chp04_TreeCreation.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "b28624ec-80a2-441a-b93c-20f09ad2141e",
   "metadata": {},
   "source": [
    "# Tree Creation\n",
    "The reasons we are going to create a tree-based model  are:\n",
    "* they handle categorical and numerical data\n",
    "* they are able to handle non-linear relationships between features and target variables\n",
    "* they provide a clear and interpretable method for feature importance\n",
    "\n",
    "Just like a tree has a trunk and branches, a tree-based model starts with a big \"trunk\" of data and then branches out into different \"leaves\" based on certain rules.\n",
    "\n",
    "Let's say we want to build a tree-based model to predict whether or not it's going to rain tomorrow. We start with a big trunk of data about the weather, like the temperature, humidity, and wind speed.\n",
    "\n",
    "Then, we look at one of those pieces of data, like the temperature. We ask ourselves, \"If the temperature is below 50 degrees, is it more likely to rain or not?\" If we think it's more likely to rain, we create a branch that says \"yes, it's going to rain\" and if we think it's less likely to rain, we create a branch that says \"no, it's not going to rain\".\n",
    "\n",
    "Then, we look at another piece of data, like the humidity, and create more branches based on whether or not we think it's more or less likely to rain. We keep doing this for all the different pieces of data until we have a big tree with lots of branches and leaves.\n",
    "\n",
    "To make a tree-based model for classification, we start by looking at each column in the data and finding the best way to split it into groups based on the different labels. We create a \"node\" with this information and then repeat the process for each node until we have a \"pure\" node or a parameter tells us to stop. To know when our separation is good, we use a metric called the Gini Coefficient.\n",
    " \n",
    "## 4.1 The Gini Coefficient\n",
    "The Gini coefficient was proposed by Corrado Gini, an Italian statistician, in the year 1912. He originally created it as a measure of inequality of income distribution, but it has since been applied to other areas, including economics, sociology, and machine learning.\n",
    "\n",
    "Gini was interested in understanding the distribution of wealth in society and wanted a way to quantify the degree of inequality. He developed the Gini coefficient as a way to measure this inequality on a scale of 0 to 1, where 0 represents perfect equality (everyone has the same wealth) and 1 represents perfect inequality (one person has all the wealth).\n",
    "\n",
    "In machine learning, the Gini coefficient can help us understand the quality of the splits (or a measure of the \"purity\" of a node in the tree) in a tree-based model and identify the nodes that are most informative for making predictions.\n",
    "When building the tree, the model splits the data into different groups based on the values of the input features, and it tries to create groups that are as \"pure\" as possible - that is, groups where all the samples belong to the same class or category. The score of 0 indicates that all the samples in the group belong to the same class (perfect purity) and a score of 1 indicates that the samples are evenly distributed among all the classes (maximum impurity)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "d8aa7c70-65d7-404b-83a2-84095325c84c",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy.random as rn\n",
    "from EffectiveXGBoost import *"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dad6205b-71cf-4f38-847f-5240244c92a9",
   "metadata": {},
   "source": [
    "```python\n",
    "pos_center = 12\n",
    "pos_count = 100\n",
    "neg_center = 7\n",
    "neg_count = 1000\n",
    "```\n",
    "These four lines define four variables: `pos_center`, `pos_count`, `neg_center`, and `neg_count`. `pos_center` and `pos_count` represent the mean and number of data points for a hypothetical positive dataset, and `neg_center` and `neg_count` represent the mean and number of data points for a hypothetical negative dataset.\n",
    "\n",
    "```python\n",
    "rs = rn.RandomState(rn.MT19937(rn.SeedSequence(42)))\n",
    "```\n",
    "This line creates a random number generator object, `rs`, using the `RandomState` function from the `numpy.random` library. The `MT19937` function generates random numbers using the Mersenne Twister algorithm, and `SeedSequence(42)` sets the initial seed for the random number generator.\n",
    "\n",
    "```python\n",
    "gini = pd.DataFrame({'value':\n",
    "np.append((pos_center) + rs.randn(pos_count),\n",
    "(neg_center) + rs.randn(neg_count)),\n",
    "'label':\n",
    "['pos']* pos_count + ['neg'] * neg_count})\n",
    "\n",
    "```\n",
    "This line creates a Pandas DataFrame called `gini`. The DataFrame has two columns: `value` and `label`. The `value` column is created by appending two arrays of random numbers to each other using `np.append()`. The first array is generated by adding `pos_center` to an array of `pos_count` random numbers generated using `rs.randn()`. The second array is generated by adding `neg_center` to an array of `neg_count` random numbers generated using `rs.randn()`. The label column is created by concatenating `pos_count` copies of the string `'pos'` with `neg_count` copies of the string `'neg'`.\n",
    "\n",
    "```python\n",
    "fig, ax = plt.subplots(figsize=(8, 4))\n",
    "```\n",
    "This line creates a new figure and axes for a plot using the `subplots()` function from the `matplotlib.pyplot` library. The `figsize` argument sets the size of the figure.\n",
    "\n",
    "```python\n",
    "_ = (gini\n",
    ".groupby('label')\n",
    "[['value']]\n",
    ".plot.hist(bins=30, alpha=.5, ax=ax, edgecolor='black')\n",
    ")\n",
    "```\n",
    "This line creates a histogram of the data in the `gini` DataFrame using the `groupby()` and `plot.hist()` functions from Pandas. The `groupby()` function groups the data by the label column, and the `[['value']]` argument selects only the `value` column for plotting. The `bins=30` argument sets the number of bins in the histogram to $30$, and the `alpha=.5` argument sets the transparency of the bars to $0.5$. The `ax=ax` argument tells the function to use the axes object created earlier to draw the plot, and the `edgecolor='black'` argument sets the color of the bar edges to black. The `_ =` syntax is used to suppress the output of the function call.\n",
    "\n",
    "```python\n",
    "ax.legend(['Negative', 'Positive'])\n",
    "```\n",
    "This line adds a legend to the plot using the `legend()` function from `matplotlib.axes`. The `['Negative', 'Positive']` argument provides the labels for the legend. The `ax` object is used to specify which axes the legend should be added to."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "e8d55750-7c98-41c5-8fa1-176ebf5261b0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pos_center = 12\n",
    "pos_count = 100\n",
    "neg_center = 7\n",
    "neg_count = 1000\n",
    "rs = rn.RandomState(rn.MT19937(rn.SeedSequence(42)))\n",
    "gini = pd.DataFrame({'value':\n",
    "np.append((pos_center) + rs.randn(pos_count),\n",
    "(neg_center) + rs.randn(neg_count)),\n",
    "'label':\n",
    "['pos']* pos_count + ['neg'] * neg_count})\n",
    "fig, ax = plt.subplots(figsize=(8, 4))\n",
    "_ = (gini\n",
    ".groupby('label')\n",
    "[['value']]\n",
    ".plot.hist(bins=30, alpha=.5, ax=ax, edgecolor='black')\n",
    ")\n",
    "_= ax.legend(['Negative', 'Positive'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d2b2f3f4-b52d-443b-9787-4e8525079e8f",
   "metadata": {},
   "source": [
    "In the graph displayed above, there is an overlap between 9 and 10 of the negative and positive. How do we find an algorithm that determines whether values are negative or positive? The Gini coefficient could be used to help us with that task. Below is a function to help us with that."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "994e1d9d-96a3-48c0-a0a1-4adc9140f6f2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "gini_pos=0.217 gini_neg=0.00202 weighted_avg=0.0241\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.024117224644432264"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def calc_gini(df, val_col, label_col, pos_val, split_point, debug=False):\n",
    "    \"\"\"\n",
    "    This function calculates the Gini impurity of a dataset. Gini impurity\n",
    "    is a measure of the probability of a random sample being classified\n",
    "    incorrectly when a feature is used to split the data. The lower the\n",
    "    impurity, the better the split.\n",
    "    Parameters:\n",
    "    df (pd.DataFrame): The dataframe containing the data\n",
    "    val_col (str): The column name of the feature used to split the data\n",
    "    label_col (str): The column name of the target variable\n",
    "    pos_val (str or int): The value of the target variable that represents\n",
    "    the positive class\n",
    "    split_point (float): The threshold used to split the data.\n",
    "    debug (bool): optional, when set to True, prints the calculated Gini\n",
    "    impurities and the final weighted average\n",
    "    Returns:\n",
    "    float: The weighted average of Gini impurity for the positive and\n",
    "    negative subsets.\n",
    "    \"\"\"\n",
    "    \n",
    "    ge_split = df[val_col] >= split_point\n",
    "    eq_pos = df[label_col] == pos_val\n",
    "    tp = df[ge_split & eq_pos].shape[0]\n",
    "    fp = df[ge_split & ~eq_pos].shape[0]\n",
    "    tn = df[~ge_split & ~eq_pos].shape[0]\n",
    "    fn = df[~ge_split & eq_pos].shape[0]\n",
    "    pos_size = tp+fp\n",
    "    neg_size = tn+fn\n",
    "    total_size = len(df)\n",
    "    if pos_size == 0:\n",
    "        gini_pos = 0\n",
    "    else:\n",
    "        gini_pos = 1 - (tp/pos_size)**2 - (fp/pos_size)**2\n",
    "    if neg_size == 0:\n",
    "        gini_neg = 0\n",
    "    else:\n",
    "        gini_neg = 1 - (tn/neg_size)**2 - (fn/neg_size)**2\n",
    "    weighted_avg = gini_pos * (pos_size/total_size) + gini_neg * (neg_size/total_size)\n",
    "\n",
    "    if debug:\n",
    "        print(f'{gini_pos=:.3} {gini_neg=:.3} {weighted_avg=:.3}')\n",
    "    return weighted_avg\n",
    "\n",
    "calc_gini(gini, val_col='value', label_col='label', pos_val='pos',split_point=9.24, debug=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "a47af193-81e4-4261-a6f0-75d518f274d2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'Split Point')"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "values = np.arange(5, 15, .1)\n",
    "ginis = []\n",
    "for v in values:\n",
    "    ginis.append(calc_gini(gini, \n",
    "                           val_col='value', label_col='label', \n",
    "                           pos_val='pos', split_point=v))\n",
    "fig, ax = plt.subplots(figsize=(8, 4))\n",
    "ax.plot(values, ginis)\n",
    "ax.set_title('Gini Coefficient')\n",
    "ax.set_ylabel('Gini Coefficient')\n",
    "ax.set_xlabel('Split Point')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4e8e2eea-e434-43e3-890f-6380d9f6d979",
   "metadata": {},
   "source": [
    "The lowest Gini coeeficient lies around 10. We want to check those values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "8a3615e9-0798-43c5-a56f-2ba14dec453e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9.6     0.013703\n",
       "9.7     0.010470\n",
       "9.8     0.007193\n",
       "9.9     0.005429\n",
       "10.0    0.007238\n",
       "10.1    0.005438\n",
       "10.2    0.005438\n",
       "10.3    0.007244\n",
       "10.4    0.009046\n",
       "10.5    0.009046\n",
       "dtype: float64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.Series(ginis, index=values).loc[9.5:10.5]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "da74e3a6-9b47-455e-a367-c5dfcae6de36",
   "metadata": {},
   "source": [
    "From the above we can see that the split point of 9.9 will maximize this coefficient"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "eedca3ae-51c6-4415-8c6f-8c109f5e9c3e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "        gini  split\n",
      "49  0.005429    9.9\n"
     ]
    }
   ],
   "source": [
    "print(pd.DataFrame({'gini':ginis, 'split':values})\n",
    "      .query('gini <= gini.min()')\n",
    "     )"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e7e1c6ca-cf75-431f-8011-655720a3d0ea",
   "metadata": {},
   "source": [
    "## 4.2 Coefficients in Trees\n",
    "\n",
    "We want to create a decision stump (or a decision tree without branches)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "01454d23-70e0-4d2f-a49f-d7acfc0b7320",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>DecisionTreeClassifier(max_depth=1)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" checked><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">DecisionTreeClassifier</label><div class=\"sk-toggleable__content\"><pre>DecisionTreeClassifier(max_depth=1)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "DecisionTreeClassifier(max_depth=1)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn import tree\n",
    "stump = tree.DecisionTreeClassifier(max_depth=1)\n",
    "stump.fit(gini[['value']], gini.label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "ee309e86-0068-4699-9a63-4da373c35bbd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(0.5, 0.75, 'value <= 9.708\\ngini = 0.165\\nsamples = 1100\\nvalue = [1000, 100]\\nclass = neg'),\n",
       " Text(0.25, 0.25, 'gini = 0.002\\nsamples = 999\\nvalue = [998, 1]\\nclass = neg'),\n",
       " Text(0.75, 0.25, 'gini = 0.039\\nsamples = 101\\nvalue = [2, 99]\\nclass = pos')]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(8, 4))\n",
    "tree.plot_tree(stump, feature_names=['value']\n",
    "               ,filled=True,class_names=stump.classes_\n",
    "               ,ax=ax)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "2240e17f-0fd5-4528-89e9-23b1c2719d2a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.005397272727272727\n"
     ]
    }
   ],
   "source": [
    "gini_pos = 0.039\n",
    "gini_neg = 0.002\n",
    "pos_size = 101\n",
    "neg_size = 999\n",
    "total_size = pos_size + neg_size\n",
    "weighted_avg = gini_pos * (pos_size/total_size) + gini_neg * (neg_size/total_size)\n",
    "print(weighted_avg)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "18c72bc3-a0a0-4da0-a845-6334aff64271",
   "metadata": {},
   "source": [
    "XGBoost goes through a tree-building process. After training a tree, there are two statistics _gi_ and _hi_. They are similar to\n",
    "Gini but also represent the behavior of the loss function. The gradient (_gi_) represents the first\n",
    "derivative of the loss function with respect to the predicted value for each instance, while the\n",
    "second derivative (_hi_) represents the curvature of the loss function.\n",
    "\n",
    "To create a stump in XGBoost, we limit the _max_depth_ and _n_estimators_ parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "4aaa8a14-c2a6-4dcf-aeb3-7ddb8d831417",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-2 {color: black;background-color: white;}#sk-container-id-2 pre{padding: 0;}#sk-container-id-2 div.sk-toggleable {background-color: white;}#sk-container-id-2 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-2 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-2 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-2 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-2 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-2 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-2 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-2 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-2 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-2 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-2 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-2 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-2 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-2 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-2 div.sk-item {position: relative;z-index: 1;}#sk-container-id-2 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-2 div.sk-item::before, #sk-container-id-2 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-2 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-2 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-2 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-2 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-2 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-2 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-2 div.sk-label-container {text-align: center;}#sk-container-id-2 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-2 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-2\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>XGBClassifier(base_score=0.5, booster=&#x27;gbtree&#x27;, callbacks=None,\n",
       "              colsample_bylevel=1, colsample_bynode=1, colsample_bytree=1,\n",
       "              early_stopping_rounds=None, enable_categorical=False,\n",
       "              eval_metric=None, gamma=0, gpu_id=-1, grow_policy=&#x27;depthwise&#x27;,\n",
       "              importance_type=None, interaction_constraints=&#x27;&#x27;,\n",
       "              learning_rate=0.300000012, max_bin=256, max_cat_to_onehot=4,\n",
       "              max_delta_step=0, max_depth=1, max_leaves=0, min_child_weight=1,\n",
       "              missing=nan, monotone_constraints=&#x27;()&#x27;, n_estimators=1, n_jobs=0,\n",
       "              num_parallel_tree=1, predictor=&#x27;auto&#x27;, random_state=0,\n",
       "              reg_alpha=0, reg_lambda=1, ...)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" checked><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">XGBClassifier</label><div class=\"sk-toggleable__content\"><pre>XGBClassifier(base_score=0.5, booster=&#x27;gbtree&#x27;, callbacks=None,\n",
       "              colsample_bylevel=1, colsample_bynode=1, colsample_bytree=1,\n",
       "              early_stopping_rounds=None, enable_categorical=False,\n",
       "              eval_metric=None, gamma=0, gpu_id=-1, grow_policy=&#x27;depthwise&#x27;,\n",
       "              importance_type=None, interaction_constraints=&#x27;&#x27;,\n",
       "              learning_rate=0.300000012, max_bin=256, max_cat_to_onehot=4,\n",
       "              max_delta_step=0, max_depth=1, max_leaves=0, min_child_weight=1,\n",
       "              missing=nan, monotone_constraints=&#x27;()&#x27;, n_estimators=1, n_jobs=0,\n",
       "              num_parallel_tree=1, predictor=&#x27;auto&#x27;, random_state=0,\n",
       "              reg_alpha=0, reg_lambda=1, ...)</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "XGBClassifier(base_score=0.5, booster='gbtree', callbacks=None,\n",
       "              colsample_bylevel=1, colsample_bynode=1, colsample_bytree=1,\n",
       "              early_stopping_rounds=None, enable_categorical=False,\n",
       "              eval_metric=None, gamma=0, gpu_id=-1, grow_policy='depthwise',\n",
       "              importance_type=None, interaction_constraints='',\n",
       "              learning_rate=0.300000012, max_bin=256, max_cat_to_onehot=4,\n",
       "              max_delta_step=0, max_depth=1, max_leaves=0, min_child_weight=1,\n",
       "              missing=nan, monotone_constraints='()', n_estimators=1, n_jobs=0,\n",
       "              num_parallel_tree=1, predictor='auto', random_state=0,\n",
       "              reg_alpha=0, reg_lambda=1, ...)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import xgboost as xgb\n",
    "xg_stump = xgb.XGBClassifier(n_estimators=1, max_depth=1)\n",
    "xg_stump.fit(gini[['value']], (gini.label== 'pos'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "65b6f6d2-2f33-4474-93cb-72a550d6159c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import os\n",
    "# create a new directory called \"img\" if it does not exist\n",
    "if not os.path.exists('img'):\n",
    "    os.makedirs('img')\n",
    "    \n",
    "fig, ax = plt.subplots(figsize=(12, 12))\n",
    "xgb.plot_tree(xg_stump, num_trees=0, ax=ax)\n",
    "plt.savefig('img/stump_xg.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "4be08846-69ea-4ff8-867f-d9f21163d7b8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "my_image_export(xg_stump, n_trees=0, filename='img/stump_xg.dot', title='A demo stump')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "28d5c4f1-6742-406a-b1bd-97c9bc69d7d4",
   "metadata": {},
   "outputs": [],
   "source": [
    "import subprocess\n",
    "def my_dot_export(xg, num_trees, filename, title='', direction='TB'):\n",
    "    \"\"\"Exports a specified number of trees from an XGBoost model as a graph\n",
    "    visualization in dot and png formats.\n",
    "    Args:\n",
    "    xg: An XGBoost model.\n",
    "    num_trees: The number of tree to export.\n",
    "    filename: The name of the file to save the exported visualization.\n",
    "    title: The title to display on the graph visualization (optional).\n",
    "    direction: The direction to lay out the graph, either 'TB' (top to\n",
    "    bottom) or 'LR' (left to right) (optional).\n",
    "    \"\"\"\n",
    "    res = xgb.to_graphviz(xg, num_trees=num_trees)\n",
    "    content = f''' node [fontname = \"Roboto Condensed\"];\n",
    "    edge [fontname = \"Roboto Thin\"];\n",
    "    label = \"{title}\"\n",
    "    fontname = \"Roboto Condensed\"\n",
    "    '''\n",
    "    out = res.source.replace('graph [ rankdir=TB ]',\n",
    "                             f'graph [ rankdir={direction} ];\\n {content}')\n",
    "    # dot -Gdpi=300 -Tpng -ocourseflow.png courseflow.dot\n",
    "    dot_filename = filename\n",
    "    with open(dot_filename, 'w') as f:\n",
    "        f.write(out)\n",
    "    png_filename = dot_filename.replace('.dot', '.png')\n",
    "    subprocess.run(f'dot -Gdpi=300 -Tpng -o{png_filename} {dot_filename}'.split())\n",
    "    \n",
    "    \n",
    "#my_dot_export(xg_stump, num_trees=0, filename='img/stump_xg.dot', title='A demo stump')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e6de3f65-3505-40fe-9044-615d4fd3146a",
   "metadata": {},
   "source": [
    "## 4.3 Another Visualization Tool\n",
    "This is a Python library for visualizing decision trees. It provides an easy-to-use interface for creating and customizing decision tree visualizations with a range of options for tree layout, node formatting, and color schemes. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "5b292bbf-b8ff-4517-8b55-a6eac3c74a66",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/svg+xml": [
       "<svg xmlns=\"http://www.w3.org/2000/svg\" xmlns:dc=\"http://purl.org/dc/elements/1.1/\" xmlns:ns2=\"http://creativecommons.org/ns#\" xmlns:rdf=\"http://www.w3.org/1999/02/22-rdf-syntax-ns#\" xmlns:xlink=\"http://www.w3.org/1999/xlink\" width=\"250.0\" height=\"182.0\" viewBox=\"0.0 0.0 249.5 181.5\">\n",
       "<g id=\"graph0\" class=\"graph\" transform=\"scale(1.0 1.0) rotate(0) translate(4 177.5)\">\n",
       "<title>G</title>\n",
       "<polygon fill=\"white\" stroke=\"transparent\" points=\"-4,4 -4,-177.5 245.5,-177.5 245.5,4 -4,4\" />\n",
       "<g id=\"clust2\" class=\"cluster\">\n",
       "<title>cluster_legend</title>\n",
       "</g>\n",
       "\n",
       "<g id=\"node1\" class=\"node\">\n",
       "<title>node0</title>\n",
       "<svg width=\"153px\" height=\"47px\" viewBox=\"0 0 204.48625 62.539464\" version=\"1.1\" preserveAspectRatio=\"xMinYMin meet\" x=\"4.5\" y=\"-164.5\">\n",
       " <metadata>\n",
       "  <rdf:RDF>\n",
       "   <ns2:Work>\n",
       "    <dc:type rdf:resource=\"http://purl.org/dc/dcmitype/StillImage\" />\n",
       "    <dc:date>2023-04-23T13:13:46.133580</dc:date>\n",
       "    <dc:format>image/svg+xml</dc:format>\n",
       "    <dc:creator>\n",
       "     <ns2:Agent>\n",
       "      <dc:title>Matplotlib v3.5.1, https://matplotlib.org/</dc:title>\n",
       "     </ns2:Agent>\n",
       "    </dc:creator>\n",
       "   </ns2:Work>\n",
       "  </rdf:RDF>\n",
       " </metadata>\n",
       " <defs>\n",
       "  <style type=\"text/css\">*{stroke-linejoin: round; stroke-linecap: butt}</style>\n",
       " </defs>\n",
       " <g id=\"figure_1\">\n",
       "  <g id=\"patch_1\">\n",
       "   <path d=\"M 0 62.539464  L 204.48625 62.539464  L 204.48625 0  L 0 0  L 0 62.539464  z \" style=\"fill: none\" />\n",
       "  </g>\n",
       "  <g id=\"axes_1\">\n",
       "   <g id=\"patch_2\">\n",
       "    <path d=\"M 20.34625 28.748839  L 204.48625 28.748839  L 204.48625 1.568839  L 20.34625 1.568839  z \" style=\"fill: #ffffff\" />\n",
       "   </g>\n",
       "   <g id=\"patch_3\">\n",
       "    <path d=\"M 36.358424 28.748839  L 44.364511 28.748839  L 44.364511 28.513515  L 36.358424 28.513515  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_4\">\n",
       "    <path d=\"M 44.364511 28.748839  L 52.370598 28.748839  L 52.370598 27.572216  L 44.364511 27.572216  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_5\">\n",
       "    <path d=\"M 52.370598 28.748839  L 60.376685 28.748839  L 60.376685 24.630657  L 52.370598 24.630657  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_6\">\n",
       "    <path d=\"M 60.376685 28.748839  L 68.382772 28.748839  L 68.382772 14.276372  L 60.376685 14.276372  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_7\">\n",
       "    <path d=\"M 68.382772 28.748839  L 76.388859 28.748839  L 76.388859 6.275333  L 68.382772 6.275333  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_8\">\n",
       "    <path d=\"M 76.388859 28.748839  L 84.394946 28.748839  L 84.394946 2.863125  L 76.388859 2.863125  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_9\">\n",
       "    <path d=\"M 84.394946 28.748839  L 92.401033 28.748839  L 92.401033 4.392735  L 84.394946 4.392735  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_10\">\n",
       "    <path d=\"M 92.401033 28.748839  L 100.40712 28.748839  L 100.40712 13.570398  L 92.401033 13.570398  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_11\">\n",
       "    <path d=\"M 100.40712 28.748839  L 108.413207 28.748839  L 108.413207 22.042086  L 100.40712 22.042086  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_12\">\n",
       "    <path d=\"M 108.413207 28.748839  L 116.419293 28.748839  L 116.419293 26.513255  L 108.413207 26.513255  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_13\">\n",
       "    <path d=\"M 116.419293 28.748839  L 124.42538 28.748839  L 124.42538 27.925203  L 116.419293 27.925203  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_14\">\n",
       "    <path d=\"M 124.42538 28.748839  L 132.431467 28.748839  L 132.431467 28.748839  L 124.42538 28.748839  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_15\">\n",
       "    <path d=\"M 132.431467 28.748839  L 140.437554 28.748839  L 140.437554 28.748839  L 132.431467 28.748839  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_16\">\n",
       "    <path d=\"M 140.437554 28.748839  L 148.443641 28.748839  L 148.443641 28.748839  L 140.437554 28.748839  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_17\">\n",
       "    <path d=\"M 148.443641 28.748839  L 156.449728 28.748839  L 156.449728 28.748839  L 148.443641 28.748839  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_18\">\n",
       "    <path d=\"M 156.449728 28.748839  L 164.455815 28.748839  L 164.455815 28.748839  L 156.449728 28.748839  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_19\">\n",
       "    <path d=\"M 164.455815 28.748839  L 172.461902 28.748839  L 172.461902 28.748839  L 164.455815 28.748839  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_20\">\n",
       "    <path d=\"M 172.461902 28.748839  L 180.467989 28.748839  L 180.467989 28.748839  L 172.461902 28.748839  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_21\">\n",
       "    <path d=\"M 180.467989 28.748839  L 188.474076 28.748839  L 188.474076 28.748839  L 180.467989 28.748839  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_22\">\n",
       "    <path d=\"M 36.358424 28.513515  L 44.364511 28.513515  L 44.364511 28.513515  L 36.358424 28.513515  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_23\">\n",
       "    <path d=\"M 44.364511 27.572216  L 52.370598 27.572216  L 52.370598 27.572216  L 44.364511 27.572216  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_24\">\n",
       "    <path d=\"M 52.370598 24.630657  L 60.376685 24.630657  L 60.376685 24.630657  L 52.370598 24.630657  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_25\">\n",
       "    <path d=\"M 60.376685 14.276372  L 68.382772 14.276372  L 68.382772 14.276372  L 60.376685 14.276372  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_26\">\n",
       "    <path d=\"M 68.382772 6.275333  L 76.388859 6.275333  L 76.388859 6.275333  L 68.382772 6.275333  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_27\">\n",
       "    <path d=\"M 76.388859 2.863125  L 84.394946 2.863125  L 84.394946 2.863125  L 76.388859 2.863125  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_28\">\n",
       "    <path d=\"M 84.394946 4.392735  L 92.401033 4.392735  L 92.401033 4.392735  L 84.394946 4.392735  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_29\">\n",
       "    <path d=\"M 92.401033 13.570398  L 100.40712 13.570398  L 100.40712 13.570398  L 92.401033 13.570398  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_30\">\n",
       "    <path d=\"M 100.40712 22.042086  L 108.413207 22.042086  L 108.413207 22.042086  L 100.40712 22.042086  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_31\">\n",
       "    <path d=\"M 108.413207 26.513255  L 116.419293 26.513255  L 116.419293 26.395593  L 108.413207 26.395593  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_32\">\n",
       "    <path d=\"M 116.419293 27.925203  L 124.42538 27.925203  L 124.42538 27.689878  L 116.419293 27.689878  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_33\">\n",
       "    <path d=\"M 124.42538 28.748839  L 132.431467 28.748839  L 132.431467 28.27819  L 124.42538 28.27819  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_34\">\n",
       "    <path d=\"M 132.431467 28.748839  L 140.437554 28.748839  L 140.437554 26.983904  L 132.431467 26.983904  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_35\">\n",
       "    <path d=\"M 140.437554 28.748839  L 148.443641 28.748839  L 148.443641 26.630917  L 140.437554 26.630917  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_36\">\n",
       "    <path d=\"M 148.443641 28.748839  L 156.449728 28.748839  L 156.449728 25.336631  L 148.443641 25.336631  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_37\">\n",
       "    <path d=\"M 156.449728 28.748839  L 164.455815 28.748839  L 164.455815 26.74858  L 156.449728 26.74858  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_38\">\n",
       "    <path d=\"M 164.455815 28.748839  L 172.461902 28.748839  L 172.461902 27.454554  L 164.455815 27.454554  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_39\">\n",
       "    <path d=\"M 172.461902 28.748839  L 180.467989 28.748839  L 180.467989 28.513515  L 172.461902 28.513515  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_40\">\n",
       "    <path d=\"M 180.467989 28.748839  L 188.474076 28.748839  L 188.474076 28.631177  L 180.467989 28.631177  z \" clip-path=\"url(#p55f6a0a312)\" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_41\">\n",
       "    <path d=\"M 117.77232 29.564239  L 114.4578 37.718239  L 121.08684 37.718239  z \" style=\"fill: #444443\" />\n",
       "   </g>\n",
       "   <g id=\"matplotlib.axis_1\">\n",
       "    <g id=\"xtick_1\">\n",
       "     <g id=\"line2d_1\">\n",
       "      <defs>\n",
       "       <path id=\"m2010f7ea5c\" d=\"M 0 0  L 0 3.5  \" style=\"stroke: #000000; stroke-width: 0.3\" />\n",
       "      </defs>\n",
       "      <g>\n",
       "       <use xlink:href=\"#m2010f7ea5c\" x=\"36.358424\" y=\"28.748839\" style=\"stroke: #000000; stroke-width: 0.3\" />\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_1\">\n",
       "      \n",
       "      <g style=\"fill: #444443\" transform=\"translate(28.574049 45.975089)scale(0.08 -0.08)\">\n",
       "       <defs>\n",
       "        <path id=\"ArialMT-33\" d=\"M 269 1209  L 831 1284  Q 928 806 1161 595  Q 1394 384 1728 384  Q 2125 384 2398 659  Q 2672 934 2672 1341  Q 2672 1728 2419 1979  Q 2166 2231 1775 2231  Q 1616 2231 1378 2169  L 1441 2663  Q 1497 2656 1531 2656  Q 1891 2656 2178 2843  Q 2466 3031 2466 3422  Q 2466 3731 2256 3934  Q 2047 4138 1716 4138  Q 1388 4138 1169 3931  Q 950 3725 888 3313  L 325 3413  Q 428 3978 793 4289  Q 1159 4600 1703 4600  Q 2078 4600 2393 4439  Q 2709 4278 2876 4000  Q 3044 3722 3044 3409  Q 3044 3113 2884 2869  Q 2725 2625 2413 2481  Q 2819 2388 3044 2092  Q 3269 1797 3269 1353  Q 3269 753 2831 336  Q 2394 -81 1725 -81  Q 1122 -81 723 278  Q 325 638 269 1209  z \" transform=\"scale(0.015625)\" />\n",
       "        <path id=\"ArialMT-2e\" d=\"M 581 0  L 581 641  L 1222 641  L 1222 0  L 581 0  z \" transform=\"scale(0.015625)\" />\n",
       "        <path id=\"ArialMT-36\" d=\"M 3184 3459  L 2625 3416  Q 2550 3747 2413 3897  Q 2184 4138 1850 4138  Q 1581 4138 1378 3988  Q 1113 3794 959 3422  Q 806 3050 800 2363  Q 1003 2672 1297 2822  Q 1591 2972 1913 2972  Q 2475 2972 2870 2558  Q 3266 2144 3266 1488  Q 3266 1056 3080 686  Q 2894 316 2569 119  Q 2244 -78 1831 -78  Q 1128 -78 684 439  Q 241 956 241 2144  Q 241 3472 731 4075  Q 1159 4600 1884 4600  Q 2425 4600 2770 4297  Q 3116 3994 3184 3459  z M 888 1484  Q 888 1194 1011 928  Q 1134 663 1356 523  Q 1578 384 1822 384  Q 2178 384 2434 671  Q 2691 959 2691 1453  Q 2691 1928 2437 2201  Q 2184 2475 1800 2475  Q 1419 2475 1153 2201  Q 888 1928 888 1484  z \" transform=\"scale(0.015625)\" />\n",
       "        <path id=\"ArialMT-34\" d=\"M 2069 0  L 2069 1097  L 81 1097  L 81 1613  L 2172 4581  L 2631 4581  L 2631 1613  L 3250 1613  L 3250 1097  L 2631 1097  L 2631 0  L 2069 0  z M 2069 1613  L 2069 3678  L 634 1613  L 2069 1613  z \" transform=\"scale(0.015625)\" />\n",
       "       </defs>\n",
       "       <use xlink:href=\"#ArialMT-33\" />\n",
       "       <use xlink:href=\"#ArialMT-2e\" x=\"55.615234\" />\n",
       "       <use xlink:href=\"#ArialMT-36\" x=\"83.398438\" />\n",
       "       <use xlink:href=\"#ArialMT-34\" x=\"139.013672\" />\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_2\">\n",
       "     <g id=\"line2d_2\">\n",
       "      <g>\n",
       "       <use xlink:href=\"#m2010f7ea5c\" x=\"117.77232\" y=\"28.748839\" style=\"stroke: #000000; stroke-width: 0.3\" />\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_2\">\n",
       "      \n",
       "      <g style=\"fill: #444443\" transform=\"translate(109.987945 45.975089)scale(0.08 -0.08)\">\n",
       "       <defs>\n",
       "        <path id=\"ArialMT-39\" d=\"M 350 1059  L 891 1109  Q 959 728 1153 556  Q 1347 384 1650 384  Q 1909 384 2104 503  Q 2300 622 2425 820  Q 2550 1019 2634 1356  Q 2719 1694 2719 2044  Q 2719 2081 2716 2156  Q 2547 1888 2255 1720  Q 1963 1553 1622 1553  Q 1053 1553 659 1965  Q 266 2378 266 3053  Q 266 3750 677 4175  Q 1088 4600 1706 4600  Q 2153 4600 2523 4359  Q 2894 4119 3086 3673  Q 3278 3228 3278 2384  Q 3278 1506 3087 986  Q 2897 466 2520 194  Q 2144 -78 1638 -78  Q 1100 -78 759 220  Q 419 519 350 1059  z M 2653 3081  Q 2653 3566 2395 3850  Q 2138 4134 1775 4134  Q 1400 4134 1122 3828  Q 844 3522 844 3034  Q 844 2597 1108 2323  Q 1372 2050 1759 2050  Q 2150 2050 2401 2323  Q 2653 2597 2653 3081  z \" transform=\"scale(0.015625)\" />\n",
       "        <path id=\"ArialMT-37\" d=\"M 303 3981  L 303 4522  L 3269 4522  L 3269 4084  Q 2831 3619 2401 2847  Q 1972 2075 1738 1259  Q 1569 684 1522 0  L 944 0  Q 953 541 1156 1306  Q 1359 2072 1739 2783  Q 2119 3494 2547 3981  L 303 3981  z \" transform=\"scale(0.015625)\" />\n",
       "        <path id=\"ArialMT-31\" d=\"M 2384 0  L 1822 0  L 1822 3584  Q 1619 3391 1289 3197  Q 959 3003 697 2906  L 697 3450  Q 1169 3672 1522 3987  Q 1875 4303 2022 4600  L 2384 4600  L 2384 0  z \" transform=\"scale(0.015625)\" />\n",
       "       </defs>\n",
       "       <use xlink:href=\"#ArialMT-39\" />\n",
       "       <use xlink:href=\"#ArialMT-2e\" x=\"55.615234\" />\n",
       "       <use xlink:href=\"#ArialMT-37\" x=\"83.398438\" />\n",
       "       <use xlink:href=\"#ArialMT-31\" x=\"139.013672\" />\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"xtick_3\">\n",
       "     <g id=\"line2d_3\">\n",
       "      <g>\n",
       "       <use xlink:href=\"#m2010f7ea5c\" x=\"188.474076\" y=\"28.748839\" style=\"stroke: #000000; stroke-width: 0.3\" />\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_3\">\n",
       "      \n",
       "      <g style=\"fill: #444443\" transform=\"translate(178.465326 45.975089)scale(0.08 -0.08)\">\n",
       "       <defs>\n",
       "        <path id=\"ArialMT-38\" d=\"M 1131 2484  Q 781 2613 612 2850  Q 444 3088 444 3419  Q 444 3919 803 4259  Q 1163 4600 1759 4600  Q 2359 4600 2725 4251  Q 3091 3903 3091 3403  Q 3091 3084 2923 2848  Q 2756 2613 2416 2484  Q 2838 2347 3058 2040  Q 3278 1734 3278 1309  Q 3278 722 2862 322  Q 2447 -78 1769 -78  Q 1091 -78 675 323  Q 259 725 259 1325  Q 259 1772 486 2073  Q 713 2375 1131 2484  z M 1019 3438  Q 1019 3113 1228 2906  Q 1438 2700 1772 2700  Q 2097 2700 2305 2904  Q 2513 3109 2513 3406  Q 2513 3716 2298 3927  Q 2084 4138 1766 4138  Q 1444 4138 1231 3931  Q 1019 3725 1019 3438  z M 838 1322  Q 838 1081 952 856  Q 1066 631 1291 507  Q 1516 384 1775 384  Q 2178 384 2440 643  Q 2703 903 2703 1303  Q 2703 1709 2433 1975  Q 2163 2241 1756 2241  Q 1359 2241 1098 1978  Q 838 1716 838 1322  z \" transform=\"scale(0.015625)\" />\n",
       "       </defs>\n",
       "       <use xlink:href=\"#ArialMT-31\" />\n",
       "       <use xlink:href=\"#ArialMT-34\" x=\"55.615234\" />\n",
       "       <use xlink:href=\"#ArialMT-2e\" x=\"111.230469\" />\n",
       "       <use xlink:href=\"#ArialMT-39\" x=\"139.013672\" />\n",
       "       <use xlink:href=\"#ArialMT-38\" x=\"194.628906\" />\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"text_4\">\n",
       "     \n",
       "     <g style=\"fill: #444443\" transform=\"translate(98.073438 60.154464)scale(0.12 -0.12)\">\n",
       "      <defs>\n",
       "       <path id=\"ArialMT-76\" d=\"M 1344 0  L 81 3319  L 675 3319  L 1388 1331  Q 1503 1009 1600 663  Q 1675 925 1809 1294  L 2547 3319  L 3125 3319  L 1869 0  L 1344 0  z \" transform=\"scale(0.015625)\" />\n",
       "       <path id=\"ArialMT-61\" d=\"M 2588 409  Q 2275 144 1986 34  Q 1697 -75 1366 -75  Q 819 -75 525 192  Q 231 459 231 875  Q 231 1119 342 1320  Q 453 1522 633 1644  Q 813 1766 1038 1828  Q 1203 1872 1538 1913  Q 2219 1994 2541 2106  Q 2544 2222 2544 2253  Q 2544 2597 2384 2738  Q 2169 2928 1744 2928  Q 1347 2928 1158 2789  Q 969 2650 878 2297  L 328 2372  Q 403 2725 575 2942  Q 747 3159 1072 3276  Q 1397 3394 1825 3394  Q 2250 3394 2515 3294  Q 2781 3194 2906 3042  Q 3031 2891 3081 2659  Q 3109 2516 3109 2141  L 3109 1391  Q 3109 606 3145 398  Q 3181 191 3288 0  L 2700 0  Q 2613 175 2588 409  z M 2541 1666  Q 2234 1541 1622 1453  Q 1275 1403 1131 1340  Q 988 1278 909 1158  Q 831 1038 831 891  Q 831 666 1001 516  Q 1172 366 1500 366  Q 1825 366 2078 508  Q 2331 650 2450 897  Q 2541 1088 2541 1459  L 2541 1666  z \" transform=\"scale(0.015625)\" />\n",
       "       <path id=\"ArialMT-6c\" d=\"M 409 0  L 409 4581  L 972 4581  L 972 0  L 409 0  z \" transform=\"scale(0.015625)\" />\n",
       "       <path id=\"ArialMT-75\" d=\"M 2597 0  L 2597 488  Q 2209 -75 1544 -75  Q 1250 -75 995 37  Q 741 150 617 320  Q 494 491 444 738  Q 409 903 409 1263  L 409 3319  L 972 3319  L 972 1478  Q 972 1038 1006 884  Q 1059 663 1231 536  Q 1403 409 1656 409  Q 1909 409 2131 539  Q 2353 669 2445 892  Q 2538 1116 2538 1541  L 2538 3319  L 3100 3319  L 3100 0  L 2597 0  z \" transform=\"scale(0.015625)\" />\n",
       "       <path id=\"ArialMT-65\" d=\"M 2694 1069  L 3275 997  Q 3138 488 2766 206  Q 2394 -75 1816 -75  Q 1088 -75 661 373  Q 234 822 234 1631  Q 234 2469 665 2931  Q 1097 3394 1784 3394  Q 2450 3394 2872 2941  Q 3294 2488 3294 1666  Q 3294 1616 3291 1516  L 816 1516  Q 847 969 1125 678  Q 1403 388 1819 388  Q 2128 388 2347 550  Q 2566 713 2694 1069  z M 847 1978  L 2700 1978  Q 2663 2397 2488 2606  Q 2219 2931 1791 2931  Q 1403 2931 1139 2672  Q 875 2413 847 1978  z \" transform=\"scale(0.015625)\" />\n",
       "      </defs>\n",
       "      <use xlink:href=\"#ArialMT-76\" />\n",
       "      <use xlink:href=\"#ArialMT-61\" x=\"50\" />\n",
       "      <use xlink:href=\"#ArialMT-6c\" x=\"105.615234\" />\n",
       "      <use xlink:href=\"#ArialMT-75\" x=\"127.832031\" />\n",
       "      <use xlink:href=\"#ArialMT-65\" x=\"183.447266\" />\n",
       "     </g>\n",
       "    </g>\n",
       "   </g>\n",
       "   <g id=\"matplotlib.axis_2\">\n",
       "    <g id=\"ytick_1\">\n",
       "     <g id=\"line2d_4\">\n",
       "      <defs>\n",
       "       <path id=\"mbb2a0fec0c\" d=\"M 0 0  L -3.5 0  \" style=\"stroke: #000000; stroke-width: 0.3\" />\n",
       "      </defs>\n",
       "      <g>\n",
       "       <use xlink:href=\"#mbb2a0fec0c\" x=\"20.34625\" y=\"28.748839\" style=\"stroke: #000000; stroke-width: 0.3\" />\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_5\">\n",
       "      \n",
       "      <g style=\"fill: #444443\" transform=\"translate(8.8975 31.611964)scale(0.08 -0.08)\">\n",
       "       <defs>\n",
       "        <path id=\"ArialMT-30\" d=\"M 266 2259  Q 266 3072 433 3567  Q 600 4063 929 4331  Q 1259 4600 1759 4600  Q 2128 4600 2406 4451  Q 2684 4303 2865 4023  Q 3047 3744 3150 3342  Q 3253 2941 3253 2259  Q 3253 1453 3087 958  Q 2922 463 2592 192  Q 2263 -78 1759 -78  Q 1097 -78 719 397  Q 266 969 266 2259  z M 844 2259  Q 844 1131 1108 757  Q 1372 384 1759 384  Q 2147 384 2411 759  Q 2675 1134 2675 2259  Q 2675 3391 2411 3762  Q 2147 4134 1753 4134  Q 1366 4134 1134 3806  Q 844 3388 844 2259  z \" transform=\"scale(0.015625)\" />\n",
       "       </defs>\n",
       "       <use xlink:href=\"#ArialMT-30\" />\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"ytick_2\">\n",
       "     <g id=\"line2d_5\">\n",
       "      <g>\n",
       "       <use xlink:href=\"#mbb2a0fec0c\" x=\"20.34625\" y=\"2.863125\" style=\"stroke: #000000; stroke-width: 0.3\" />\n",
       "      </g>\n",
       "     </g>\n",
       "     <g id=\"text_6\">\n",
       "      \n",
       "      <g style=\"fill: #444443\" transform=\"translate(-0 5.72625)scale(0.08 -0.08)\">\n",
       "       <defs>\n",
       "        <path id=\"ArialMT-32\" d=\"M 3222 541  L 3222 0  L 194 0  Q 188 203 259 391  Q 375 700 629 1000  Q 884 1300 1366 1694  Q 2113 2306 2375 2664  Q 2638 3022 2638 3341  Q 2638 3675 2398 3904  Q 2159 4134 1775 4134  Q 1369 4134 1125 3890  Q 881 3647 878 3216  L 300 3275  Q 359 3922 746 4261  Q 1134 4600 1788 4600  Q 2447 4600 2831 4234  Q 3216 3869 3216 3328  Q 3216 3053 3103 2787  Q 2991 2522 2730 2228  Q 2469 1934 1863 1422  Q 1356 997 1212 845  Q 1069 694 975 541  L 3222 541  z \" transform=\"scale(0.015625)\" />\n",
       "       </defs>\n",
       "       <use xlink:href=\"#ArialMT-32\" />\n",
       "       <use xlink:href=\"#ArialMT-32\" x=\"55.615234\" />\n",
       "       <use xlink:href=\"#ArialMT-30\" x=\"111.230469\" />\n",
       "      </g>\n",
       "     </g>\n",
       "    </g>\n",
       "   </g>\n",
       "   <g id=\"patch_42\">\n",
       "    <path d=\"M 20.34625 28.748839  L 20.34625 1.568839  \" style=\"fill: none; stroke: #444443; stroke-width: 0.3; stroke-linejoin: miter; stroke-linecap: square\" />\n",
       "   </g>\n",
       "   <g id=\"patch_43\">\n",
       "    <path d=\"M 20.34625 28.748839  L 204.48625 28.748839  \" style=\"fill: none; stroke: #444443; stroke-width: 0.3; stroke-linejoin: miter; stroke-linecap: square\" />\n",
       "   </g>\n",
       "  </g>\n",
       " </g>\n",
       " <defs>\n",
       "  <clipPath id=\"p55f6a0a312\">\n",
       "   <rect x=\"20.34625\" y=\"1.568839\" width=\"184.14\" height=\"27.18\" />\n",
       "  </clipPath>\n",
       " </defs>\n",
       "</svg></g>\n",
       "\n",
       "<g id=\"node2\" class=\"node\">\n",
       "<title>leaf1</title>\n",
       "<polygon fill=\"none\" stroke=\"#444443\" stroke-width=\"0\" points=\"74,-83 7,-83 7,0 74,0 74,-83\" />\n",
       "<svg width=\"59px\" height=\"75px\" viewBox=\"0 0 78.52 99.516237\" version=\"1.1\" preserveAspectRatio=\"xMinYMin meet\" x=\"11.5\" y=\"-78.5\">\n",
       " <metadata>\n",
       "  <rdf:RDF>\n",
       "   <ns2:Work>\n",
       "    <dc:type rdf:resource=\"http://purl.org/dc/dcmitype/StillImage\" />\n",
       "    <dc:date>2023-04-23T13:13:46.256590</dc:date>\n",
       "    <dc:format>image/svg+xml</dc:format>\n",
       "    <dc:creator>\n",
       "     <ns2:Agent>\n",
       "      <dc:title>Matplotlib v3.5.1, https://matplotlib.org/</dc:title>\n",
       "     </ns2:Agent>\n",
       "    </dc:creator>\n",
       "   </ns2:Work>\n",
       "  </rdf:RDF>\n",
       " </metadata>\n",
       " <defs>\n",
       "  <style type=\"text/css\">*{stroke-linejoin: round; stroke-linecap: butt}</style>\n",
       " </defs>\n",
       " <g id=\"figure_1\">\n",
       "  <g id=\"patch_1\">\n",
       "   <path d=\"M 0 99.516237  L 78.52 99.516237  L 78.52 0  L 0 0  L 0 99.516237  z \" style=\"fill: none\" />\n",
       "  </g>\n",
       "  <g id=\"axes_1\">\n",
       "   <g id=\"patch_2\">\n",
       "    <path d=\"M 78.52 39.26  C 78.52 34.109618 77.506509 29.009215 75.537414 24.250108  C 73.568318 19.491 70.681835 15.165554 67.042829 11.520822  C 63.403823 7.876089 59.082921 4.982808 54.326915 3.006232  C 49.57091 1.029656 44.472107 0.008147 39.321731 0.000049  C 34.171355 -0.00805 29.069365 0.99742 24.307168 2.95903  C 19.54497 4.92064 15.21499 7.800319 11.564541 11.433589  C 7.914091 15.06686 5.01402 19.383208 3.029968 24.136099  C 1.045916 28.888991 0.016391 33.986181 0.000194 39.136538  C -0.016002 44.286895 0.981444 49.390459 2.935564 54.155735  C 4.889683 58.921012 7.76255 63.255514 11.390076 66.911672  C 15.017603 70.56783 19.329385 73.474684 24.079151 75.466207  C 28.828917 77.45773 33.924482 78.495268 39.074807 78.519563  C 44.225132 78.543858 49.330259 77.554437 54.098602 75.607813  C 58.866944 73.661188 63.205958 70.795141 66.867816 67.173368  C 70.529673 63.551595 73.443304 59.244388 75.442292 54.49776  C 77.441281 49.751131 78.486831 44.657204 78.519223 39.506923  L 39.26 39.26  L 78.52 39.26  z \" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_3\">\n",
       "    <path d=\"M 78.519223 39.506923  C 78.519482 39.46577 78.519676 39.424616 78.519806 39.383462  C 78.519935 39.342308 78.52 39.301154 78.52 39.26  L 39.26 39.26  L 78.519223 39.506923  z \" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"text_1\">\n",
       "    \n",
       "    <g style=\"fill: #444443\" transform=\"translate(26.622031 88.102831)scale(0.09 -0.09)\">\n",
       "     <defs>\n",
       "      <path id=\"ArialMT-6e\" d=\"M 422 0  L 422 3319  L 928 3319  L 928 2847  Q 1294 3394 1984 3394  Q 2284 3394 2536 3286  Q 2788 3178 2913 3003  Q 3038 2828 3088 2588  Q 3119 2431 3119 2041  L 3119 0  L 2556 0  L 2556 2019  Q 2556 2363 2490 2533  Q 2425 2703 2258 2804  Q 2091 2906 1866 2906  Q 1506 2906 1245 2678  Q 984 2450 984 1813  L 984 0  L 422 0  z \" transform=\"scale(0.015625)\" />\n",
       "      <path id=\"ArialMT-3d\" d=\"M 3381 2694  L 356 2694  L 356 3219  L 3381 3219  L 3381 2694  z M 3381 1303  L 356 1303  L 356 1828  L 3381 1828  L 3381 1303  z \" transform=\"scale(0.015625)\" />\n",
       "      <path id=\"ArialMT-39\" d=\"M 350 1059  L 891 1109  Q 959 728 1153 556  Q 1347 384 1650 384  Q 1909 384 2104 503  Q 2300 622 2425 820  Q 2550 1019 2634 1356  Q 2719 1694 2719 2044  Q 2719 2081 2716 2156  Q 2547 1888 2255 1720  Q 1963 1553 1622 1553  Q 1053 1553 659 1965  Q 266 2378 266 3053  Q 266 3750 677 4175  Q 1088 4600 1706 4600  Q 2153 4600 2523 4359  Q 2894 4119 3086 3673  Q 3278 3228 3278 2384  Q 3278 1506 3087 986  Q 2897 466 2520 194  Q 2144 -78 1638 -78  Q 1100 -78 759 220  Q 419 519 350 1059  z M 2653 3081  Q 2653 3566 2395 3850  Q 2138 4134 1775 4134  Q 1400 4134 1122 3828  Q 844 3522 844 3034  Q 844 2597 1108 2323  Q 1372 2050 1759 2050  Q 2150 2050 2401 2323  Q 2653 2597 2653 3081  z \" transform=\"scale(0.015625)\" />\n",
       "     </defs>\n",
       "     <use xlink:href=\"#ArialMT-6e\" />\n",
       "     <use xlink:href=\"#ArialMT-3d\" x=\"55.615234\" />\n",
       "     <use xlink:href=\"#ArialMT-39\" x=\"114.013672\" />\n",
       "     <use xlink:href=\"#ArialMT-39\" x=\"169.628906\" />\n",
       "     <use xlink:href=\"#ArialMT-39\" x=\"225.244141\" />\n",
       "    </g>\n",
       "    \n",
       "    <g style=\"fill: #444443\" transform=\"translate(22.247891 97.622019)scale(0.09 -0.09)\">\n",
       "     <defs>\n",
       "      <path id=\"ArialMT-65\" d=\"M 2694 1069  L 3275 997  Q 3138 488 2766 206  Q 2394 -75 1816 -75  Q 1088 -75 661 373  Q 234 822 234 1631  Q 234 2469 665 2931  Q 1097 3394 1784 3394  Q 2450 3394 2872 2941  Q 3294 2488 3294 1666  Q 3294 1616 3291 1516  L 816 1516  Q 847 969 1125 678  Q 1403 388 1819 388  Q 2128 388 2347 550  Q 2566 713 2694 1069  z M 847 1978  L 2700 1978  Q 2663 2397 2488 2606  Q 2219 2931 1791 2931  Q 1403 2931 1139 2672  Q 875 2413 847 1978  z \" transform=\"scale(0.015625)\" />\n",
       "      <path id=\"ArialMT-67\" d=\"M 319 -275  L 866 -356  Q 900 -609 1056 -725  Q 1266 -881 1628 -881  Q 2019 -881 2231 -725  Q 2444 -569 2519 -288  Q 2563 -116 2559 434  Q 2191 0 1641 0  Q 956 0 581 494  Q 206 988 206 1678  Q 206 2153 378 2554  Q 550 2956 876 3175  Q 1203 3394 1644 3394  Q 2231 3394 2613 2919  L 2613 3319  L 3131 3319  L 3131 450  Q 3131 -325 2973 -648  Q 2816 -972 2473 -1159  Q 2131 -1347 1631 -1347  Q 1038 -1347 672 -1080  Q 306 -813 319 -275  z M 784 1719  Q 784 1066 1043 766  Q 1303 466 1694 466  Q 2081 466 2343 764  Q 2606 1063 2606 1700  Q 2606 2309 2336 2618  Q 2066 2928 1684 2928  Q 1309 2928 1046 2623  Q 784 2319 784 1719  z \" transform=\"scale(0.015625)\" />\n",
       "      <path id=\"ArialMT-61\" d=\"M 2588 409  Q 2275 144 1986 34  Q 1697 -75 1366 -75  Q 819 -75 525 192  Q 231 459 231 875  Q 231 1119 342 1320  Q 453 1522 633 1644  Q 813 1766 1038 1828  Q 1203 1872 1538 1913  Q 2219 1994 2541 2106  Q 2544 2222 2544 2253  Q 2544 2597 2384 2738  Q 2169 2928 1744 2928  Q 1347 2928 1158 2789  Q 969 2650 878 2297  L 328 2372  Q 403 2725 575 2942  Q 747 3159 1072 3276  Q 1397 3394 1825 3394  Q 2250 3394 2515 3294  Q 2781 3194 2906 3042  Q 3031 2891 3081 2659  Q 3109 2516 3109 2141  L 3109 1391  Q 3109 606 3145 398  Q 3181 191 3288 0  L 2700 0  Q 2613 175 2588 409  z M 2541 1666  Q 2234 1541 1622 1453  Q 1275 1403 1131 1340  Q 988 1278 909 1158  Q 831 1038 831 891  Q 831 666 1001 516  Q 1172 366 1500 366  Q 1825 366 2078 508  Q 2331 650 2450 897  Q 2541 1088 2541 1459  L 2541 1666  z \" transform=\"scale(0.015625)\" />\n",
       "      <path id=\"ArialMT-74\" d=\"M 1650 503  L 1731 6  Q 1494 -44 1306 -44  Q 1000 -44 831 53  Q 663 150 594 308  Q 525 466 525 972  L 525 2881  L 113 2881  L 113 3319  L 525 3319  L 525 4141  L 1084 4478  L 1084 3319  L 1650 3319  L 1650 2881  L 1084 2881  L 1084 941  Q 1084 700 1114 631  Q 1144 563 1211 522  Q 1278 481 1403 481  Q 1497 481 1650 503  z \" transform=\"scale(0.015625)\" />\n",
       "      <path id=\"ArialMT-69\" d=\"M 425 3934  L 425 4581  L 988 4581  L 988 3934  L 425 3934  z M 425 0  L 425 3319  L 988 3319  L 988 0  L 425 0  z \" transform=\"scale(0.015625)\" />\n",
       "      <path id=\"ArialMT-76\" d=\"M 1344 0  L 81 3319  L 675 3319  L 1388 1331  Q 1503 1009 1600 663  Q 1675 925 1809 1294  L 2547 3319  L 3125 3319  L 1869 0  L 1344 0  z \" transform=\"scale(0.015625)\" />\n",
       "     </defs>\n",
       "     <use xlink:href=\"#ArialMT-6e\" />\n",
       "     <use xlink:href=\"#ArialMT-65\" x=\"55.615234\" />\n",
       "     <use xlink:href=\"#ArialMT-67\" x=\"111.230469\" />\n",
       "     <use xlink:href=\"#ArialMT-61\" x=\"166.845703\" />\n",
       "     <use xlink:href=\"#ArialMT-74\" x=\"222.460938\" />\n",
       "     <use xlink:href=\"#ArialMT-69\" x=\"250.244141\" />\n",
       "     <use xlink:href=\"#ArialMT-76\" x=\"272.460938\" />\n",
       "     <use xlink:href=\"#ArialMT-65\" x=\"322.460938\" />\n",
       "    </g>\n",
       "   </g>\n",
       "  </g>\n",
       " </g>\n",
       "</svg></g>\n",
       "\n",
       "<g id=\"edge1\" class=\"edge\">\n",
       "<title>node0-&gt;leaf1</title>\n",
       "<path fill=\"none\" stroke=\"#444443\" stroke-width=\"0.3\" d=\"M69.68,-113.98C66.27,-105.65 62.42,-96.21 58.66,-86.98\" />\n",
       "<polygon fill=\"#444443\" stroke=\"#444443\" stroke-width=\"0.3\" points=\"59.86,-86.22 57.05,-83.05 57.26,-87.28 59.86,-86.22\" />\n",
       "<text text-anchor=\"start\" x=\"65.5\" y=\"-94.8\" font-family=\"Arial\" font-size=\"14.00\" fill=\"#444443\"> &#160;&lt;</text>\n",
       "</g>\n",
       "\n",
       "<g id=\"node3\" class=\"node\">\n",
       "<title>leaf2</title>\n",
       "<polygon fill=\"none\" stroke=\"#444443\" stroke-width=\"0\" points=\"155,-82.5 88,-82.5 88,-0.5 155,-0.5 155,-82.5\" />\n",
       "<svg width=\"59px\" height=\"74px\" viewBox=\"0 0 78.519998 99.410763\" version=\"1.1\" preserveAspectRatio=\"xMinYMin meet\" x=\"92.5\" y=\"-78.5\">\n",
       " <metadata>\n",
       "  <rdf:RDF>\n",
       "   <ns2:Work>\n",
       "    <dc:type rdf:resource=\"http://purl.org/dc/dcmitype/StillImage\" />\n",
       "    <dc:date>2023-04-23T13:13:46.320089</dc:date>\n",
       "    <dc:format>image/svg+xml</dc:format>\n",
       "    <dc:creator>\n",
       "     <ns2:Agent>\n",
       "      <dc:title>Matplotlib v3.5.1, https://matplotlib.org/</dc:title>\n",
       "     </ns2:Agent>\n",
       "    </dc:creator>\n",
       "   </ns2:Work>\n",
       "  </rdf:RDF>\n",
       " </metadata>\n",
       " <defs>\n",
       "  <style type=\"text/css\">*{stroke-linejoin: round; stroke-linecap: butt}</style>\n",
       " </defs>\n",
       " <g id=\"figure_1\">\n",
       "  <g id=\"patch_1\">\n",
       "   <path d=\"M 0 99.410763  L 78.519998 99.410763  L 78.519998 0  L 0 0  L 0 99.410763  z \" style=\"fill: none\" />\n",
       "  </g>\n",
       "  <g id=\"axes_1\">\n",
       "   <g id=\"patch_2\">\n",
       "    <path d=\"M 78.519998 39.259994  C 78.519998 38.44581 78.49467 37.631823 78.444053 36.819214  C 78.393435 36.006605 78.317552 35.195767 78.216513 34.387877  L 39.259998 39.259994  L 78.519998 39.259994  z \" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"patch_3\">\n",
       "    <path d=\"M 78.216513 34.387877  C 77.58946 29.374081 76.000332 24.528165 73.536543 20.116692  C 71.072755 15.705219 67.780326 11.810589 63.840366 8.647038  C 59.900406 5.483486 55.386507 3.110103 50.54697 1.657443  C 45.707432 0.204783 40.632651 -0.300021 35.601779 0.1708  C 30.570907 0.641622 25.677914 2.079274 21.191951 4.404674  C 16.705989 6.730075 12.710849 9.899789 9.426295 13.739457  C 6.141742 17.579125 3.629125 22.017028 2.026658 26.809047  C 0.424192 31.601066 -0.238192 36.657694 0.075942 41.700774  C 0.390076 46.743855 1.674862 51.679193 3.859624 56.235305  C 6.044387 60.791417 9.08832 64.883202 12.823981 68.28558  C 16.559643 71.687958 20.917258 74.337378 25.657123 76.0881  C 30.396987 77.838822 35.430568 78.658146 40.480979 78.501003  C 45.53139 78.34386 50.504297 77.213184 55.126151 75.171173  C 59.748005 73.129161 63.932477 70.213955 67.449387 66.585914  C 70.966298 62.957872 73.749957 58.684762 75.647242 54.001638  C 77.544526 49.318513 78.519998 44.312848 78.519998 39.259992  L 39.259998 39.259994  L 78.216513 34.387877  z \" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "   </g>\n",
       "   <g id=\"text_1\">\n",
       "    \n",
       "    <g style=\"fill: #444443\" transform=\"translate(26.622029 88.102825)scale(0.09 -0.09)\">\n",
       "     <defs>\n",
       "      <path id=\"ArialMT-6e\" d=\"M 422 0  L 422 3319  L 928 3319  L 928 2847  Q 1294 3394 1984 3394  Q 2284 3394 2536 3286  Q 2788 3178 2913 3003  Q 3038 2828 3088 2588  Q 3119 2431 3119 2041  L 3119 0  L 2556 0  L 2556 2019  Q 2556 2363 2490 2533  Q 2425 2703 2258 2804  Q 2091 2906 1866 2906  Q 1506 2906 1245 2678  Q 984 2450 984 1813  L 984 0  L 422 0  z \" transform=\"scale(0.015625)\" />\n",
       "      <path id=\"ArialMT-3d\" d=\"M 3381 2694  L 356 2694  L 356 3219  L 3381 3219  L 3381 2694  z M 3381 1303  L 356 1303  L 356 1828  L 3381 1828  L 3381 1303  z \" transform=\"scale(0.015625)\" />\n",
       "      <path id=\"ArialMT-31\" d=\"M 2384 0  L 1822 0  L 1822 3584  Q 1619 3391 1289 3197  Q 959 3003 697 2906  L 697 3450  Q 1169 3672 1522 3987  Q 1875 4303 2022 4600  L 2384 4600  L 2384 0  z \" transform=\"scale(0.015625)\" />\n",
       "      <path id=\"ArialMT-30\" d=\"M 266 2259  Q 266 3072 433 3567  Q 600 4063 929 4331  Q 1259 4600 1759 4600  Q 2128 4600 2406 4451  Q 2684 4303 2865 4023  Q 3047 3744 3150 3342  Q 3253 2941 3253 2259  Q 3253 1453 3087 958  Q 2922 463 2592 192  Q 2263 -78 1759 -78  Q 1097 -78 719 397  Q 266 969 266 2259  z M 844 2259  Q 844 1131 1108 757  Q 1372 384 1759 384  Q 2147 384 2411 759  Q 2675 1134 2675 2259  Q 2675 3391 2411 3762  Q 2147 4134 1753 4134  Q 1366 4134 1134 3806  Q 844 3388 844 2259  z \" transform=\"scale(0.015625)\" />\n",
       "     </defs>\n",
       "     <use xlink:href=\"#ArialMT-6e\" />\n",
       "     <use xlink:href=\"#ArialMT-3d\" x=\"55.615234\" />\n",
       "     <use xlink:href=\"#ArialMT-31\" x=\"114.013672\" />\n",
       "     <use xlink:href=\"#ArialMT-30\" x=\"169.628906\" />\n",
       "     <use xlink:href=\"#ArialMT-31\" x=\"225.244141\" />\n",
       "    </g>\n",
       "    \n",
       "    <g style=\"fill: #444443\" transform=\"translate(24.002888 97.622013)scale(0.09 -0.09)\">\n",
       "     <defs>\n",
       "      <path id=\"ArialMT-70\" d=\"M 422 -1272  L 422 3319  L 934 3319  L 934 2888  Q 1116 3141 1344 3267  Q 1572 3394 1897 3394  Q 2322 3394 2647 3175  Q 2972 2956 3137 2557  Q 3303 2159 3303 1684  Q 3303 1175 3120 767  Q 2938 359 2589 142  Q 2241 -75 1856 -75  Q 1575 -75 1351 44  Q 1128 163 984 344  L 984 -1272  L 422 -1272  z M 931 1641  Q 931 1000 1190 694  Q 1450 388 1819 388  Q 2194 388 2461 705  Q 2728 1022 2728 1688  Q 2728 2322 2467 2637  Q 2206 2953 1844 2953  Q 1484 2953 1207 2617  Q 931 2281 931 1641  z \" transform=\"scale(0.015625)\" />\n",
       "      <path id=\"ArialMT-6f\" d=\"M 213 1659  Q 213 2581 725 3025  Q 1153 3394 1769 3394  Q 2453 3394 2887 2945  Q 3322 2497 3322 1706  Q 3322 1066 3130 698  Q 2938 331 2570 128  Q 2203 -75 1769 -75  Q 1072 -75 642 372  Q 213 819 213 1659  z M 791 1659  Q 791 1022 1069 705  Q 1347 388 1769 388  Q 2188 388 2466 706  Q 2744 1025 2744 1678  Q 2744 2294 2464 2611  Q 2184 2928 1769 2928  Q 1347 2928 1069 2612  Q 791 2297 791 1659  z \" transform=\"scale(0.015625)\" />\n",
       "      <path id=\"ArialMT-73\" d=\"M 197 991  L 753 1078  Q 800 744 1014 566  Q 1228 388 1613 388  Q 2000 388 2187 545  Q 2375 703 2375 916  Q 2375 1106 2209 1216  Q 2094 1291 1634 1406  Q 1016 1563 777 1677  Q 538 1791 414 1992  Q 291 2194 291 2438  Q 291 2659 392 2848  Q 494 3038 669 3163  Q 800 3259 1026 3326  Q 1253 3394 1513 3394  Q 1903 3394 2198 3281  Q 2494 3169 2634 2976  Q 2775 2784 2828 2463  L 2278 2388  Q 2241 2644 2061 2787  Q 1881 2931 1553 2931  Q 1166 2931 1000 2803  Q 834 2675 834 2503  Q 834 2394 903 2306  Q 972 2216 1119 2156  Q 1203 2125 1616 2013  Q 2213 1853 2448 1751  Q 2684 1650 2818 1456  Q 2953 1263 2953 975  Q 2953 694 2789 445  Q 2625 197 2315 61  Q 2006 -75 1616 -75  Q 969 -75 630 194  Q 291 463 197 991  z \" transform=\"scale(0.015625)\" />\n",
       "      <path id=\"ArialMT-69\" d=\"M 425 3934  L 425 4581  L 988 4581  L 988 3934  L 425 3934  z M 425 0  L 425 3319  L 988 3319  L 988 0  L 425 0  z \" transform=\"scale(0.015625)\" />\n",
       "      <path id=\"ArialMT-74\" d=\"M 1650 503  L 1731 6  Q 1494 -44 1306 -44  Q 1000 -44 831 53  Q 663 150 594 308  Q 525 466 525 972  L 525 2881  L 113 2881  L 113 3319  L 525 3319  L 525 4141  L 1084 4478  L 1084 3319  L 1650 3319  L 1650 2881  L 1084 2881  L 1084 941  Q 1084 700 1114 631  Q 1144 563 1211 522  Q 1278 481 1403 481  Q 1497 481 1650 503  z \" transform=\"scale(0.015625)\" />\n",
       "      <path id=\"ArialMT-76\" d=\"M 1344 0  L 81 3319  L 675 3319  L 1388 1331  Q 1503 1009 1600 663  Q 1675 925 1809 1294  L 2547 3319  L 3125 3319  L 1869 0  L 1344 0  z \" transform=\"scale(0.015625)\" />\n",
       "      <path id=\"ArialMT-65\" d=\"M 2694 1069  L 3275 997  Q 3138 488 2766 206  Q 2394 -75 1816 -75  Q 1088 -75 661 373  Q 234 822 234 1631  Q 234 2469 665 2931  Q 1097 3394 1784 3394  Q 2450 3394 2872 2941  Q 3294 2488 3294 1666  Q 3294 1616 3291 1516  L 816 1516  Q 847 969 1125 678  Q 1403 388 1819 388  Q 2128 388 2347 550  Q 2566 713 2694 1069  z M 847 1978  L 2700 1978  Q 2663 2397 2488 2606  Q 2219 2931 1791 2931  Q 1403 2931 1139 2672  Q 875 2413 847 1978  z \" transform=\"scale(0.015625)\" />\n",
       "     </defs>\n",
       "     <use xlink:href=\"#ArialMT-70\" />\n",
       "     <use xlink:href=\"#ArialMT-6f\" x=\"55.615234\" />\n",
       "     <use xlink:href=\"#ArialMT-73\" x=\"111.230469\" />\n",
       "     <use xlink:href=\"#ArialMT-69\" x=\"161.230469\" />\n",
       "     <use xlink:href=\"#ArialMT-74\" x=\"183.447266\" />\n",
       "     <use xlink:href=\"#ArialMT-69\" x=\"211.230469\" />\n",
       "     <use xlink:href=\"#ArialMT-76\" x=\"233.447266\" />\n",
       "     <use xlink:href=\"#ArialMT-65\" x=\"283.447266\" />\n",
       "    </g>\n",
       "   </g>\n",
       "  </g>\n",
       " </g>\n",
       "</svg></g>\n",
       "\n",
       "<g id=\"edge2\" class=\"edge\">\n",
       "<title>node0-&gt;leaf2</title>\n",
       "<path fill=\"none\" stroke=\"#444443\" stroke-width=\"0.3\" d=\"M91.6,-113.98C95.12,-105.55 99.12,-95.98 103.03,-86.65\" />\n",
       "<polygon fill=\"#444443\" stroke=\"#444443\" stroke-width=\"0.3\" points=\"104.44,-86.9 104.69,-82.67 101.86,-85.82 104.44,-86.9\" />\n",
       "<text text-anchor=\"start\" x=\"101.5\" y=\"-94.8\" font-family=\"Arial\" font-size=\"14.00\" fill=\"#444443\"> &#160;&#8805;</text>\n",
       "</g>\n",
       "\n",
       "\n",
       "<g id=\"node4\" class=\"node\">\n",
       "<title>legend</title>\n",
       "<polygon fill=\"none\" stroke=\"black\" stroke-width=\"0\" points=\"233,-165.5 168,-165.5 168,-117.5 233,-117.5 233,-165.5\" />\n",
       "<svg width=\"61px\" height=\"44px\" viewBox=\"0 0 82.304687 58.539062\" version=\"1.1\" preserveAspectRatio=\"xMinYMin meet\" x=\"170.5\" y=\"-163.5\">\n",
       " <metadata>\n",
       "  <rdf:RDF>\n",
       "   <ns2:Work>\n",
       "    <dc:type rdf:resource=\"http://purl.org/dc/dcmitype/StillImage\" />\n",
       "    <dc:date>2023-04-23T13:13:45.854333</dc:date>\n",
       "    <dc:format>image/svg+xml</dc:format>\n",
       "    <dc:creator>\n",
       "     <ns2:Agent>\n",
       "      <dc:title>Matplotlib v3.5.1, https://matplotlib.org/</dc:title>\n",
       "     </ns2:Agent>\n",
       "    </dc:creator>\n",
       "   </ns2:Work>\n",
       "  </rdf:RDF>\n",
       " </metadata>\n",
       " <defs>\n",
       "  <style type=\"text/css\">*{stroke-linejoin: round; stroke-linecap: butt}</style>\n",
       " </defs>\n",
       " <g id=\"figure_1\">\n",
       "  <g id=\"patch_1\">\n",
       "   <path d=\"M 0 58.539062  L 82.304687 58.539062  L 82.304687 0  L 0 0  L 0 58.539062  z \" style=\"fill: none\" />\n",
       "  </g>\n",
       "  <g id=\"axes_1\">\n",
       "   <g id=\"legend_1\">\n",
       "    <g id=\"patch_2\">\n",
       "     <path d=\"M 2 58.539062  L 75.304688 58.539062  Q 77.304688 58.539062 77.304688 56.539062  L 77.304688 7  Q 77.304688 5 75.304688 5  L 2 5  Q 0 5 0 7  L 0 56.539062  Q 0 58.539062 2 58.539062  z \" style=\"fill: #ffffff; opacity: 0.8; stroke: #444443; stroke-width: 0.5; stroke-linejoin: miter\" />\n",
       "    </g>\n",
       "    <g id=\"text_1\">\n",
       "     \n",
       "     <g style=\"fill: #444443\" transform=\"translate(19.760156 20.157812)scale(0.1 -0.1)\">\n",
       "      <defs>\n",
       "       <path id=\"Arial-BoldMT-70\" d=\"M 434 3319  L 1253 3319  L 1253 2831  Q 1413 3081 1684 3237  Q 1956 3394 2288 3394  Q 2866 3394 3269 2941  Q 3672 2488 3672 1678  Q 3672 847 3265 386  Q 2859 -75 2281 -75  Q 2006 -75 1782 34  Q 1559 144 1313 409  L 1313 -1263  L 434 -1263  L 434 3319  z M 1303 1716  Q 1303 1156 1525 889  Q 1747 622 2066 622  Q 2372 622 2575 867  Q 2778 1113 2778 1672  Q 2778 2194 2568 2447  Q 2359 2700 2050 2700  Q 1728 2700 1515 2451  Q 1303 2203 1303 1716  z \" transform=\"scale(0.015625)\" />\n",
       "       <path id=\"Arial-BoldMT-6f\" d=\"M 256 1706  Q 256 2144 472 2553  Q 688 2963 1083 3178  Q 1478 3394 1966 3394  Q 2719 3394 3200 2905  Q 3681 2416 3681 1669  Q 3681 916 3195 420  Q 2709 -75 1972 -75  Q 1516 -75 1102 131  Q 688 338 472 736  Q 256 1134 256 1706  z M 1156 1659  Q 1156 1166 1390 903  Q 1625 641 1969 641  Q 2313 641 2545 903  Q 2778 1166 2778 1666  Q 2778 2153 2545 2415  Q 2313 2678 1969 2678  Q 1625 2678 1390 2415  Q 1156 2153 1156 1659  z \" transform=\"scale(0.015625)\" />\n",
       "       <path id=\"Arial-BoldMT-73\" d=\"M 150 947  L 1031 1081  Q 1088 825 1259 692  Q 1431 559 1741 559  Q 2081 559 2253 684  Q 2369 772 2369 919  Q 2369 1019 2306 1084  Q 2241 1147 2013 1200  Q 950 1434 666 1628  Q 272 1897 272 2375  Q 272 2806 612 3100  Q 953 3394 1669 3394  Q 2350 3394 2681 3172  Q 3013 2950 3138 2516  L 2309 2363  Q 2256 2556 2107 2659  Q 1959 2763 1684 2763  Q 1338 2763 1188 2666  Q 1088 2597 1088 2488  Q 1088 2394 1175 2328  Q 1294 2241 1995 2081  Q 2697 1922 2975 1691  Q 3250 1456 3250 1038  Q 3250 581 2869 253  Q 2488 -75 1741 -75  Q 1063 -75 667 200  Q 272 475 150 947  z \" transform=\"scale(0.015625)\" />\n",
       "       <path id=\"Arial-BoldMT-69\" d=\"M 459 3769  L 459 4581  L 1338 4581  L 1338 3769  L 459 3769  z M 459 0  L 459 3319  L 1338 3319  L 1338 0  L 459 0  z \" transform=\"scale(0.015625)\" />\n",
       "       <path id=\"Arial-BoldMT-74\" d=\"M 1981 3319  L 1981 2619  L 1381 2619  L 1381 1281  Q 1381 875 1398 808  Q 1416 741 1477 697  Q 1538 653 1625 653  Q 1747 653 1978 738  L 2053 56  Q 1747 -75 1359 -75  Q 1122 -75 931 4  Q 741 84 652 211  Q 563 338 528 553  Q 500 706 500 1172  L 500 2619  L 97 2619  L 97 3319  L 500 3319  L 500 3978  L 1381 4491  L 1381 3319  L 1981 3319  z \" transform=\"scale(0.015625)\" />\n",
       "       <path id=\"Arial-BoldMT-76\" d=\"M 1372 0  L 34 3319  L 956 3319  L 1581 1625  L 1763 1059  Q 1834 1275 1853 1344  Q 1897 1484 1947 1625  L 2578 3319  L 3481 3319  L 2163 0  L 1372 0  z \" transform=\"scale(0.015625)\" />\n",
       "       <path id=\"Arial-BoldMT-65\" d=\"M 2381 1056  L 3256 909  Q 3088 428 2723 176  Q 2359 -75 1813 -75  Q 947 -75 531 491  Q 203 944 203 1634  Q 203 2459 634 2926  Q 1066 3394 1725 3394  Q 2466 3394 2894 2905  Q 3322 2416 3303 1406  L 1103 1406  Q 1113 1016 1316 798  Q 1519 581 1822 581  Q 2028 581 2168 693  Q 2309 806 2381 1056  z M 2431 1944  Q 2422 2325 2234 2523  Q 2047 2722 1778 2722  Q 1491 2722 1303 2513  Q 1116 2303 1119 1944  L 2431 1944  z \" transform=\"scale(0.015625)\" />\n",
       "      </defs>\n",
       "      <use xlink:href=\"#Arial-BoldMT-70\" />\n",
       "      <use xlink:href=\"#Arial-BoldMT-6f\" x=\"61.083984\" />\n",
       "      <use xlink:href=\"#Arial-BoldMT-73\" x=\"122.167969\" />\n",
       "      <use xlink:href=\"#Arial-BoldMT-69\" x=\"177.783203\" />\n",
       "      <use xlink:href=\"#Arial-BoldMT-74\" x=\"205.566406\" />\n",
       "      <use xlink:href=\"#Arial-BoldMT-69\" x=\"238.867188\" />\n",
       "      <use xlink:href=\"#Arial-BoldMT-76\" x=\"266.650391\" />\n",
       "      <use xlink:href=\"#Arial-BoldMT-65\" x=\"322.265625\" />\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"patch_3\">\n",
       "     <path d=\"M 8 34.289062  L 28 34.289062  L 28 27.289062  L 8 27.289062  z \" style=\"fill: #fefebb; stroke: #444443; stroke-width: 0.4; stroke-linejoin: miter\" />\n",
       "    </g>\n",
       "    <g id=\"text_2\">\n",
       "     \n",
       "     <g style=\"fill: #444443\" transform=\"translate(31.5 34.289062)scale(0.1 -0.1)\">\n",
       "      <defs>\n",
       "       <path id=\"ArialMT-6e\" d=\"M 422 0  L 422 3319  L 928 3319  L 928 2847  Q 1294 3394 1984 3394  Q 2284 3394 2536 3286  Q 2788 3178 2913 3003  Q 3038 2828 3088 2588  Q 3119 2431 3119 2041  L 3119 0  L 2556 0  L 2556 2019  Q 2556 2363 2490 2533  Q 2425 2703 2258 2804  Q 2091 2906 1866 2906  Q 1506 2906 1245 2678  Q 984 2450 984 1813  L 984 0  L 422 0  z \" transform=\"scale(0.015625)\" />\n",
       "       <path id=\"ArialMT-65\" d=\"M 2694 1069  L 3275 997  Q 3138 488 2766 206  Q 2394 -75 1816 -75  Q 1088 -75 661 373  Q 234 822 234 1631  Q 234 2469 665 2931  Q 1097 3394 1784 3394  Q 2450 3394 2872 2941  Q 3294 2488 3294 1666  Q 3294 1616 3291 1516  L 816 1516  Q 847 969 1125 678  Q 1403 388 1819 388  Q 2128 388 2347 550  Q 2566 713 2694 1069  z M 847 1978  L 2700 1978  Q 2663 2397 2488 2606  Q 2219 2931 1791 2931  Q 1403 2931 1139 2672  Q 875 2413 847 1978  z \" transform=\"scale(0.015625)\" />\n",
       "       <path id=\"ArialMT-67\" d=\"M 319 -275  L 866 -356  Q 900 -609 1056 -725  Q 1266 -881 1628 -881  Q 2019 -881 2231 -725  Q 2444 -569 2519 -288  Q 2563 -116 2559 434  Q 2191 0 1641 0  Q 956 0 581 494  Q 206 988 206 1678  Q 206 2153 378 2554  Q 550 2956 876 3175  Q 1203 3394 1644 3394  Q 2231 3394 2613 2919  L 2613 3319  L 3131 3319  L 3131 450  Q 3131 -325 2973 -648  Q 2816 -972 2473 -1159  Q 2131 -1347 1631 -1347  Q 1038 -1347 672 -1080  Q 306 -813 319 -275  z M 784 1719  Q 784 1066 1043 766  Q 1303 466 1694 466  Q 2081 466 2343 764  Q 2606 1063 2606 1700  Q 2606 2309 2336 2618  Q 2066 2928 1684 2928  Q 1309 2928 1046 2623  Q 784 2319 784 1719  z \" transform=\"scale(0.015625)\" />\n",
       "       <path id=\"ArialMT-61\" d=\"M 2588 409  Q 2275 144 1986 34  Q 1697 -75 1366 -75  Q 819 -75 525 192  Q 231 459 231 875  Q 231 1119 342 1320  Q 453 1522 633 1644  Q 813 1766 1038 1828  Q 1203 1872 1538 1913  Q 2219 1994 2541 2106  Q 2544 2222 2544 2253  Q 2544 2597 2384 2738  Q 2169 2928 1744 2928  Q 1347 2928 1158 2789  Q 969 2650 878 2297  L 328 2372  Q 403 2725 575 2942  Q 747 3159 1072 3276  Q 1397 3394 1825 3394  Q 2250 3394 2515 3294  Q 2781 3194 2906 3042  Q 3031 2891 3081 2659  Q 3109 2516 3109 2141  L 3109 1391  Q 3109 606 3145 398  Q 3181 191 3288 0  L 2700 0  Q 2613 175 2588 409  z M 2541 1666  Q 2234 1541 1622 1453  Q 1275 1403 1131 1340  Q 988 1278 909 1158  Q 831 1038 831 891  Q 831 666 1001 516  Q 1172 366 1500 366  Q 1825 366 2078 508  Q 2331 650 2450 897  Q 2541 1088 2541 1459  L 2541 1666  z \" transform=\"scale(0.015625)\" />\n",
       "       <path id=\"ArialMT-74\" d=\"M 1650 503  L 1731 6  Q 1494 -44 1306 -44  Q 1000 -44 831 53  Q 663 150 594 308  Q 525 466 525 972  L 525 2881  L 113 2881  L 113 3319  L 525 3319  L 525 4141  L 1084 4478  L 1084 3319  L 1650 3319  L 1650 2881  L 1084 2881  L 1084 941  Q 1084 700 1114 631  Q 1144 563 1211 522  Q 1278 481 1403 481  Q 1497 481 1650 503  z \" transform=\"scale(0.015625)\" />\n",
       "       <path id=\"ArialMT-69\" d=\"M 425 3934  L 425 4581  L 988 4581  L 988 3934  L 425 3934  z M 425 0  L 425 3319  L 988 3319  L 988 0  L 425 0  z \" transform=\"scale(0.015625)\" />\n",
       "       <path id=\"ArialMT-76\" d=\"M 1344 0  L 81 3319  L 675 3319  L 1388 1331  Q 1503 1009 1600 663  Q 1675 925 1809 1294  L 2547 3319  L 3125 3319  L 1869 0  L 1344 0  z \" transform=\"scale(0.015625)\" />\n",
       "      </defs>\n",
       "      <use xlink:href=\"#ArialMT-6e\" />\n",
       "      <use xlink:href=\"#ArialMT-65\" x=\"55.615234\" />\n",
       "      <use xlink:href=\"#ArialMT-67\" x=\"111.230469\" />\n",
       "      <use xlink:href=\"#ArialMT-61\" x=\"166.845703\" />\n",
       "      <use xlink:href=\"#ArialMT-74\" x=\"222.460938\" />\n",
       "      <use xlink:href=\"#ArialMT-69\" x=\"250.244141\" />\n",
       "      <use xlink:href=\"#ArialMT-76\" x=\"272.460938\" />\n",
       "      <use xlink:href=\"#ArialMT-65\" x=\"322.460938\" />\n",
       "     </g>\n",
       "    </g>\n",
       "    <g id=\"patch_4\">\n",
       "     <path d=\"M 8 48.551562  L 28 48.551562  L 28 41.551562  L 8 41.551562  z \" style=\"fill: #a1dab4; stroke: #444443; stroke-width: 0.4; stroke-linejoin: miter\" />\n",
       "    </g>\n",
       "    <g id=\"text_3\">\n",
       "     \n",
       "     <g style=\"fill: #444443\" transform=\"translate(31.5 48.551562)scale(0.1 -0.1)\">\n",
       "      <defs>\n",
       "       <path id=\"ArialMT-70\" d=\"M 422 -1272  L 422 3319  L 934 3319  L 934 2888  Q 1116 3141 1344 3267  Q 1572 3394 1897 3394  Q 2322 3394 2647 3175  Q 2972 2956 3137 2557  Q 3303 2159 3303 1684  Q 3303 1175 3120 767  Q 2938 359 2589 142  Q 2241 -75 1856 -75  Q 1575 -75 1351 44  Q 1128 163 984 344  L 984 -1272  L 422 -1272  z M 931 1641  Q 931 1000 1190 694  Q 1450 388 1819 388  Q 2194 388 2461 705  Q 2728 1022 2728 1688  Q 2728 2322 2467 2637  Q 2206 2953 1844 2953  Q 1484 2953 1207 2617  Q 931 2281 931 1641  z \" transform=\"scale(0.015625)\" />\n",
       "       <path id=\"ArialMT-6f\" d=\"M 213 1659  Q 213 2581 725 3025  Q 1153 3394 1769 3394  Q 2453 3394 2887 2945  Q 3322 2497 3322 1706  Q 3322 1066 3130 698  Q 2938 331 2570 128  Q 2203 -75 1769 -75  Q 1072 -75 642 372  Q 213 819 213 1659  z M 791 1659  Q 791 1022 1069 705  Q 1347 388 1769 388  Q 2188 388 2466 706  Q 2744 1025 2744 1678  Q 2744 2294 2464 2611  Q 2184 2928 1769 2928  Q 1347 2928 1069 2612  Q 791 2297 791 1659  z \" transform=\"scale(0.015625)\" />\n",
       "       <path id=\"ArialMT-73\" d=\"M 197 991  L 753 1078  Q 800 744 1014 566  Q 1228 388 1613 388  Q 2000 388 2187 545  Q 2375 703 2375 916  Q 2375 1106 2209 1216  Q 2094 1291 1634 1406  Q 1016 1563 777 1677  Q 538 1791 414 1992  Q 291 2194 291 2438  Q 291 2659 392 2848  Q 494 3038 669 3163  Q 800 3259 1026 3326  Q 1253 3394 1513 3394  Q 1903 3394 2198 3281  Q 2494 3169 2634 2976  Q 2775 2784 2828 2463  L 2278 2388  Q 2241 2644 2061 2787  Q 1881 2931 1553 2931  Q 1166 2931 1000 2803  Q 834 2675 834 2503  Q 834 2394 903 2306  Q 972 2216 1119 2156  Q 1203 2125 1616 2013  Q 2213 1853 2448 1751  Q 2684 1650 2818 1456  Q 2953 1263 2953 975  Q 2953 694 2789 445  Q 2625 197 2315 61  Q 2006 -75 1616 -75  Q 969 -75 630 194  Q 291 463 197 991  z \" transform=\"scale(0.015625)\" />\n",
       "      </defs>\n",
       "      <use xlink:href=\"#ArialMT-70\" />\n",
       "      <use xlink:href=\"#ArialMT-6f\" x=\"55.615234\" />\n",
       "      <use xlink:href=\"#ArialMT-73\" x=\"111.230469\" />\n",
       "      <use xlink:href=\"#ArialMT-69\" x=\"161.230469\" />\n",
       "      <use xlink:href=\"#ArialMT-74\" x=\"183.447266\" />\n",
       "      <use xlink:href=\"#ArialMT-69\" x=\"211.230469\" />\n",
       "      <use xlink:href=\"#ArialMT-76\" x=\"233.447266\" />\n",
       "      <use xlink:href=\"#ArialMT-65\" x=\"283.447266\" />\n",
       "     </g>\n",
       "    </g>\n",
       "   </g>\n",
       "  </g>\n",
       " </g>\n",
       "</svg></g>\n",
       "</g>\n",
       "</svg>"
      ],
      "text/plain": [
       "<dtreeviz.utils.DTreeVizRender at 0x24a4ede08e0>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import dtreeviz\n",
    "viz = dtreeviz.model(xg_stump, X_train=gini[['value']], \n",
    "                     y_train=gini.label=='pos', \n",
    "                     target_name='positive',\n",
    "                     feature_names=['value'], class_names=['negative', 'positive'],\n",
    "                     tree_index=0)\n",
    "viz.view()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "81170b23-e574-4a4a-ae4e-25abeaac6e13",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'img\\\\stump_xg.dot'"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "graph = xgb.to_graphviz(xg_stump, num_trees=0)\n",
    "graph.render('img/stump_xg', format='dot')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "94c195e2-16cf-45b6-a27c-820d1f1c0bb4",
   "metadata": {},
   "outputs": [],
   "source": [
    "my_dot_export(xg_stump, num_trees=0, filename='img/stump_xg.dot', title='A demo stump')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0df3e3a4-4c2d-4815-95c7-d3cfa634295e",
   "metadata": {},
   "source": [
    "```python\n",
    "import subprocess\n",
    "def my_dot_export(xg, num_trees, filename, title='', direction='TB'):\n",
    "    \"\"\"Exports a specified number of trees from an XGBoost model as a graph\n",
    "    visualization in dot and png formats.\n",
    "    Args:\n",
    "    xg: An XGBoost model.\n",
    "    num_trees: The number of tree to export.\n",
    "    filename: The name of the file to save the exported visualization.\n",
    "    title: The title to display on the graph visualization (optional).\n",
    "    direction: The direction to lay out the graph, either 'TB' (top to\n",
    "    bottom) or 'LR' (left to right) (optional).\n",
    "    \"\"\"\n",
    "    res = xgb.to_graphviz(xg, num_trees=num_trees)\n",
    "    content = f''' node [fontname = \"Roboto Condensed\"];\n",
    "    edge [fontname = \"Roboto Thin\"];\n",
    "    label = \"{title}\"\n",
    "    fontname = \"Roboto Condensed\"\n",
    "    '''\n",
    "    out = res.source.replace('graph [ rankdir=TB ]',\n",
    "                             f'graph [ rankdir={direction} ];\\n {content}')\n",
    "    # dot -Gdpi=300 -Tpng -ocourseflow.png courseflow.dot\n",
    "    dot_filename = filename\n",
    "    with open(dot_filename, 'w') as f:\n",
    "        f.write(out)\n",
    "    png_filename = dot_filename.replace('.dot', '.png')\n",
    "    subprocess.run(f'dot -Gdpi=300 -Tpng -o{png_filename} {dot_filename}'.split())\n",
    "```\n",
    "\n",
    "The `subprocess` module allows you to spawn new processes, connect to their input/output/error pipes, and obtain their return codes.\n",
    "\n",
    "```python\n",
    "def my_dot_export(xg, num_trees, filename, title='', direction='TB'):\n",
    "    \"\"\"\n",
    "    Exports a specified number of trees from an XGBoost model as a graph visualization in dot and png formats.\n",
    "\n",
    "    Args:\n",
    "    xg: An XGBoost model.\n",
    "    num_trees: The number of tree to export.\n",
    "    filename: The name of the file to save the exported visualization.\n",
    "    title: The title to display on the graph visualization (optional).\n",
    "    direction: The direction to lay out the graph, either 'TB' (top to bottom) or 'LR' (left to right) (optional).\n",
    "    \"\"\"\n",
    "\n",
    "    res = xgb.to_graphviz(xg, num_trees=num_trees)\n",
    "```\n",
    "\n",
    "This function `my_dot_export` takes an XGBoost model, a number of trees to export, a filename to save the exported visualization, an optional title, and an optional direction for the graph layout as input.\n",
    "\n",
    "It then uses `xgb.to_graphviz` to generate the graph visualization in the DOT format, given the number of trees to export.\n",
    "\n",
    "```python\n",
    "content = f''' node [fontname = \"Roboto Condensed\"];\n",
    "edge [fontname = \"Roboto Thin\"];\n",
    "label = \"{title}\"\n",
    "fontname = \"Roboto Condensed\"\n",
    "'''\n",
    "out = res.source.replace('graph [ rankdir=TB ]',f'graph [ rankdir={direction} ];\\n {content}')\n",
    "```\n",
    "This block of code sets up the content and styling of the exported visualization, including the font typeface and title.\n",
    "\n",
    "The `DOT` output from `xgb.to_graphviz` is then modified by replacing the default layout `TB` with the specified layout `direction`, and appending the content and styling setup.\n",
    "\n",
    "```python\n",
    "dot_filename = filename\n",
    "with open(dot_filename, 'w') as f:\n",
    "    f.write(out)\n",
    "png_filename = dot_filename.replace('.dot', '.png')\n",
    "subprocess.run(f'dot -Gdpi=300 -Tpng -o{png_filename} {dot_filename}'.split())\n",
    "```\n",
    "This block of code saves the modified `DOT` output to a file with the specified filename.\n",
    "\n",
    "It then uses the s`ubprocess` module to run the Graphviz `dot` command to convert the DOT file to a PNG image, which is saved with the same name as the DOT file but with a `.png` extension.\n",
    "\n",
    "Overall, this function generates a graph visualization of the specified number of trees from an XGBoost model in the DOT format, and saves it as both a DOT and PNG file."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0c85029d-7457-40a1-9203-9d4eb2d10893",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}