoZ9qaWqKh8ef1OP6Za+9njgNAAD0Ly5TBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEgmxgEAACCZGAcAAIBkYhwAAACSiXEAAABINrDYA0Bf805bW4wbO3a/6xnHONwmTjgjWpqaelyvqqmJFzZsPOzHAACAUiTGobd1dcWmb1zX4/KxV1+bc4zDrKWpKR4ef1KP65e99nrKMQAAoBS5TB0AAACSiXEAAABIJsYBAAAgmRgHAACAZGIcAAAAkvk1dQB6hVvVAQAcPDEOQK9wqzoAgIPnMnUAAABIJsYBAAAgmRgHAACAZGIcAAAAkolxAAAASCbGAQAAIJlbm0Ef9E5bW4wbO/aA+xzKMQ70+v7E/bUBAPiwxDj0RV1dsekb1+13l2OvvvaQjnHA1/cj7q8NAMCH5TJ1AAAASCbGAQAAIJkYBwAAgGRiHAAAAJKJcQAAAEjm19SBw6I3bq/WXxzo1mgRbo8GANDXiHHg8OiN26v1Ewe6NVqE26MBAPQ1LlMHAACAZGIcAAAAkolxAAAASCbGAQAAIJkYBwAAgGR+TR0oWb1xS7ADHaM3bil2oNu89cYt3jLeB6XlSPhMHAkzAJDrUM/9/em/HWIcKFm9cUuwAx2jV24pdoDbvPXGLd5S3gcl5Uj4TBwJMwCQ61DP/f3pvx0uUwcAAIBkYhwAAACS9fkYX7p0aYwdOzYGDx4c9fX18dJLLxV7JAAAAPq5Ph3jP/jBD2LevHlxyy23xMaNG+O0006LSZMmxY4dO4o9GgAAAP1Yn/4Bt7vvvjuuvPLKuOKKKyIiYvny5fEv//Iv8fd///dx4403fmD/jo6O6OjoKDxv+/+/cNze3p4z8Ee0t7Mz2v/v//a7T1dXV+z6zW/2u76/Yxzq6zNm6I1jeB9H1gydnZ37/f9fZ2fnAf9dHuoxDvT6gzlGb3wmMt7HoToSZuB3joT/PY6EGQDIdajn/r70345hw4ZFWVlZj+tlXV1dXYnzpHnvvffij/7oj+Kf/umfYsqUKYXtM2bMiNbW1njiiSc+8Jpbb701brvttsQpAQAA6Iva2tpi+PDhPa732W/Gf/WrX8XevXujqqqq2/aqqqr42c9+ts/XzJ8/P+bNm1d43tnZGW+//XaMHDky3n333Rg9enRs3759v/9CodS0t7f7bNMn+WzTV/ls0xf5XNMXDRs2bL/rfTbGP4ry8vIoLy/vtm3EiBEREYXLC4YPH+4EQZ/ks01f5bNNX+WzTV/kc01/0md/wO3YY4+No446KlpaWrptb2lpierq6iJNBQAAAH04xgcNGhQTJkyI1atXF7Z1dnbG6tWro6GhoYiTAQAA0N/16cvU582bFzNmzIgzzzwzzj777Fi8eHHs2rWr8OvqH0Z5eXnccsstH7iMHUqdzzZ9lc82fZXPNn2RzzX9UZ/9NfX33XvvvfHtb387mpub4/TTT48lS5ZEfX19sccCAACgH+vzMQ4AAABHmj77N+MAAABwpBLjAAAAkEyMAwAAQDIxDgAAAMnE+EFaunRpjB07NgYPHhz19fXx0ksvFXskOCS33nprlJWVdXuMGzeu2GPBh7J27dq4+OKLo7a2NsrKymLlypXd1ru6uuLmm2+OmpqaGDJkSDQ2NsaWLVuKMyx8CAf6bF9++eUfOIdfdNFFxRkWPoRFixbFWWedFcOGDYtRo0bFlClTYvPmzd322b17d8yePTtGjhwZxxxzTEydOjVaWlqKNDEcPmL8IPzgBz+IefPmxS233BIbN26M0047LSZNmhQ7duwo9mhwSE4++eRoamoqPJ5//vlijwQfyq5du+K0006LpUuX7nP9rrvuiiVLlsTy5ctj/fr1MXTo0Jg0aVLs3r07eVL4cA702Y6IuOiii7qdwx999NHECeGjWbNmTcyePTtefPHFePrpp2PPnj1x4YUXxq5duwr7XHPNNfHkk0/GY489FmvWrIk333wzLr300iJODYeHW5sdhPr6+jjrrLPi3nvvjYiIzs7OGD16dFx11VVx4403Fnk6+GhuvfXWWLlyZbz66qvFHgV6RVlZWTz++OMxZcqUiPjtt+K1tbXx9a9/Pa699tqIiGhra4uqqqp48MEHY9q0aUWcFg7eH362I377zXhra+sHvjGHUvO///u/MWrUqFizZk187nOfi7a2tvj4xz8eK1asiC9/+csREfGzn/0sTjzxxFi3bl2cc845RZ4Yeo9vxg/gvffeiw0bNkRjY2Nh24ABA6KxsTHWrVtXxMng0G3ZsiVqa2vjk5/8ZEyfPj22bdtW7JGg12zdujWam5u7nb8rKiqivr7e+Zs+4bnnnotRo0bFH//xH8fXvva1eOutt4o9EnxobW1tERFRWVkZEREbNmyIPXv2dDt3jxs3LsaMGePcTZ8jxg/gV7/6Vezduzeqqqq6ba+qqorm5uYiTQWHrr6+Ph588MFYtWpVLFu2LLZu3Rp/8id/Eu+++26xR4Ne8f452vmbvuiiiy6Khx56KFavXh133nlnrFmzJiZPnhx79+4t9mhw0Do7O2Pu3LkxceLEGD9+fET89tw9aNCgGDFiRLd9nbvpiwYWewCgOCZPnlz451NPPTXq6+vjuOOOix/+8Icxc+bMIk4GwIH8/p9ZnHLKKXHqqafG8ccfH88991xccMEFRZwMDt7s2bPjtdde85s19Fu+GT+AY489No466qgP/IJjS0tLVFdXF2kq6H0jRoyIT3/60/HGG28UexToFe+fo52/6Q8++clPxrHHHuscTsmYM2dOPPXUU/Hss8/GJz7xicL26urqeO+996K1tbXb/s7d9EVi/AAGDRoUEyZMiNWrVxe2dXZ2xurVq6OhoaGIk0Hv2rlzZ/ziF7+ImpqaYo8CvaKuri6qq6u7nb/b29tj/fr1zt/0Of/93/8db731lnM4R7yurq6YM2dOPP744/HMM89EXV1dt/UJEybE0Ucf3e3cvXnz5ti2bZtzN32Oy9QPwrx582LGjBlx5plnxtlnnx2LFy+OXbt2xRVXXFHs0eAju/baa+Piiy+O4447Lt5888245ZZb4qijjoqvfvWrxR4NDtrOnTu7fRO4devWePXVV6OysjLGjBkTc+fOjdtvvz1OOOGEqKuriwULFkRtbW23X6WGI9H+PtuVlZVx2223xdSpU6O6ujp+8YtfxPXXXx+f+tSnYtKkSUWcGg5s9uzZsWLFinjiiSdi2LBhhb8Dr6ioiCFDhkRFRUXMnDkz5s2bF5WVlTF8+PC46qqroqGhwS+p0+e4tdlBuvfee+Pb3/52NDc3x+mnnx5LliyJ+vr6Yo8FH9m0adNi7dq18dZbb8XHP/7xOPfcc+OOO+6I448/vtijwUF77rnn4vzzz//A9hkzZsSDDz4YXV1dccstt8T9998fra2tce6558Z9990Xn/70p4swLRy8/X22ly1bFlOmTImf/OQn0draGrW1tXHhhRfGN7/5zQ/8YCEcacrKyva5/YEHHojLL788IiJ2794dX//61+PRRx+Njo6OmDRpUtx3330uU6fPEeMAAACQzN+MAwAAQDIxDgAAAMnEOAAAACQT4wAAAJBMjAMAAEAyMQ4AAADJxDgAAAAkE+MAAACQTIwDAABAMjEOAAAAycQ4AAAAJPt/l2/jV/EJhn0AAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 1000x1000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.displot([tortuosidad, tortuosidad_generados], multiple=\"dodge\", height=10, aspect=1, legend=False, palette = enmax_palette, bins = 50)\n",
    "ax = plt.gca()\n",
    "#ax.legend(labels=[\"Synthetic Samples\", \"Real samples\"], fontsize = 25)\n",
    "#ax.set_title(\"Tortuosity P{} {}\".format(p, dataset_name), fontsize=35)\n",
    "#ax.set_xlabel(\"Tortuosity\", fontsize = 25)\n",
    "ax.set_ylabel(\" \", fontsize = 15)\n",
    "plt.xticks(fontsize=10)\n",
    "plt.yticks(fontsize=10)\n",
    "\n",
    "\n",
    "# get the patches for each dataset\n",
    "patches = ax.patches\n",
    "l = int(len(patches)/2)\n",
    "patches_tortuosidad = patches[:l]\n",
    "patches_tortuosidad_generados = patches[l:]\n",
    "\n",
    "# get the counts for ratio_radios\n",
    "counts_tortuosidad = [patch.get_height() for patch in patches_tortuosidad]\n",
    "\n",
    "# get the counts for ratio_radios_generados\n",
    "counts_tortuosidad_generados = [patch.get_height() for patch in patches_tortuosidad_generados]\n",
    "#ax.xaxis.set_major_locator(MaxNLocator(nbins=1, prune='both'))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 830,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8135569708258236"
      ]
     },
     "execution_count": 830,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.dot(counts_tortuosidad, counts_tortuosidad_generados)/(np.linalg.norm(counts_tortuosidad)*np.linalg.norm(counts_tortuosidad_generados))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Largo total"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 831,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.displot([largo, largo_generados], multiple=\"dodge\", height=10, aspect=1, legend=False, palette=enmax_palette)\n",
    "ax = plt.gca()\n",
    "#ax.legend(labels=[\"Synthetic Samples\", \"Real samples\" ], fontsize = 25)\n",
    "#ax.set_title(\"Length P{} {}\".format(p, dataset_name), fontsize=35)\n",
    "\n",
    "#ax.set_xlabel(\"Length\", fontsize = 25)\n",
    "ax.set_ylabel(\"  \", fontsize = 25)\n",
    "plt.xticks(fontsize=15)\n",
    "plt.yticks(fontsize=15)\n",
    "# get the patches for each dataset\n",
    "patches = ax.patches\n",
    "l = int(len(patches)/2)\n",
    "patches_length = patches[:l]\n",
    "patches_length_generados = patches[l:]\n",
    "\n",
    "\n",
    "\n",
    "counts = [patch.get_height() for patch in patches]\n",
    "# get the counts for ratio_radios\n",
    "counts_length = [patch.get_height() for patch in patches_length]\n",
    "\n",
    "# get the counts for ratio_radios_generados\n",
    "counts_length_generados = [patch.get_height() for patch in patches_length_generados]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 832,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8658592985269553"
      ]
     },
     "execution_count": 832,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.dot(counts_length, counts_length_generados)/(np.linalg.norm(counts_length)*np.linalg.norm(counts_length_generados))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "pytorch",
   "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": "3.10.13"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}