PDC_sampler_version0.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import numpy as np\n",
    "import copy as cp\n",
    "import os.path\n",
    "import matplotlib.pyplot as plt\n",
    "import sys\n",
    "import pickle\n",
    "import scipy.misc as scm\n",
    "import seaborn as sns\n",
    "import matplotlib.cm as cm\n",
    "\n",
    "\n",
    "from scipy import stats\n",
    "import sklearn\n",
    "from sklearn import datasets\n",
    "from sklearn.semi_supervised import label_propagation\n",
    "from sklearn.metrics import classification_report, confusion_matrix\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn import svm\n",
    "from sklearn.metrics import accuracy_score, f1_score\n",
    "from sklearn.metrics import precision_recall_fscore_support\n",
    "\n",
    "from PIL import Image\n",
    "from pylab import *\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "system = 'SiO2-Al2O3-MgO'\n",
    "field_size = []  #                                                                                                         \n",
    "LP_algorithm = 'LP'                                                                    \n",
    "US_strategy = 'LC'      \n",
    "initial_random = 9 \n",
    "total_trials = 400#250 #400                                                                                                 \n",
    "total_episodes = 1\n",
    "is_output_img = True\n",
    "use_cross_selection = True\n",
    "fig_title = \"Sampling by LP+LC for SiO$_2$-Al$_2$O$_3$-MgO\" \n",
    "\n",
    "\n",
    "\n",
    "img = Image.open('PD_examples/SiO2-Al2O3-MgO_labeled.png')\n",
    "phase_class = 10\n",
    "    \n",
    "    \n",
    "rgb = img.convert('RGB')\n",
    "img_array = np.asarray(rgb)/255.\n",
    "\n",
    "\n",
    "xlabel = ' '\n",
    "ylabel = ' '"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "field_range = [[140, 305], [1453, 1425]] \n",
    "field_size = [field_range[1][0] - field_range[0][0], field_range[1][1] - field_range[0][1]]\n",
    "axis_range = [[0, 10000], [0, 17321]] #x_axis, y_axis\n",
    "\n",
    "\n",
    "def drange(begin, end, step):\n",
    "    n = begin\n",
    "    while n+step < end:\n",
    "        yield n\n",
    "        n += step\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "#grid point list\n",
    "\n",
    "\n",
    "\n",
    "temp_range = [0, 10000, round(10000./40)] \n",
    "prs_range = [0, 17321, round(17321./40)] \n",
    "\n",
    "if system == 'SiO2-Al2O3-MgO':\n",
    "    grid_list = []\n",
    "    grid_list_list = [[], [], [], [], [], [], [], [], [], []]\n",
    "    sgn=1\n",
    "    for y in range(prs_range[0], prs_range[1], prs_range[2]):\n",
    "        for x in range(temp_range[0], temp_range[1], temp_range[2]):\n",
    "            shift = 0 if sgn==1 else 0.5*temp_range[2]\n",
    "            x=round(x+shift)\n",
    "            yf = round(field_range[1][1] - ((float(y) - prs_range[0])/(prs_range[1] - prs_range[0]))*field_size[1]) \n",
    "            xf = round(field_range[0][0] + ((float(x)- temp_range[0])/(temp_range[1] - temp_range[0]))*field_size[0])\n",
    "            \n",
    "            rgb = img_array[yf][xf]\n",
    "            if np.linalg.norm(rgb - np.array([0,0,0]))<0.1:\n",
    "                continue\n",
    "            elif np.linalg.norm(rgb - np.array([1,0,0]))<0.1:\n",
    "                grid_list.append([y, x, 0])\n",
    "                grid_list_list[0].append([y,x])\n",
    "            elif np.linalg.norm(rgb - np.array([0,1,0]))<0.1:\n",
    "                grid_list.append([y, x, 1])\n",
    "                grid_list_list[1].append([y,x])\n",
    "            elif np.linalg.norm(rgb - np.array([0,0,1]))<0.1:\n",
    "                grid_list.append([y, x, 2])\n",
    "                grid_list_list[2].append([y,x])\n",
    "            elif np.linalg.norm(rgb - np.array([0,1,1]))<0.1:\n",
    "                grid_list.append([y, x, 3])\n",
    "                grid_list_list[3].append([y,x])\n",
    "            elif np.linalg.norm(rgb - np.array([1,0,1]))<0.1:\n",
    "                grid_list.append([y, x, 4])\n",
    "                grid_list_list[4].append([y,x])\n",
    "            elif np.linalg.norm(rgb - np.array([1,1,0]))<0.1:\n",
    "                grid_list.append([y, x, 5])\n",
    "                grid_list_list[5].append([y,x])\n",
    "            elif np.linalg.norm(rgb - np.array([1,1,1]))<0.1:\n",
    "                grid_list.append([y, x, 6])\n",
    "                grid_list_list[6].append([y,x])\n",
    "            elif np.linalg.norm(rgb - np.array([0.5, 0, 0]))<0.1:\n",
    "                grid_list.append([y, x, 7])\n",
    "                grid_list_list[7].append([y,x])\n",
    "            elif np.linalg.norm(rgb - np.array([0, 0.5, 0]))<0.1:\n",
    "                grid_list.append([y, x, 8])\n",
    "                grid_list_list[8].append([y,x])\n",
    "            elif np.linalg.norm(rgb - np.array([0, 0, 0.5]))<0.1:\n",
    "                grid_list.append([y, x, 9])\n",
    "                grid_list_list[9].append([y,x])\n",
    "                \n",
    "        sgn = sgn*(-1)\n",
    "        \n",
    "    grid_list = np.array(grid_list)\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/utils/validation.py:475: DataConversionWarning: Data with input dtype int64 was converted to float64 by StandardScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/utils/validation.py:475: DataConversionWarning: Data with input dtype int64 was converted to float64 by StandardScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n"
     ]
    }
   ],
   "source": [
    "#Standardization\n",
    "\n",
    "ss = StandardScaler()\n",
    "ss.fit(grid_list[:,0:2])\n",
    "grid_list_std = ss.transform(grid_list[:,0:2])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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": [
    "color_list = [cm.rainbow(float(i)/(phase_class-1)) for i in range(phase_class)]\n",
    "\n",
    "for i in range(phase_class):\n",
    "    plt.scatter(np.array(grid_list_list[i])[:,1],np.array(grid_list_list[i])[:,0], s=2, color = color_list[i])\n",
    "if system == 'H2O_detail':\n",
    "    plt.xlim(-60,30)\n",
    "    plt.ylim(0,1000)\n",
    "elif system == 'H2O':\n",
    "    plt.xlim(-273,374)\n",
    "    plt.ylim(0,80)\n",
    "plt.grid()\n",
    "plt.savefig('grid_label.png', dpi=300)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def cross_selection(prev_grid_id, unlabeled_indices, grid_list, system):\n",
    "\n",
    "    if system == 'SiO2-Al2O3-MgO':\n",
    "        \n",
    "        candidate_indices = []\n",
    "        for ui in unlabeled_indices:\n",
    "            difx_step = (grid_list[prev_grid_id][1] - grid_list[ui][1])/(round(10000./40)/2)\n",
    "            if grid_list[ui][0] == grid_list[prev_grid_id][0] or np.abs(grid_list[prev_grid_id][0]-grid_list[ui][0]) == np.abs(round(17321./40)*difx_step):\n",
    "                candidate_indices.append(ui)\n",
    "        if len(candidate_indices)>0:\n",
    "            return candidate_indices\n",
    "        else:\n",
    "            return unlabeled_indices\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "trial: 9\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/semi_supervised/label_propagation.py:288: ConvergenceWarning: max_iter=1000 was reached without convergence.\n",
      "  category=ConvergenceWarning\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.00      0.00      0.00         1\n",
      "          1       0.00      0.00      0.00        10\n",
      "          2       0.54      0.77      0.63        77\n",
      "          3       0.69      0.92      0.79        71\n",
      "          4       0.00      0.00      0.00        57\n",
      "          5       0.00      0.00      0.00        24\n",
      "          6       0.00      0.00      0.00        21\n",
      "          7       0.55      0.82      0.66        98\n",
      "          8       0.80      0.80      0.80       270\n",
      "          9       0.88      0.93      0.90       191\n",
      "\n",
      "avg / total       0.64      0.73      0.68       820\n",
      "\n",
      "macro_report (0.34587887374543425, 0.4224754825273118, 0.3779578633485591, None) [2 3 7 8 9]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/semi_supervised/label_propagation.py:288: ConvergenceWarning: max_iter=1000 was reached without convergence.\n",
      "  category=ConvergenceWarning\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "classes [2 3 7 8 9]\n",
      "[1732 1000    9]\n",
      "[14722  4750     3]\n",
      "[9093 3125    7]\n",
      "[3897 1625    9]\n",
      "[9959 3375    7]\n",
      "[15588  5000     3]\n",
      "[15588  4500     3]\n",
      "[8660 6500    2]\n",
      "[3464 5000    8]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "trial: 10\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/semi_supervised/label_propagation.py:288: ConvergenceWarning: max_iter=1000 was reached without convergence.\n",
      "  category=ConvergenceWarning\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.00      0.00      0.00         1\n",
      "          1       0.00      0.00      0.00        10\n",
      "          2       0.82      0.61      0.70        77\n",
      "          3       0.67      0.92      0.77        71\n",
      "          4       0.00      0.00      0.00        57\n",
      "          5       0.00      0.00      0.00        24\n",
      "          6       0.00      0.00      0.00        21\n",
      "          7       0.62      0.64      0.63        98\n",
      "          8       0.65      0.99      0.79       270\n",
      "          9       0.92      0.74      0.82       191\n",
      "\n",
      "avg / total       0.64      0.71      0.66       820\n",
      "\n",
      "macro_report (0.36880480137790944, 0.39047878009680237, 0.37167257610377047, None) [2 3 7 8 9]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/semi_supervised/label_propagation.py:288: ConvergenceWarning: max_iter=1000 was reached without convergence.\n",
      "  category=ConvergenceWarning\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "classes [2 3 7 8 9]\n",
      "[1732 1000    9]\n",
      "[14722  4750     3]\n",
      "[9093 3125    7]\n",
      "[3897 1625    9]\n",
      "[9959 3375    7]\n",
      "[15588  5000     3]\n",
      "[15588  4500     3]\n",
      "[8660 6500    2]\n",
      "[3464 5000    8]\n",
      "[5629 4375    8]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/matplotlib/cbook/deprecation.py:107: MatplotlibDeprecationWarning: Passing one of 'on', 'true', 'off', 'false' as a boolean is deprecated; use an actual boolean (True/False) instead.\n",
      "  warnings.warn(message, mplDeprecation, stacklevel=1)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "trial: 11\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/semi_supervised/label_propagation.py:288: ConvergenceWarning: max_iter=1000 was reached without convergence.\n",
      "  category=ConvergenceWarning\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n",
      "/Users/terayama/Downloads/ENTER/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.00      0.00      0.00         1\n",
      "          1       0.00      0.00      0.00        10\n",
      "          2       0.94      0.60      0.73        77\n",
      "          3       0.67      0.92      0.77        71\n",
      "          4       0.00      0.00      0.00        57\n",
      "          5       0.76      0.79      0.78        24\n",
      "          6       0.00      0.00      0.00        21\n",
      "          7       0.68      0.64      0.66        98\n",
      "          8       0.66      0.97      0.79       270\n",
      "          9       0.92      0.76      0.83       191\n",
      "\n",
      "avg / total       0.68      0.73      0.69       820\n",
      "\n",
      "macro_report (0.46327042254944334, 0.46858913445061223, 0.4564347216773042, None) [2 3 5 7 8 9]\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-8-91a025d6193d>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m    142\u001b[0m             \u001b[0;32melif\u001b[0m \u001b[0mLP_algorithm\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;34m'LS_rbf'\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    143\u001b[0m                 \u001b[0mlp_model\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlabel_propagation\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mLabelSpreading\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mgamma\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mgamma_F\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 144\u001b[0;31m             \u001b[0mlp_model\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mgrid_list_std\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlabel_train\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    145\u001b[0m             \u001b[0mpredicted_labels_F\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlp_model\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtransduction_\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0munlabeled_indices\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    146\u001b[0m             \u001b[0mpredicted_all_labels_F\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlp_model\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtransduction_\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/Downloads/ENTER/lib/python3.6/site-packages/sklearn/semi_supervised/label_propagation.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, X, y)\u001b[0m\n\u001b[1;32m    410\u001b[0m             )\n\u001b[1;32m    411\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0malpha\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 412\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0msuper\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mLabelPropagation\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    413\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    414\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/Downloads/ENTER/lib/python3.6/site-packages/sklearn/semi_supervised/label_propagation.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, X, y)\u001b[0m\n\u001b[1;32m    270\u001b[0m             \u001b[0ml_previous\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlabel_distributions_\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    271\u001b[0m             self.label_distributions_ = safe_sparse_dot(\n\u001b[0;32m--> 272\u001b[0;31m                 graph_matrix, self.label_distributions_)\n\u001b[0m\u001b[1;32m    273\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    274\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_variant\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;34m'propagation'\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/Downloads/ENTER/lib/python3.6/site-packages/sklearn/utils/extmath.py\u001b[0m in \u001b[0;36msafe_sparse_dot\u001b[0;34m(a, b, dense_output)\u001b[0m\n\u001b[1;32m    138\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mret\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    139\u001b[0m     \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 140\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ma\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    141\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    142\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x324 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x324 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "random_seed = 46\n",
    "np.random.seed(random_seed)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "for episode in range(total_episodes):\n",
    "    random_list = np.random.permutation(range(len(grid_list))) \n",
    "    \n",
    "    \n",
    "    for trial in range(initial_random):\n",
    "        if trial == 0:\n",
    "            labeled_indices = random_list[0:1]\n",
    "            unlabeled_indices = np.array([i for i in range(len(grid_list)) if i not in labeled_indices])\n",
    "        else:\n",
    "            label_train = np.copy(grid_list[:,2])\n",
    "            label_train[unlabeled_indices] = -1\n",
    "            lp_LS_model = label_propagation.LabelSpreading()\n",
    "            lp_LS_model.fit(grid_list_std, label_train)\n",
    "            predicted_LS_all_labels = lp_LS_model.transduction_\n",
    "            true_labels = np.array(grid_list[:,2])[unlabeled_indices]\n",
    "\n",
    "            macro_report = precision_recall_fscore_support(np.copy(grid_list[:,2]), predicted_LS_all_labels, average=\"macro\")\n",
    "\n",
    "    \n",
    "            if use_cross_selection:\n",
    "                cand_indices = cross_selection(labeled_indices[-1], unlabeled_indices, grid_list, system)\n",
    "                next_index = [np.random.permutation(cand_indices)[0]]\n",
    "                labeled_indices = np.concatenate([labeled_indices, np.array(next_index)])\n",
    "                unlabeled_indices = np.array([i for i in range(len(grid_list)) if i not in labeled_indices])\n",
    "            else:\n",
    "                next_index = [np.random.permutation(unlabeled_indices)[0]]\n",
    "                labeled_indices = np.concatenate([labeled_indices, np.array(next_index)])\n",
    "                unlabeled_indices = np.array([i for i in range(len(grid_list)) if i not in labeled_indices])\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "    f1_score_list = []\n",
    "\n",
    "    for trial in range(total_trials - initial_random+1):\n",
    "        print('trial:', trial+initial_random)\n",
    "\n",
    "\n",
    "        label_train = np.copy(grid_list[:,2])\n",
    "        label_train[unlabeled_indices] = -1\n",
    "\n",
    "            \n",
    "        if LP_algorithm == 'LS_rbf':\n",
    "            lp_model = label_propagation.LabelSpreading()\n",
    "        elif LP_algorithm == 'LS_knn':\n",
    "            lp_model = label_propagation.LabelSpreading(kernel='knn')\n",
    "        elif LP_algorithm == 'LP':\n",
    "            lp_model = label_propagation.LabelPropagation()\n",
    "            \n",
    "            \n",
    "        \n",
    "        if LP_algorithm in ['LS_rbf', 'LS_knn', 'LP'] :\n",
    "            lp_model.fit(grid_list_std, label_train)\n",
    "            predicted_labels = lp_model.transduction_[unlabeled_indices]\n",
    "            predicted_all_labels = lp_model.transduction_\n",
    "            label_distributions = lp_model.label_distributions_[unlabeled_indices]\n",
    "            label_distributions_all = lp_model.label_distributions_\n",
    "            classes = lp_model.classes_\n",
    "        \n",
    "        \n",
    "        print(classification_report(np.copy(grid_list[:,2]), predicted_all_labels))\n",
    "        macro_report = precision_recall_fscore_support(np.copy(grid_list[:,2]), predicted_all_labels, average=\"macro\")\n",
    "        print('macro_report', macro_report, classes)\n",
    "        \n",
    "       \n",
    "\n",
    "\n",
    "\n",
    "        if US_strategy == 'E':\n",
    "            pred_entropies = stats.distributions.entropy(label_distributions.T)\n",
    "            u_score_list = pred_entropies/np.max(pred_entropies)\n",
    "            if use_cross_selection:\n",
    "                cand_indices = cross_selection(labeled_indices[-1], unlabeled_indices, grid_list, system)\n",
    "                pred_entropies_all = stats.distributions.entropy(label_distributions_all.T)\n",
    "                cand_E_list = []\n",
    "                for i in cand_indices:\n",
    "                    cand_E_list.append(pred_entropies_all[i])\n",
    "                uncertainty_index = [cand_indices[np.argmax(cand_E_list)]]\n",
    "            else:\n",
    "                uncertainty_index = [unlabeled_indices[np.argmax(pred_entropies)]]\n",
    "            \n",
    "            \n",
    "        elif US_strategy == 'LC':\n",
    "\n",
    "            u_score_list = 1- np.max(label_distributions, axis = 1)\n",
    "            if use_cross_selection:\n",
    "                cand_indices = cross_selection(labeled_indices[-1], unlabeled_indices, grid_list, system)\n",
    "                cand_LC_list = []\n",
    "                for i in cand_indices:\n",
    "                    cand_LC_list.append(label_distributions_all[i])\n",
    "                \n",
    "                uncertainty_index = [cand_indices[np.argmax(1- np.max(np.array(cand_LC_list), axis = 1))]]\n",
    "            else:\n",
    "                uncertainty_index = [unlabeled_indices[np.argmax(1- np.max(label_distributions, axis = 1))]]\n",
    "            \n",
    "        elif US_strategy == 'MS':\n",
    "\n",
    "            u_score_list = []\n",
    "            for pro_dist in label_distributions:\n",
    "                pro_ordered = np.sort(pro_dist)[::-1]\n",
    "                margin = pro_ordered[0] - pro_ordered[1]\n",
    "                u_score_list.append(margin)\n",
    "\n",
    "            if use_cross_selection:\n",
    "                cand_indices = cross_selection(labeled_indices[-1], unlabeled_indices, grid_list, system)\n",
    "                cand_margin_list = []\n",
    "                for i in cand_indices:\n",
    "                    pro_ordered = np.sort(label_distributions_all[i])[::-1]\n",
    "                    margin = pro_ordered[0] - pro_ordered[1]\n",
    "                    cand_margin_list.append(margin)\n",
    "                uncertainty_index = [cand_indices[np.argmin(cand_margin_list)]]    \n",
    "            else:\n",
    "                uncertainty_index = [unlabeled_indices[np.argmin(u_score_list)]]\n",
    "            u_score_list = 1-  np.array(u_score_list)\n",
    "        elif US_strategy == 'RS':\n",
    "            if use_cross_selection:\n",
    "                cand_indices = cross_selection(labeled_indices[-1], unlabeled_indices, grid_list, system)\n",
    "                uncertainty_index = [np.random.permutation(cand_indices)[0]]\n",
    "            else:\n",
    "                uncertainty_index = [np.random.permutation(unlabeled_indices)[0]]\n",
    "            u_score_list = [0.5 for i in range(len(label_distributions))]\n",
    "            \n",
    "        \n",
    "        \n",
    "            \n",
    "            \n",
    "\n",
    "        if is_output_img and (trial+initial_random) % 1 == 0 :\n",
    "            if system == 'SiO2-Al2O3-MgO':\n",
    "                gamma_F = 75\n",
    "\n",
    "            if LP_algorithm == 'LP':\n",
    "                lp_model = label_propagation.LabelPropagation(gamma=gamma_F) #H2O_detailの時gamma=25\n",
    "            elif LP_algorithm == 'LS_rbf':\n",
    "                lp_model = label_propagation.LabelSpreading(gamma=gamma_F)\n",
    "            lp_model.fit(grid_list_std, label_train)\n",
    "            predicted_labels_F = lp_model.transduction_[unlabeled_indices]\n",
    "            predicted_all_labels_F = lp_model.transduction_\n",
    "            label_distributions_F = lp_model.label_distributions_[unlabeled_indices]\n",
    "            label_distributions_all_F = lp_model.label_distributions_\n",
    "            classes = lp_model.classes_\n",
    "            print(\"classes\", classes)\n",
    "            \n",
    "            \n",
    "            macro_report_F = precision_recall_fscore_support(np.copy(grid_list[:,2]), predicted_all_labels_F, average=\"macro\")\n",
    "\n",
    "            \n",
    "            plt.rcParams['xtick.direction'] = 'in'\n",
    "            plt.rcParams['ytick.direction'] = 'in'\n",
    "            \n",
    "            plt.figure(figsize=(12, 4.5))\n",
    "            plt.rcParams[\"font.size\"] = 13\n",
    "            ax = plt.subplot(131)\n",
    "            count=0\n",
    "            init_num = 9 if initial_random == -1 else initial_random \n",
    "            for i in labeled_indices:\n",
    "                print(grid_list[i])\n",
    "                if count < init_num:\n",
    "                    plt.scatter(grid_list[i][1], grid_list[i][0], s =18, c=color_list[grid_list[i][2]], marker=\"^\")\n",
    "                else:\n",
    "                    plt.scatter(grid_list[i][1], grid_list[i][0], s =18, c=color_list[grid_list[i][2]])\n",
    "                count+=1\n",
    "\n",
    "            plt.xlim(temp_range[0], temp_range[1])\n",
    "            plt.ylim(prs_range[0], prs_range[1])\n",
    "            plt.xlabel(xlabel)\n",
    "            plt.ylabel(ylabel)\n",
    "\n",
    "            ax.tick_params(labelbottom=\"off\",bottom=\"off\")\n",
    "            ax.tick_params(labelleft=\"off\",left=\"off\")\n",
    "            plt.gca().spines['right'].set_visible(False)\n",
    "            plt.gca().spines['top'].set_visible(False)\n",
    "            plt.gca().spines['left'].set_visible(False)\n",
    "            plt.gca().spines['bottom'].set_visible(False)\n",
    "            x = np.linspace(-100,10000,4)  \n",
    "            y = 0*x -150         \n",
    "            plt.plot(x-120 ,y,color='black')\n",
    "            y = 2*1.732*x-90              \n",
    "            plt.plot(x- 120  ,y,color='black')\n",
    "            y = 2*1.732*10000 - 2*1.732*x -90           \n",
    "            plt.plot(x- 120 ,y,color='black')\n",
    "            plt.xlim(temp_range[0]-120, temp_range[1]-120)\n",
    "            plt.ylim(prs_range[0]-150, prs_range[1]-90)\n",
    "\n",
    "  \n",
    "                \n",
    "            plt.title('Checked points ('+str(trial+initial_random)+')', size = 16)\n",
    "\n",
    "            x = np.arange(0, 300)/300.\n",
    "            x = temp_range[0]+(temp_range[1] - temp_range[0])*x\n",
    "            y = np.arange(0, 300)/300.\n",
    "            y = prs_range[0]+(prs_range[1] - prs_range[0])*y\n",
    "            \n",
    "            mesh_list = []\n",
    "            for yy in y:\n",
    "                for xx in x:\n",
    "                    mesh_list.append([yy, xx])\n",
    "            pred_label = lp_model.predict(ss.transform(mesh_list))\n",
    "     \n",
    "            X, Y = np.meshgrid(x, y)\n",
    "            Z = np.reshape(pred_label, (len(y), len(x)))\n",
    "            z_min = 0\n",
    "            z_max = phase_class\n",
    "\n",
    "           \n",
    "            fig = plt.subplot(132)\n",
    "\n",
    "            u_score_colors = cm.Greens(u_score_list)\n",
    "            \n",
    "            for i in range(len(grid_list[unlabeled_indices])):\n",
    "                if i% 1 == 0:\n",
    "                    \n",
    "                    plt.scatter(grid_list[unlabeled_indices][i][1], grid_list[unlabeled_indices][i][0], c= u_score_colors[i], s=16)\n",
    "                    \n",
    "            plt.colorbar\n",
    "            plt.xlim(temp_range[0], temp_range[1])\n",
    "            plt.ylim(prs_range[0], prs_range[1])\n",
    "            plt.xlabel(xlabel)\n",
    "            plt.ylabel(ylabel)\n",
    "            if system == 'SiO2-Al2O3-MgO':\n",
    "                fig.tick_params(labelbottom=\"off\",bottom=\"off\")\n",
    "                fig.tick_params(labelleft=\"off\",left=\"off\")\n",
    "                plt.gca().spines['right'].set_visible(False)\n",
    "                plt.gca().spines['top'].set_visible(False)\n",
    "                plt.gca().spines['left'].set_visible(False)\n",
    "                plt.gca().spines['bottom'].set_visible(False)\n",
    "                x = np.linspace(-100,10000,4)  # xの値域(0, 1, 2, 3)\n",
    "                y = 0*x -150         # 直線の式\n",
    "                plt.plot(x-120 ,y,color='black')\n",
    "                y = 2*1.732*x-90              # 直線の式\n",
    "                plt.plot(x- 120  ,y,color='black')\n",
    "                y = 2*1.732*10000 - 2*1.732*x -90           # 直線の式\n",
    "                plt.plot(x- 120 ,y,color='black')\n",
    "                plt.xlim(temp_range[0]-120, temp_range[1]-120)\n",
    "                plt.ylim(prs_range[0]-150, prs_range[1]-90)\n",
    "            plt.title('Uncertainty score', size = 16)\n",
    "            \n",
    "            ax = plt.subplot(133)\n",
    "            \"\"\"\n",
    "            plt.imshow(Z, cmap='RdBu', vmin=z_min, vmax=z_max,shape = (500, 500),\n",
    "                       #extent=[X.min(), X.max(), Y.min(), Y.max()*100], #extent=[X.min()*10, X.max()*10, Y.min(), Y.max()],\n",
    "                       interpolation='nearest', origin='lower')\n",
    "            \"\"\"\n",
    "            #print('Z', Z)\n",
    "            #print('Z/float(phase_class+10)',Z/float(phase_class+10))\n",
    "            plt.pcolormesh(X, Y, Z, cmap='rainbow', vmin=0, vmax=phase_class-1)\n",
    "            #plt.title('Estimated phase diagram')\n",
    "            plt.xlabel(xlabel)\n",
    "            plt.ylabel(ylabel)\n",
    "            plt.xlim(temp_range[0], temp_range[1])\n",
    "            plt.ylim(prs_range[0], prs_range[1])\n",
    "            if system == 'SiO2-Al2O3-MgO':\n",
    "                ax.tick_params(labelbottom=\"off\",bottom=\"off\")\n",
    "                ax.tick_params(labelleft=\"off\",left=\"off\")\n",
    "                plt.gca().spines['right'].set_visible(False)\n",
    "                plt.gca().spines['top'].set_visible(False)\n",
    "                plt.gca().spines['left'].set_visible(False)\n",
    "                plt.gca().spines['bottom'].set_visible(False)\n",
    "                x = np.linspace(0,10000,4)  \n",
    "                y = 0*x -5             #\n",
    "                plt.plot(x- 15  ,y,color='black')\n",
    "                y = 2*1.732*x -5               \n",
    "                plt.plot(x- 15  ,y,color='black')\n",
    "                y = 2*1.732*10000 - 2*1.732*x -5              \n",
    "                plt.plot(x- 15  ,y,color='black')\n",
    "            \n",
    "                x = np.arange(-100,5000,1)\n",
    "                y1 = 2*1.732*x \n",
    "                y2 = 0*x+20000\n",
    "                plt.fill_between(x,y1,y2,facecolor='white')\n",
    "                x = np.arange(5000,10000,1)\n",
    "                y1 = 10000*2*1.732 - 2*1.732*x \n",
    "                y2 = 0*x+20000\n",
    "                plt.fill_between(x,y1,y2,facecolor='white')\n",
    "            \n",
    "            \n",
    "            plt.title('Estimated (Macro-F1: '+\"{0:.2f}\".format(macro_report[2])+')', size = 16)\n",
    "            \n",
    "\n",
    "            plt.suptitle(fig_title, fontsize=20)\n",
    "            plt.tight_layout()\n",
    "            plt.subplots_adjust(top=0.8)\n",
    "\n",
    "            output_img_dir = 'snapshot/'+system+'_'+LP_algorithm+'_'+US_strategy+'_CS'+str(use_cross_selection)+'_init'+str(initial_random)+'_seed'+str(random_seed)+'_trial'+str(trial)+'.png'\n",
    "            plt.savefig(output_img_dir, dpi = 300)\n",
    "        \n",
    "        labeled_indices = np.concatenate([labeled_indices, np.array(uncertainty_index)])\n",
    "        unlabeled_indices = np.array([i for i in range(len(grid_list)) if i not in labeled_indices])\n",
    "\n",
    "\n",
    "  \n",
    "        \n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "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": 1
}