abalone_homework (1).ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import packages\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "names = ['Sex','Length','Diameter','Height','Whole weight','Shucked weight','Viscera weight','Shell weight','Rings']\n",
    "df = pd.read_csv('abalone.csv',header=None,names=names,sep=';')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Sex</th>\n",
       "      <th>Length</th>\n",
       "      <th>Diameter</th>\n",
       "      <th>Height</th>\n",
       "      <th>Whole weight</th>\n",
       "      <th>Shucked weight</th>\n",
       "      <th>Viscera weight</th>\n",
       "      <th>Shell weight</th>\n",
       "      <th>Rings</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>M</td>\n",
       "      <td>0.455</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.095</td>\n",
       "      <td>0.5140</td>\n",
       "      <td>0.2245</td>\n",
       "      <td>0.1010</td>\n",
       "      <td>0.150</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>M</td>\n",
       "      <td>0.350</td>\n",
       "      <td>0.265</td>\n",
       "      <td>0.090</td>\n",
       "      <td>0.2255</td>\n",
       "      <td>0.0995</td>\n",
       "      <td>0.0485</td>\n",
       "      <td>0.070</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>F</td>\n",
       "      <td>0.530</td>\n",
       "      <td>0.420</td>\n",
       "      <td>0.135</td>\n",
       "      <td>0.6770</td>\n",
       "      <td>0.2565</td>\n",
       "      <td>0.1415</td>\n",
       "      <td>0.210</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>M</td>\n",
       "      <td>0.440</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.125</td>\n",
       "      <td>0.5160</td>\n",
       "      <td>0.2155</td>\n",
       "      <td>0.1140</td>\n",
       "      <td>0.155</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>I</td>\n",
       "      <td>0.330</td>\n",
       "      <td>0.255</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.2050</td>\n",
       "      <td>0.0895</td>\n",
       "      <td>0.0395</td>\n",
       "      <td>0.055</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Sex  Length  Diameter  Height  Whole weight  Shucked weight  Viscera weight  \\\n",
       "0   M   0.455     0.365   0.095        0.5140          0.2245          0.1010   \n",
       "1   M   0.350     0.265   0.090        0.2255          0.0995          0.0485   \n",
       "2   F   0.530     0.420   0.135        0.6770          0.2565          0.1415   \n",
       "3   M   0.440     0.365   0.125        0.5160          0.2155          0.1140   \n",
       "4   I   0.330     0.255   0.080        0.2050          0.0895          0.0395   \n",
       "\n",
       "   Shell weight  Rings  \n",
       "0         0.150     15  \n",
       "1         0.070      7  \n",
       "2         0.210      9  \n",
       "3         0.155     10  \n",
       "4         0.055      7  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()\n",
    "#df.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# We change the M,F and I categorical variables as numerical using 0,1 and 2.\n",
    "replace_list = {\"Sex\" : {\"M\": 0, \"F\" : 1, \"I\": 2}}\n",
    "df.replace(replace_list,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Sex</th>\n",
       "      <th>Length</th>\n",
       "      <th>Diameter</th>\n",
       "      <th>Height</th>\n",
       "      <th>Whole weight</th>\n",
       "      <th>Shucked weight</th>\n",
       "      <th>Viscera weight</th>\n",
       "      <th>Shell weight</th>\n",
       "      <th>Rings</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0.455</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.095</td>\n",
       "      <td>0.5140</td>\n",
       "      <td>0.2245</td>\n",
       "      <td>0.1010</td>\n",
       "      <td>0.150</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0.350</td>\n",
       "      <td>0.265</td>\n",
       "      <td>0.090</td>\n",
       "      <td>0.2255</td>\n",
       "      <td>0.0995</td>\n",
       "      <td>0.0485</td>\n",
       "      <td>0.070</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0.530</td>\n",
       "      <td>0.420</td>\n",
       "      <td>0.135</td>\n",
       "      <td>0.6770</td>\n",
       "      <td>0.2565</td>\n",
       "      <td>0.1415</td>\n",
       "      <td>0.210</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0.440</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.125</td>\n",
       "      <td>0.5160</td>\n",
       "      <td>0.2155</td>\n",
       "      <td>0.1140</td>\n",
       "      <td>0.155</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2</td>\n",
       "      <td>0.330</td>\n",
       "      <td>0.255</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.2050</td>\n",
       "      <td>0.0895</td>\n",
       "      <td>0.0395</td>\n",
       "      <td>0.055</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Sex  Length  Diameter  Height  Whole weight  Shucked weight  \\\n",
       "0    0   0.455     0.365   0.095        0.5140          0.2245   \n",
       "1    0   0.350     0.265   0.090        0.2255          0.0995   \n",
       "2    1   0.530     0.420   0.135        0.6770          0.2565   \n",
       "3    0   0.440     0.365   0.125        0.5160          0.2155   \n",
       "4    2   0.330     0.255   0.080        0.2050          0.0895   \n",
       "\n",
       "   Viscera weight  Shell weight  Rings  \n",
       "0          0.1010         0.150     15  \n",
       "1          0.0485         0.070      7  \n",
       "2          0.1415         0.210      9  \n",
       "3          0.1140         0.155     10  \n",
       "4          0.0395         0.055      7  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import Normalizer\n",
    "from sklearn import metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "#!pip install mglearn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "import mglearn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "#mglearn.plots.plot_dbscan()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Take Abalone dataset. Run some clustering method (except k-means) with different number of clusters.Which number of clusters is better?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# abalone dataset\n",
    "X = df.drop(['Rings','Sex'],axis=1)\n",
    "label_rings=df.Rings\n",
    "scaler = Normalizer()\n",
    "scaler.fit(X)\n",
    "X_scaled = scaler.transform(X)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# AgglomerativeClustering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.cluster import AgglomerativeClustering\n",
    "agg = AgglomerativeClustering(n_clusters=2)\n",
    "label_pred= agg.fit_predict(X_scaled)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0,0.5,'Feature 1')"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot the cluster assignments\n",
    "plt.scatter(X_scaled[:, 0], X_scaled[:, 1], c=label_pred, cmap=\"plasma\")\n",
    "plt.xlabel(\"Feature 0\")\n",
    "plt.ylabel(\"Feature 1\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ARI = 0.032\n"
     ]
    }
   ],
   "source": [
    "print(\"ARI =\", metrics.adjusted_rand_score(label_rings, label_pred).round(3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "agg = AgglomerativeClustering(n_clusters=5)\n",
    "label_pred= agg.fit_predict(X_scaled)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([2, 0, 1, ..., 3, 3, 4], dtype=int64)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "label_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0,0.5,'Feature 1')"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(X_scaled[:, 0], X_scaled[:, 1], c=label_pred, cmap=\"plasma\")\n",
    "plt.xlabel(\"Feature 0\")\n",
    "plt.ylabel(\"Feature 1\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ARI = 0.06\n"
     ]
    }
   ],
   "source": [
    "print(\"ARI =\", metrics.adjusted_rand_score(label_rings, label_pred).round(2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ARI = 0.052\n"
     ]
    }
   ],
   "source": [
    "agg = AgglomerativeClustering(n_clusters=8)\n",
    "label_pred= agg.fit_predict(X_scaled)\n",
    "print(\"ARI =\", metrics.adjusted_rand_score(label_rings, label_pred).round(3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0,0.5,'Feature 1')"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(X_scaled[:, 0], X_scaled[:, 1], c=label_pred, cmap=\"plasma\")\n",
    "plt.xlabel(\"Feature 0\")\n",
    "plt.ylabel(\"Feature 1\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### comment: When the cluster no 5 , it gives highest score. Therefore we can say that 5 no of cluster is better."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.Take Abalone dataset. Reduce it’s dimensionality to 3 with Kernel PCA, then run clustering with 10 clusters.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Original shape: (4177, 7)\n",
      "Reduced shape: (4177, 3)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.decomposition import KernelPCA\n",
    "# keep the first three principal components of the data\n",
    "pca = KernelPCA(n_components=3, kernel='cosine',fit_inverse_transform=True)\n",
    "# fit PCA model to abalone dataset\n",
    "pca.fit(X_scaled)\n",
    "# transform data onto the first three principal components\n",
    "X_pca = pca.transform(X_scaled)\n",
    "print(\"Original shape: {}\".format(str(X_scaled.shape)))\n",
    "print(\"Reduced shape: {}\".format(str(X_pca.shape)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1346, 1986, 2468, 2162, 4162, 1014, 3064, 3023, 1668,  233],\n",
       "      dtype=int64)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.cluster import KMeans\n",
    "from sklearn.metrics import pairwise_distances_argmin_min\n",
    "km = KMeans(n_clusters=10).fit(X_pca)\n",
    "closest, _ = pairwise_distances_argmin_min(km.cluster_centers_, X_pca)\n",
    "closest"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### the closest gave output as array for the ten clusters. So X[3064] is the closest point in X to centroid 0, and X[1545] is the closest to centroid 1 and so on."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.cluster import KMeans\n",
    "kmeans = KMeans(n_clusters=10)\n",
    "kmeans.fit(X_pca)\n",
    "y_kmeans = kmeans.predict(X_pca)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(X_pca[:, 0], X_pca[:, 1], c=y_kmeans, s=50, cmap='viridis')\n",
    "\n",
    "centers = kmeans.cluster_centers_\n",
    "plt.scatter(centers[:, 0], centers[:, 1], c='black', s=200, alpha=0.5);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10, 3)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "centers.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3. Label propagation:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.mixture import GaussianMixture\n",
    "import matplotlib.pyplot as plt\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "import numpy as np\n",
    "%matplotlib inline\n",
    "from sklearn import datasets"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# cluster Abalone dataset without age (rings) feature using Gaussian Mixture Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "#abalone Dataset without ring\n",
    "X_df = df.drop(['Rings'],axis=1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "#X_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import Normalizer\n",
    "scaler = Normalizer()\n",
    "scaler.fit(X_df)\n",
    "X_scaled = scaler.transform(X_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Gaussian Mixture Model\n",
    "gmm = GaussianMixture(n_components=12)\n",
    "gmm.fit(X_scaled)\n",
    "labels=gmm.predict(X_scaled)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x27c8f7088d0>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(X_scaled[:, 0], X_scaled[:, 1], c=labels, s=40, cmap='viridis')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "#print(proba_lists[:5].round(3))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### take randomly 20 samples with age from the original Abalone dataset, predict their probabilities with GaussianMM (\"predict_proba\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Sex</th>\n",
       "      <th>Length</th>\n",
       "      <th>Diameter</th>\n",
       "      <th>Height</th>\n",
       "      <th>Whole weight</th>\n",
       "      <th>Shucked weight</th>\n",
       "      <th>Viscera weight</th>\n",
       "      <th>Shell weight</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3955</th>\n",
       "      <td>2</td>\n",
       "      <td>0.490</td>\n",
       "      <td>0.385</td>\n",
       "      <td>0.120</td>\n",
       "      <td>0.5910</td>\n",
       "      <td>0.2710</td>\n",
       "      <td>0.1125</td>\n",
       "      <td>0.1775</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>0</td>\n",
       "      <td>0.470</td>\n",
       "      <td>0.370</td>\n",
       "      <td>0.120</td>\n",
       "      <td>0.5795</td>\n",
       "      <td>0.2930</td>\n",
       "      <td>0.2270</td>\n",
       "      <td>0.1400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1304</th>\n",
       "      <td>0</td>\n",
       "      <td>0.535</td>\n",
       "      <td>0.435</td>\n",
       "      <td>0.150</td>\n",
       "      <td>0.7170</td>\n",
       "      <td>0.3475</td>\n",
       "      <td>0.1445</td>\n",
       "      <td>0.1940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>0</td>\n",
       "      <td>0.665</td>\n",
       "      <td>0.525</td>\n",
       "      <td>0.165</td>\n",
       "      <td>1.3380</td>\n",
       "      <td>0.5515</td>\n",
       "      <td>0.3575</td>\n",
       "      <td>0.3500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1989</th>\n",
       "      <td>2</td>\n",
       "      <td>0.280</td>\n",
       "      <td>0.210</td>\n",
       "      <td>0.075</td>\n",
       "      <td>0.1195</td>\n",
       "      <td>0.0530</td>\n",
       "      <td>0.0265</td>\n",
       "      <td>0.0300</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Sex  Length  Diameter  Height  Whole weight  Shucked weight  \\\n",
       "3955    2   0.490     0.385   0.120        0.5910          0.2710   \n",
       "46      0   0.470     0.370   0.120        0.5795          0.2930   \n",
       "1304    0   0.535     0.435   0.150        0.7170          0.3475   \n",
       "32      0   0.665     0.525   0.165        1.3380          0.5515   \n",
       "1989    2   0.280     0.210   0.075        0.1195          0.0530   \n",
       "\n",
       "      Viscera weight  Shell weight  \n",
       "3955          0.1125        0.1775  \n",
       "46            0.2270        0.1400  \n",
       "1304          0.1445        0.1940  \n",
       "32            0.3575        0.3500  \n",
       "1989          0.0265        0.0300  "
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_small=df.sample(20)\n",
    "rings= df['Rings'].sample(20)\n",
    "del X_small['Rings']\n",
    "X_small.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = Normalizer()\n",
    "scaler.fit(X_small)\n",
    "X1_scaled = scaler.transform(X_small)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "gmm = GaussianMixture(n_components=12)\n",
    "gmm.fit(X1_scaled)\n",
    "labels=gmm.predict(X1_scaled)\n",
    "cluster = gmm.predict_proba(X1_scaled)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 0. 0. 0. 0. 1. 0. 0. 0.]\n",
      " [0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
      " [0. 0. 0. 0. 1. 0. 0. 0. 0. 0. 0. 0.]]\n"
     ]
    }
   ],
   "source": [
    "print(cluster[:5].round(3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x27c8f779048>"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD8CAYAAACMwORRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAGntJREFUeJzt3XmYXXWd5/H39661pCpbVUhMUiSQRIgJhKYS4giogLKoQRERBBQ3hB70aZdWFId2sJ/ufnBaZ5xOK/SMI/Igm9pN1CCCsksCAUIgQEISCAlkI3tS212+80cV6SJ1i7pVde8995z6vJ4nz1P3d07O/fyo4pNT557F3B0REYmWWNABRESk9FTuIiIRpHIXEYkglbuISASp3EVEIkjlLiISQSp3EZEIUrmLiESQyl1EJIISQb1xU1OTT5s2Lai3FxEJpSeffPINd28eaL3Ayn3atGmsWLEiqLcXEQklM9tYzHo6LCMiEkEqdxGRCFK5i4hEkMpdRCSCVO4iIhGkchcRiSCVu4hIBKncRUQiSOUuIhJBKncRkQgKdbnn83mymVzQMUREqk5g95YZjv172/jJ9/6dh5euIpfLM+Ndk/nr//4xjpl3ZNDRRESqQuj23HO5PN+4YDEPL11FNpPD885Lz27m6ot/ysa1W4OOJyJSFUJX7k8+tIYdr+/uczgm05nl1n+5L6BUIiLVJXTlvn71a3S0Z/qM5/POiyuLuhOmiEjkha7cmyaNJl2TLLhswuSxFU4jIlKdQlfup5x9HPFEvM94ujbJBVecFkAiEZHqE7pyr6lL80+3XEHTxNHU1qepG1VDuibJZ75+Nq3vPSboeCIiVSGUp0LOeNdkbnrkGtau2kT7wS6OmddCbX066FgiIlUjlOUOEIvFdF67iEg/QndYRkREBqZyFxGJIJW7iEgEqdxFRCJI5S4iEkEqdxGRCFK5i4hEkMpdRCSCVO4iIhGkchcRiSCVu4hIBKncRUQiSOUuIhJBKncRkQgK7S1/d3Tu4tE3nqQ918G8McdyTMPRmFnQsUREqkIoy/3P2x/jxvW34ThZz/Lb1//EnNGz+NYxXyJufR/BJyIy0oTusMyurj3cuP42Mp4h61kAOvNdPLd3LfdufTTgdCIi1aGocjezs8xsjZmtM7OrCyy/zMx2mNnKnj9fKH3Ubo/tfLrgeGe+iz9ue7hcbysiEioDHpYxsziwGPgAsBl4wsyWuPvzh616u7tfVYaMb9GZ6yLnucLL8l3lfnsRkVAoZs99AbDO3Te4exdwG3BueWP1b96YY0nG+v6blLAEJ42bF0AiEZHqU0y5TwY29Xq9uWfscB83s1Vm9iszm1poQ2Z2uZmtMLMVO3bsGEJcOGpUCyeNO550LHVoLGEJGhL1nDv5jCFtU0Qkaoop90LnF/phr38LTHP344D7gJsKbcjdb3T3VndvbW5uHlzSXq6cdgknLD+R3Dcb6PxKPRNubeGaiVcxOtkw5G2KiERJMeW+Gei9Jz4FeL33Cu6+0907e17+G3BiaeL15e78/ZW/4NEbnqfz9Ry5Xc7Lf9jO1R+9gZ3b9pbrbUVEQqWYcn8CmGlm080sBVwILOm9gplN6vVyEfBC6SK+1YtPb+TZZevp7MgcGstl8xzc38kdP/lzud5WRCRUBix3d88CVwH30F3ad7j7ajO7zswW9az2FTNbbWbPAF8BLitX4GceW09XZ6bPeC6b4/H7y/ZviohIqBR1haq7LwWWHjZ2ba+vvw18u7TRChvVWEMilaCro2/B1zfWViKCiEjVC90Vqiefc3zB8XRtinMvO7nCaUREqlPoyn3M+FF880efIl2TpKYuRTKVIFWT5OSz53L6x8r2Oa6ISKiE8sZh7zlzLsf95b/xl3uepb2tixPeM5MjZ00MOpaISNUIZbkDNIyp48xPnhR0DBGRqhS6wzIiIjIwlbuISASp3EVEIkjlLiISQSp3EZEIUrmLiESQyl1EJIJU7iIiERTKi5i68h08tP0XPLvnPrLeRUvdXE6f+EWa0i1BRxMRqQqh23N3d27d+G2e2v17OvIHyHoXGw4+xU0vf5VdXa8PvAERkREgdOW+8eAz7OjYSM573/LXyeQ7+cuO2wPLJSJSTUJX7q+1v0jm0BP9/pOT59W2VQEkEhGpPqEr9/rEGJKWKrhsVGJchdOIiFSn0JX7MY2n4AXG4yQ5afx5Fc8jIlKNQlfuqVgNMYv3Gc+RYVRifACJRESqT+jKfdWee+nKtxVcdveWH1c4jYhIdQpdua/Z92i/y3Z2bqpgEhGR6hW6co9bKK+7EhGpqNCV+9jU5H6XJWM1FUwiIlK9QlfutYmGfpelY/UVTCIiUr1CV+7T608gQd/z3I04xzSeHEAiEZHqE7pyn1Q7ixkNC0ha+tBYjAR18QYWNn08wGQiItUjdOUOsGjy3zJ79PtIWg0xEkysncHF066nPjE26GgiIlUhlOW+dMuPeX7vg2S8gzxZtnWs59aN36EtuzfoaCIiVSF05b61fT0v7nuEjHccGst5hrbsHpbv/HWAyUREqkfoyv3lg0+S92yf8RxZXnybC5xEREaS0JV7wlKYFY6djKULjouIjDShK/djGk/GsD7jSUtzwthzAkgkIlJ9QlfuDckmPjDxShKWItbzCNik1TC1bg4njD074HQiItUhlDdqmTf2TKbVz+P5vQ/QlW/nqFEnMrVuDmZ99+hFREaiUJY7wJjUEfyX5k8GHUNEpCoVdVjGzM4yszVmts7Mrn6b9c43Mzez1tJFFBGRwRqw3M0sDiwGzgZmAxeZ2ewC6zUAXwGWlzqkiIgMTjGHZRYA69x9A4CZ3QacCzx/2HrfB64HvlHShCIybNl8np88s5z/t/op9nV1MGf8EVxz0vuYP3FK0NGkTIo5LDMZ6P2Io809Y4eY2QnAVHf/3dttyMwuN7MVZrZix44dgw4rIkPz1Qd+z+KVy9jZ0UYmn+fpHVu45O47eWLr5qCjSZkUU+6FTkHxQwu7ryj6EfD1gTbk7je6e6u7tzY3NxefUkSGbOO+3dyzcR3tubde2d2Ry/IPjz8YUCopt2LKfTMwtdfrKcDrvV43AHOAB8zsFWAhsEQfqopUh5U7tpKIFf5fffXObRVOI5VSTLk/Acw0s+lmlgIuBJa8udDd97p7k7tPc/dpwDJgkbuvKEtiERmUptq6fpeNTunRlFE1YLm7exa4CrgHeAG4w91Xm9l1Zrao3AFFZHgWTpxKQzLV5/hqTTzB5+acGEgmKT9z94HXKoPW1lZfsUI79yKVsG7PTi5eegcHMl0AZPI5PjT9nfyPU88m3s8hG6lOZvakuw942DuUV6jm884fH32BX9+7krb2Lk5tncGF55zI6IbaoKOJVKUZY8bz2EVXsHzLJt5ob+P4CRNpaRgTdCwpo1CW+3U/uZsHn1hHR2cGgE1b9/D7B1fzi3+6lDGN/R9fFBnJYma8+x0tQceQCgnd72NrXt7GA4+/dKjYATLZHHv2t3PL73SYR0QEQljuy1e9Qjab6zOeyeZ44PGXAkgkIlJ9QlfuNekk8Xjh2Ol0KI8yiYiUXOjK/f0LZhUcr0klOO+M4yucRkSkOoWu3JvHjeJvP3c66WSCZKI7fm06yQmzp7LotOMCTiciUh1CeRzjw++bS+ucI7n3Ly/S1tHFwuOmcdw7J+tJTCIiPUJZ7gATmxq5dNGCoGOIiFSl0Ja7SBjt3nWAl9ZuY+y4embMPEK/bUrZqNxFKiCfd/73/7yHP9z9DKlkglwuz4QJjfzD9Z9k4qThXymay7exs+0+MvldNKZPpCE9twSpJcxU7iIV8Os7l3PvPc+S6cqR6eq+TmPz5l188+u3ctMtVwxrD35fx1M8t+1zOI57BrM4o9PzmX3ET4lZqlRTkJAJ3dkyImF05+2P09GRectYPu/s3nWQ1c8N/WlIee/kue1fIOcHyPtBnC7y3s7ezsfZtPenw40tIaZyF6mAvXvbCo6bwRs79g95u7vbHwHve8V23jvYsv/WIW9Xwk/lLlIB06Y1FRzPZvPMnDVxyNvN5vfhFL5tdy5/cMjblfBTuYtUwBevOK3P7THS6QTzTzqKyVPGDXm7o2sWAH333MEYXTN/yNuV8FO5i1RA6/yj+LvrPk7LkeMBqK1L8dHzWvnu331sWNutSUxmQv15xKz3swxixKyW6WO/NaxtS7jpbBmRClmw8GgWLDyafN6JxUp3fvuM8dcxKjWX1/b/jGxuN40185k25m+oS80o2XtI+KjcRSqslMUOYGZMaryASY0XlHS7Em46LCMiEkEqdxGRCFK5i4hEkMpdRCSCVO4iIhGkchcRiSCdCinSD/csdN6Pdz4A1ojVnoclZwYdS6QoKneRAtw78V2XQnYteBsQx9tuwRu+Qaz+00HHExmQDsuIFOAHb4LMiz3FDt33b+mA/T/Ac1uDjCZSFJW7SCHtvwY6Ci/r+GNFo4gMhcpdpKBsP+N5INPPMpHqoXIXKaTmHCBZYEEM0u+vdBqRQVO5ixRg9V+A+ESgptdgLdR9CkscFVgukWLpbBmRAiw2GsbfhbffCR33QqwRq/sUpE4JOppIUVTuIv2w2Cis/rNQ/9mgo4gMmg7LiIhEUFHlbmZnmdkaM1tnZlcXWH6FmT1rZivN7BEzm136qCIiUqwBy93M4sBi4GxgNnBRgfL+pbvPdfd5wPXAD0ueVEREilbMnvsCYJ27b3D3LuA24NzeK7j7vl4v6wEvXUQRERmsYj5QnQxs6vV6M3DS4SuZ2X8FvgakgNMKbcjMLgcuB2hpaRlsVhERKVIxe+6FnubbZ8/c3Re7+9HAt4DvFtqQu9/o7q3u3trc3Dy4pCIiUrRiyn0zMLXX6ynA62+z/m3AR4cTSkREhqeYcn8CmGlm080sBVwILOm9gpn1vsn1h4CXShdRREQGa8Bj7u6eNbOrgHuAOPAzd19tZtcBK9x9CXCVmZ1B9x2VdgOfKWfornyWn697jDtfWUFHLsPJE2Zw1bHvZ3Ld2HK+rYhIaJh7MCe2tLa2+ooVKwb999ydLz52M0/vfJWOfPed+2IYo5JpfvP+K5lUO7rUUUVEqoaZPenurQOtF7orVJ/etYmVuzYdKnaAPE5btov/s/bhAJOJiFSP0JX7ip2v0Jnre6/trOd5ZPu6ABKJiFSf0JV7Y7KWVCxecNnoVF2F04iIVKfQlftZk98F1vfU+9p4kkuP6nNtlYjIiBS6ch+TquOHrZ+gNp6kLp6iJpYgHUvw4SnH8eEpxwUdT4Zp58E2/vXR5XzlN79j8SPLeOPAwaAjiYRS6M6WedOBTAf3b11DW7aLhc1HceSo8SVMJ0F4ftt2Lrn5TjL5HJ3ZHOlEnEQsxs0Xf4I5k44IOp5IVYjs2TJvGpWs4SNTj+eT0+er2CPiG3fdzYGuLjqzOQA6szkOdmX42l1LCWonRCSsQlvuEi1b9x9g0569BZdt2XeA1/buK7hMRApTuUtVeLs9cwPy2nMXGRSVu1SFiQ2jmNTYUHBZU30dU8foymORwVC5S1UwM67/yFnUJZMk493XMSRiMeqSSa5fdBZW4PRXEelfMQ/rEKmIeZMnsfTyT3PzipW8uH0Hs5qbuLR1HlO01y4yaCp3qSrvGN3It04/NegYIqGnwzIiIhEU6nLvyGXYn+kIOoaISNUJ5WGZnZ0HuPbpJTy6fT2OM7V+HNce/yEWNE0POpqISFUI3Z57Np/j4of+L49sX0fGc2Q9z8sH3uDKZb/khb1bgo4nIlIVQlfuD217iV1dB8l6/i3jnbkMN6x5KKBUIiLVJXTlvnbfNtqyXX3GHXhee+4iIkAIy31y3Rhq46mCy6bqAdkiIkAIy/0D75hNKhbn8OsVa+JJvjjrlEAyiYhUm9CVe008yU0nf5aW+nHUxpPUJ9LUJ1J8Z+7ZLGw+Kuh4IhJCma4MD/9mOXf+82958t5nyOfzA/+lKhfKUyFnNE7g96d/mfX7d9CW7eKY0RNJxUM5FREJ2OaXtvD1915L+8EOMp0ZkqkkE1qa+OGD19E4vvDN7MIgdHvubzIzZjRO4LhxU1TsIjJk3zvvB+zetpf2/R1ku3K0H+jgtZe28KPLbwg62rCEttxFRIZr05rX2Pry9j7PE8hmciz7/ZN0tncGlGz4VO4iMmId2NNGPtf/8fXO9r6nXYeFyl1ERqxXX9hMpjNTcFnzlHE0jB1V4USlo3IXkRHr9uvv6nfZh7/0wVA/JEblLiIj1o7NOwuOp2qSNDY1VjhNaancRWTEannnOwqOx2Ixps+ZWuE0paVyF5ER67K/v4h03VtvZ5JMJ5g+t4VZrUcHlKo0VO4iMmLNP3Me3/z5VTRNHkcynSCZTvCej53EP97z3VAfb4eQXqEqIlIqp57/bk75+EL2vrGP2lE1pGvTQUcqCZW7iIx4ZsaY5tFBxyipog7LmNlZZrbGzNaZ2dUFln/NzJ43s1Vm9iczO7L0UUVEpFgDlruZxYHFwNnAbOAiM5t92GpPA63ufhzwK+D6UgcVEZHiFbPnvgBY5+4b3L0LuA04t/cK7n6/u7f1vFwGTCltTBERGYxiyn0ysKnX6809Y/35PHD3cEKJiMjwFPOBaqHzgbzAGGZ2CdAKvLef5ZcDlwO0tLQUGVFERAarmD33zUDvS7WmAK8fvpKZnQFcAyxy94L3yXT3G9291d1bm5ubh5JXRESKUEy5PwHMNLPpZpYCLgSW9F7BzE4AbqC72LeXPqaIiAzGgOXu7lngKuAe4AXgDndfbWbXmdmintV+AIwC7jSzlWa2pJ/NiYhIBRR1EZO7LwWWHjZ2ba+vzyhxLhERGQZdoSoiMkxrX9/B9f/xIE9teI1UIs5HWmfzNx8+mfqa1MB/uUxU7iIiw/DK9t18+n/dTltX9xOdsrk8/778OVZt3MKtX/0UsVgwNyDTXSFFRIbh3+5dTkcm+5axrmyOjdt3s2ztqwGlUrmLiAzLUxteI+99L/1pz2R47tWtASTqpnIXERmG5sbCD9GuSSZpaqyvcJr/pHIXERmGy047kdpU348vYwYfnDczgEQ97x/YO4uIRMBpc2fw2dPmk0rEqU+nqE+nGF1Xw0++dB6jaoJ78IfOlhERGaYrzlzIhScfz1MbXqMunaL16Ckk4sHuO6vcRURKYEx9LafNnRF0jEN0WEZEJIJU7iIiEaRyFxGJIJW7iEgEqdxFRCJI5S4iEkEqdxGRCFK5i4hEkMpdRCSCVO4iImWUzzvb9hxgf3tnRd9Xtx8QESmT+555iX/8zf3sa+8gn3dOmtnC9z91JuMb6sr+3tpzFxEpgxXrNvOdW/7Ajn0H6czkyOTyLFv7Kp/7lzvI5/s+3KPUVO4iImXwr394rM/j97L5PNv2HmD5S+V//J7KXUSkDDZs21lwPJvLs35r4WWlpHIXESmDKeNHFxxPxmO0NI0p+/ur3EVEyuBLZy6kJvnWc1ZiZjTW1fCeY6eV/f1V7iIiZXDKsdP51sfex6iaFHXpJOlEnNlTj+DnX76AeKz81atTIUVEyuTj757LovmzeXn7bhpqU0wa21ix91a5i4iUUTIRZ9Y7mir+vjosIyISQSp3EZEIUrmLiESQyl1EJIJU7iIiEaRyFxGJIJW7iEgEFVXuZnaWma0xs3VmdnWB5aea2VNmljWz80sfU0REBmPAcjezOLAYOBuYDVxkZrMPW+1V4DLgl6UOKCIig1fMFaoLgHXuvgHAzG4DzgWef3MFd3+lZ1m+DBlFRGSQijksMxnY1Ov15p4xERGpUsXsuVuBsSE9I8rMLgcu73l5wMzWDGU7h2kC3ijBdsJEcx4ZRtqcR9p8YWhzPrKYlYop983A1F6vpwCvDzIMAO5+I3DjUP5uf8xshbu3lnKb1U5zHhlG2pxH2nyhvHMu5rDME8BMM5tuZingQmBJOcKIiEhpDFju7p4FrgLuAV4A7nD31WZ2nZktAjCz+Wa2GfgEcIOZrS5naBEReXtF3c/d3ZcCSw8bu7bX10/QfbgmCCU9zBMSmvPIMNLmPNLmC2Wcs7kP6bNRERGpYrr9gIhIBIWm3Iu4BULazG7vWb7czKZVPmVpFTHnr5nZ82a2ysz+ZGZFnSJVzQaac6/1zjczN7NQn11RzHzN7IKe7/NqMwv9VeBF/Fy3mNn9ZvZ0z8/2OUHkLBUz+5mZbTez5/pZbmb2457/HqvM7K9K8sbuXvV/gDiwHjgKSAHPALMPW+evgZ/2fH0hcHvQuSsw5/cDdT1fXzkS5tyzXgPwELAMaA06d5m/xzOBp4GxPa8nBJ27AnO+Ebiy5+vZwCtB5x7mnE8F/gp4rp/l5wB3031N0UJgeSneNyx77odugeDuXcCbt0Do7Vzgpp6vfwWcbmaFLsAKiwHn7O73u3tbz8tlBPehdqkU830G+D5wPdBRyXBlUMx8vwgsdvfdAO6+vcIZS62YOTvQ2PP1aIZ4XU21cPeHgF1vs8q5wC+82zJgjJlNGu77hqXci7kFwqF1vPv0zb3A+IqkK4/B3vbh83T/6x9mA87ZzE4Aprr77yoZrEyK+R7PAmaZ2aNmtszMzqpYuvIoZs7fAy7pOb16KfDlykQLTFlu8VLUqZBVoJhbIJTsNglVouj5mNklQCvw3rImKr+3nbOZxYAf0X0H0igo5nucoPvQzPvo/s3sYTOb4+57ypytXIqZ80XAz939n83s3cDNPXOO6o0Jy9JdYdlzL+YWCIfWMbME3b/Ovd2vQtWuqNs+mNkZwDXAInfvrFC2chlozg3AHOABM3uF7uOTS0L8oWqxP9d3uXvG3V8G1tBd9mFVzJw/D9wB4O6PATV034Mlqkp2i5fewlLuxdwCYQnwmZ6vzwf+7D2fVoTUgHPuOURxA93FHvZjsTDAnN19r7s3ufs0d59G9+cMi9x9RTBxh62Yn+v/oPuDc8ysie7DNBsqmrK0ipnzq8DpAGZ2LN3lvqOiKStrCfDpnrNmFgJ73X3LsLca9CfJg/jE+RxgLd2ftF/TM3Yd3f9zQ/cPwJ3AOuBx4KigM1dgzvcB24CVPX+WBJ253HM+bN0HCPHZMkV+jw34Id3PT3gWuDDozBWY82zgUbrPpFkJfDDozMOc763AFiBD917654ErgCt6fY8X9/z3eLZUP9O6QlVEJILCclhGREQGQeUuIhJBKncRkQhSuYuIRJDKXUQkglTuIiIRpHIXEYkglbuISAT9f9i7vP9MUkPQAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(X1_scaled[:, 0], X1_scaled[:, 1], c=labels, s=40, cmap='viridis')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Transform 'Rings' to multiple class\n",
    "rings=pd.get_dummies(rings)\n",
    "ring_values= np.array(rings.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "#rings.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "#ring_values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Get cluster x class matrix\n",
    "cluster_age = cluster.T @ rings\n",
    "# normalize so probabilities sum up to 1\n",
    "cluster_age = cluster_age / cluster_age.sum(1)[:,None]\n",
    "#cluster_age"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Get samples x class matrix\n",
    "predicted_classes= gmm.predict_proba(X_df) @ cluster_age\n",
    "#predicted_classes[:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Predict rings by multiplying 'Rings' values of classes with class probabilities\n",
    "\n",
    "predicted_age= predicted_classes @ ring_values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4177,)"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predicted_age.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 7. ,  7. , 13.5, ...,  9.6, 13.5, 13.5])"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predicted_age.round(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}