Untitled.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import numpy.linalg as al\n",
    "from cvxpy import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def Mmatrix(data, nu, m1, dim1, m2, dim2, m3, dim3, m4, dim4):\n",
    "    # data is the sufficient statistic data\n",
    "    num_sam,d_num = np.shape(data)\n",
    "    # number of sample is 990\n",
    "    # number of d is 64\n",
    "#     print(num_sam)\n",
    "    \n",
    "    #m1: number of Bernoulli sample  = 8\n",
    "    #dim1: T(x) dimension of Bernoulli sample = 1\n",
    "    \n",
    "    #m2: number of Gaussian sample = 8\n",
    "    #dim2: T(x) dimension of Gaussian sample = 1\n",
    "        \n",
    "    #m3: number of Gamma sample = 8\n",
    "    #dim3: T(x) dimension of Gamma sample = 2\n",
    "    \n",
    "    #m4: number of Dirichlet sample  = 8 \n",
    "     #dim4: T(x) dimension of Dirichlet sample = 3\n",
    "\n",
    "    endpoint1 = m1 * dim1 # 8\n",
    "    \n",
    "    ber_data = data[:,0:endpoint1] # index 0 to 7 are Bernoulli\n",
    "    #print(ber_data.shape)\n",
    "    \n",
    "    endpoint2 = endpoint1 + m2 * 2 * dim2 # 8 + 16 = 24\n",
    "\n",
    "    # for gaussian case, we only need x part in [x x^2]\n",
    "    gauss_data = np.zeros(np.shape(ber_data))\n",
    "    k=0;\n",
    "    for i in range(endpoint1, endpoint2, 2): # i in range(8,24,2)\n",
    "        gauss_data[:,k] = data[:,i]\n",
    "        k=k+1\n",
    "    \n",
    "    #print(gauss_data.shape)\n",
    "    endpoint3 = endpoint2 + m3 * dim3 # 24 + 8*2 = 40\n",
    "\n",
    "    gamma_data = data[:,endpoint2:endpoint3] # index 24 to 40 are Gamma\n",
    "    #print(gamma_data.shape)\n",
    "    endpoint4 = endpoint3 + 2*m4 * dim4 # 40 + 8*3 = 64\n",
    "    \n",
    "    # the rest of them are dirichlet distribution\n",
    "    dirich_data = data[:,endpoint3:endpoint4]\n",
    "    #print(dirich_data.shape)\n",
    "    \n",
    "    vec = np.ones([990,1])\n",
    "    # construct new data\n",
    "    new_data = np.concatenate((vec, ber_data, gauss_data, gamma_data, dirich_data), axis=1)\n",
    "    #print(new_data.shape)\n",
    "    \n",
    "    M = np.transpose(np.matrix(new_data))*np.matrix(new_data)/num_sam\n",
    "    \n",
    "    return M\n",
    "def Dmatrix(m1, dim1, m2, dim2, m3, dim3, m4, dim4):\n",
    "    #m1: number of Bernoulli sample  = 8\n",
    "    #dim1: T(x) dimension of Bernoulli sample = 1\n",
    "    \n",
    "    #m2: number of Gaussian sample = 8\n",
    "    #dim2: T(x) dimension of Gaussian sample = 1\n",
    "        \n",
    "    #m3: number of Gamma sample = 8\n",
    "    #dim3: T(x) dimension of Gamma sample = 2\n",
    "    \n",
    "    #m4: number of Dirichlet sample  = 8 \n",
    "     #dim4: T(x) dimension of Dirichlet sample = 3\n",
    "    \n",
    "    # only Bernoulli has value. The rest are continuous random variable\n",
    "    l0 = 0\n",
    "    l1 = np.ones(dim1*m1)\n",
    "    l2 = np.zeros(dim2*m2)\n",
    "    l3 = np.zeros(dim3*m3)\n",
    "    l4 = np.zeros(dim4*m4)\n",
    "    \n",
    "    #l1 = np.diag(temp);\n",
    "    temp = np.hstack((l0,l1,l2,l3,l4)) # makes an array in the form of [ l0 l1 l2 l3 l4 ] \n",
    "    D = np.diag(temp)\n",
    "    return D"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test for\n",
      "sufficient_statistics_1.csv\n",
      "()\n",
      "modifiedEdge_1.csv\n",
      "()\n"
     ]
    }
   ],
   "source": [
    "k = [1]\n",
    "i = 0\n",
    "file_name = 'sufficient_statistics_'+str(k[i])+'.csv'\n",
    "file_name2 = 'modifiedEdge_'+str(k[i])+'.csv'\n",
    "\n",
    "print('test for')\n",
    "print(file_name)\n",
    "print()\n",
    "print(file_name2)\n",
    "print()\n",
    "\n",
    "\n",
    "# number of Bernoulli sample, Bernoulli T(x) dimension\n",
    "# number of Gamma sample, Gamma T(x) dimension\n",
    "# number of Gaussian sample, Gaussian T(x) dimension\n",
    "# number of Dirichlet sample, Dirichlet T(x) dimension\n",
    "# respectively\n",
    "\n",
    "m_r = [1, 1, 2, 3] # the dimension of node potential for 4 type of nodes\n",
    "n_t = [8, 8, 8, 8] # the number of nodes for each type\n",
    "numOfType = len(m_r)\n",
    "if len(m_r) == len(n_t):\n",
    "    d = np.sum(np.multiply(m_r, n_t)) # the dimension of augmented edge paramter matrix\n",
    "else:\n",
    "    print('m_r and n_t should have the same length')\n",
    "# n_r[0], m_t[0], n_t[1], m_r[1], n_r[2], m_t[2], n_r[3], m_t[3]\n",
    "data = np.genfromtxt(file_name, delimiter=\",\")\n",
    "M = Mmatrix(data, 1, 8, 1, 8, 1, 8, 2, 8, 3)\n",
    "D = Dmatrix(8, 1, 8, 1, 8, 2, 8, 3) \n",
    "A = M+D\n",
    "A = np.matrix(A) #+ np.identity(57)\n",
    "A = (A + np.transpose(A))/2\n",
    "\n",
    "# # Generate A for a test\n",
    "# #A = np.random.randn(57,57)\n",
    "\n",
    "# A = (A + np.transpose(A))/2\n",
    "\n",
    "# Lamvec, Vec = la.eig(A)\n",
    "# min_lamb = min(Lamvec)\n",
    "# A = A + np.identity(d+1)*(-1*(min_lamb) + 1)    \n",
    "# Lamvec,Vec = la.eig(A)\n",
    "# min_lamb = min(Lamvec)\n",
    "# if min_lamb> 0:\n",
    "#     print('PSD! with min_lamb',min_lamb)\n",
    "# else:\n",
    "#     print('Negatve Definite with min_lamb', min_lamb)\n",
    "# print (A)\n",
    "    \n",
    "# #np.savetxt('A.csv', A, delimiter=',')\n",
    "\n",
    "# ### initialize Theta ###\n",
    "# theta = np.ones([57, 57])*0.1  # theta is a 57 x 57 matrix with each entries with 0.1\n",
    "\n",
    "# #print(theta)\n",
    "\n",
    "# Z = np.ones([57,57])*0.2 #57x57 identity matrix ------- when I made this matrix to 0 matrix, in theta update, it calculates (Z - U) as (0 mat - 0 mat) gives error\n",
    "# U = np.ones([57, 57])*0.1 #0 matirx\n",
    "# K = 600 # maximum number of ADMM iteration"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "A =  [[ 1.    0.39  0.71  0.46  0.34  0.4   0.31  0.59  0.69  0.45  0.54  0.52\n",
      "   0.53  0.49  0.51  0.61  0.49]\n",
      " [ 0.39  1.39  0.26  0.19  0.14  0.15  0.11  0.22  0.27  0.17  0.2   0.22\n",
      "   0.2   0.19  0.2   0.22  0.19]\n",
      " [ 0.71  0.26  1.71  0.33  0.25  0.28  0.22  0.42  0.49  0.32  0.39  0.37\n",
      "   0.39  0.34  0.37  0.44  0.36]\n",
      " [ 0.46  0.19  0.33  1.46  0.16  0.2   0.13  0.26  0.33  0.21  0.27  0.25\n",
      "   0.27  0.24  0.25  0.3   0.24]\n",
      " [ 0.34  0.14  0.25  0.16  1.34  0.15  0.12  0.22  0.25  0.17  0.2   0.21\n",
      "   0.2   0.2   0.2   0.24  0.2 ]\n",
      " [ 0.4   0.15  0.28  0.2   0.15  1.42  0.14  0.26  0.3   0.2   0.25  0.24\n",
      "   0.25  0.22  0.25  0.28  0.24]\n",
      " [ 0.31  0.11  0.22  0.13  0.12  0.14  1.33  0.22  0.24  0.18  0.21  0.2\n",
      "   0.2   0.21  0.23  0.26  0.22]\n",
      " [ 0.59  0.22  0.42  0.26  0.22  0.26  0.22  1.62  0.45  0.31  0.38  0.36\n",
      "   0.37  0.36  0.37  0.45  0.37]\n",
      " [ 0.69  0.27  0.49  0.33  0.25  0.3   0.24  0.45  1.73  0.38  0.42  0.42\n",
      "   0.43  0.42  0.43  0.51  0.43]\n",
      " [ 0.45  0.17  0.32  0.21  0.17  0.2   0.18  0.31  0.38  0.53  0.3   0.3\n",
      "   0.33  0.31  0.33  0.36  0.33]\n",
      " [ 0.54  0.2   0.39  0.27  0.2   0.25  0.21  0.38  0.42  0.3   0.64  0.36\n",
      "   0.38  0.37  0.38  0.46  0.41]\n",
      " [ 0.52  0.22  0.37  0.25  0.21  0.24  0.2   0.36  0.42  0.3   0.36  0.63\n",
      "   0.38  0.36  0.39  0.44  0.4 ]\n",
      " [ 0.53  0.2   0.39  0.27  0.2   0.25  0.2   0.37  0.43  0.33  0.38  0.38\n",
      "   0.64  0.39  0.39  0.47  0.42]\n",
      " [ 0.49  0.19  0.34  0.24  0.2   0.22  0.21  0.36  0.42  0.31  0.37  0.36\n",
      "   0.39  0.6   0.4   0.46  0.43]\n",
      " [ 0.51  0.2   0.37  0.25  0.2   0.25  0.23  0.37  0.43  0.33  0.38  0.39\n",
      "   0.39  0.4   0.65  0.49  0.43]\n",
      " [ 0.61  0.22  0.44  0.3   0.24  0.28  0.26  0.45  0.51  0.36  0.46  0.44\n",
      "   0.47  0.46  0.49  0.86  0.49]\n",
      " [ 0.49  0.19  0.36  0.24  0.2   0.24  0.22  0.37  0.43  0.33  0.41  0.4\n",
      "   0.42  0.43  0.43  0.49  0.69]]\n",
      "inv(A) =  [[ 4.03 -0.33 -0.59 -0.31 -0.19 -0.23 -0.11 -0.31 -0.37 -0.68 -0.83 -0.7\n",
      "  -0.57 -0.29 -0.19 -0.34  0.33]\n",
      " [-0.33  0.81  0.01 -0.01 -0.01  0.    0.    0.01  0.    0.01  0.01 -0.04\n",
      "   0.02 -0.02 -0.    0.04 -0.01]\n",
      " [-0.59  0.01  0.83 -0.   -0.01  0.   -0.   -0.    0.    0.02 -0.01  0.03\n",
      "  -0.03  0.04 -0.01 -0.01 -0.02]\n",
      " [-0.31 -0.01 -0.    0.8  -0.   -0.01  0.01  0.01 -0.    0.01 -0.02 -0.01\n",
      "  -0.04 -0.01 -0.02 -0.02  0.01]\n",
      " [-0.19 -0.01 -0.01 -0.    0.82 -0.   -0.   -0.   -0.   -0.02 -0.   -0.03\n",
      "   0.01 -0.03 -0.03 -0.02 -0.02]\n",
      " [-0.23  0.    0.   -0.01 -0.    0.8  -0.01 -0.   -0.01  0.01 -0.04 -0.03\n",
      "  -0.03  0.02 -0.05 -0.01 -0.03]\n",
      " [-0.11  0.   -0.    0.01 -0.   -0.01  0.83 -0.01 -0.   -0.03 -0.01 -0.01\n",
      "   0.01 -0.04 -0.06 -0.06 -0.05]\n",
      " [-0.31  0.01 -0.    0.01 -0.   -0.   -0.01  0.8  -0.   -0.04 -0.04 -0.02\n",
      "  -0.02 -0.06 -0.02 -0.07 -0.05]\n",
      " [-0.37  0.    0.   -0.   -0.   -0.01 -0.   -0.    0.82 -0.1  -0.02 -0.04\n",
      "  -0.02 -0.06 -0.05 -0.05 -0.04]\n",
      " [-0.68  0.01  0.02  0.01 -0.02  0.01 -0.03 -0.04 -0.1   3.61 -0.07 -0.08\n",
      "  -0.43 -0.27 -0.39 -0.05 -0.4 ]\n",
      " [-0.83  0.01 -0.01 -0.02 -0.   -0.04 -0.01 -0.04 -0.02 -0.07  3.51 -0.11\n",
      "  -0.2  -0.36 -0.2  -0.35 -0.63]\n",
      " [-0.7  -0.04  0.03 -0.01 -0.03 -0.03 -0.01 -0.02 -0.04 -0.08 -0.11  3.45\n",
      "  -0.34 -0.26 -0.41 -0.27 -0.52]\n",
      " [-0.57  0.02 -0.03 -0.04  0.01 -0.03  0.01 -0.02 -0.02 -0.43 -0.2  -0.34\n",
      "   3.75 -0.51 -0.16 -0.47 -0.53]\n",
      " [-0.29 -0.02  0.04 -0.01 -0.03  0.02 -0.04 -0.06 -0.06 -0.27 -0.36 -0.26\n",
      "  -0.51  4.11 -0.6  -0.4  -0.8 ]\n",
      " [-0.19 -0.   -0.01 -0.02 -0.03 -0.05 -0.06 -0.02 -0.05 -0.39 -0.2  -0.41\n",
      "  -0.16 -0.6   3.69 -0.66 -0.59]\n",
      " [-0.34  0.04 -0.01 -0.02 -0.02 -0.01 -0.06 -0.07 -0.05 -0.05 -0.35 -0.27\n",
      "  -0.47 -0.4  -0.66  2.93 -0.4 ]\n",
      " [ 0.33 -0.01 -0.02  0.01 -0.02 -0.03 -0.05 -0.05 -0.04 -0.4  -0.63 -0.52\n",
      "  -0.53 -0.8  -0.59 -0.4   3.63]]\n",
      "status: optimal\n",
      "optimal value 9.35965128344\n",
      "9.35965128344 [[ 3.82 -0.31 -0.56 -0.29 -0.18 -0.22 -0.11 -0.29 -0.35 -0.62 -0.76 -0.65\n",
      "  -0.53 -0.27 -0.18 -0.33  0.28]\n",
      " [-0.31  0.81  0.01 -0.01 -0.01  0.    0.    0.    0.    0.    0.01 -0.04\n",
      "   0.01 -0.02 -0.    0.04 -0.  ]\n",
      " [-0.56  0.01  0.83 -0.   -0.01  0.   -0.   -0.   -0.    0.01 -0.02  0.02\n",
      "  -0.04  0.04 -0.02 -0.02 -0.01]\n",
      " [-0.29 -0.01 -0.    0.8  -0.   -0.01  0.01  0.01 -0.01  0.01 -0.03 -0.01\n",
      "  -0.05 -0.01 -0.02 -0.02  0.01]\n",
      " [-0.18 -0.01 -0.01 -0.    0.82 -0.   -0.   -0.   -0.   -0.03 -0.01 -0.03\n",
      "   0.01 -0.03 -0.03 -0.02 -0.01]\n",
      " [-0.22  0.    0.   -0.01 -0.    0.81 -0.01 -0.   -0.01  0.01 -0.04 -0.04\n",
      "  -0.03  0.02 -0.05 -0.01 -0.02]\n",
      " [-0.11  0.   -0.    0.01 -0.   -0.01  0.83 -0.01 -0.   -0.03 -0.01 -0.02\n",
      "   0.01 -0.04 -0.06 -0.06 -0.05]\n",
      " [-0.29  0.   -0.    0.01 -0.   -0.   -0.01  0.81 -0.   -0.04 -0.05 -0.02\n",
      "  -0.02 -0.06 -0.02 -0.07 -0.04]\n",
      " [-0.35  0.   -0.   -0.01 -0.   -0.01 -0.   -0.    0.82 -0.1  -0.02 -0.05\n",
      "  -0.02 -0.06 -0.05 -0.06 -0.03]\n",
      " [-0.62  0.    0.01  0.01 -0.03  0.01 -0.03 -0.04 -0.1   3.59 -0.09 -0.09\n",
      "  -0.43 -0.27 -0.39 -0.05 -0.39]\n",
      " [-0.76  0.01 -0.02 -0.03 -0.01 -0.04 -0.01 -0.05 -0.02 -0.09  3.5  -0.13\n",
      "  -0.22 -0.36 -0.2  -0.35 -0.62]\n",
      " [-0.65 -0.04  0.02 -0.01 -0.03 -0.04 -0.02 -0.02 -0.05 -0.09 -0.13  3.45\n",
      "  -0.35 -0.27 -0.41 -0.27 -0.51]\n",
      " [-0.53  0.01 -0.04 -0.05  0.01 -0.03  0.01 -0.02 -0.02 -0.43 -0.22 -0.35\n",
      "   3.73 -0.5  -0.17 -0.47 -0.52]\n",
      " [-0.27 -0.02  0.04 -0.01 -0.03  0.02 -0.04 -0.06 -0.06 -0.27 -0.36 -0.27\n",
      "  -0.5   4.   -0.57 -0.38 -0.76]\n",
      " [-0.18 -0.   -0.02 -0.02 -0.03 -0.05 -0.06 -0.02 -0.05 -0.39 -0.2  -0.41\n",
      "  -0.17 -0.57  3.67 -0.66 -0.6 ]\n",
      " [-0.33  0.04 -0.02 -0.02 -0.02 -0.01 -0.06 -0.07 -0.06 -0.05 -0.35 -0.27\n",
      "  -0.47 -0.38 -0.66  2.96 -0.43]\n",
      " [ 0.28 -0.   -0.01  0.01 -0.01 -0.02 -0.05 -0.04 -0.03 -0.39 -0.62 -0.51\n",
      "  -0.52 -0.76 -0.6  -0.43  3.65]]\n",
      "status: optimal\n",
      "optimal value 58.1072399293\n",
      "58.1072399293 [[ 0.09  0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.\n",
      "   0.    0.    0.    0.  ]\n",
      " [ 0.    0.09  0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.\n",
      "   0.    0.    0.    0.  ]\n",
      " [ 0.    0.    0.09  0.    0.    0.    0.    0.    0.    0.    0.    0.    0.\n",
      "   0.    0.    0.    0.  ]\n",
      " [ 0.    0.    0.    0.09  0.    0.    0.    0.    0.    0.    0.    0.    0.\n",
      "   0.    0.    0.    0.  ]\n",
      " [ 0.    0.    0.    0.    0.09  0.    0.    0.    0.    0.    0.    0.    0.\n",
      "   0.    0.    0.    0.  ]\n",
      " [ 0.    0.    0.    0.    0.    0.09  0.    0.    0.    0.    0.    0.    0.\n",
      "   0.    0.    0.    0.  ]\n",
      " [ 0.    0.    0.    0.    0.    0.    0.09  0.    0.    0.    0.    0.    0.\n",
      "   0.    0.    0.    0.  ]\n",
      " [ 0.    0.    0.    0.    0.    0.    0.    0.09  0.    0.    0.    0.    0.\n",
      "   0.    0.    0.    0.  ]\n",
      " [ 0.    0.    0.    0.    0.    0.    0.    0.    0.09  0.    0.    0.    0.\n",
      "   0.    0.    0.    0.  ]\n",
      " [ 0.    0.    0.    0.    0.    0.    0.    0.    0.    0.1   0.    0.    0.\n",
      "   0.    0.    0.    0.  ]\n",
      " [ 0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.1   0.    0.\n",
      "   0.    0.    0.    0.  ]\n",
      " [ 0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.1\n",
      "   0.    0.    0.    0.    0.  ]\n",
      " [ 0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.\n",
      "   0.1   0.    0.    0.    0.  ]\n",
      " [ 0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.\n",
      "   0.1   0.    0.    0.  ]\n",
      " [ 0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.\n",
      "   0.    0.1   0.    0.  ]\n",
      " [ 0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.\n",
      "   0.    0.    0.09  0.  ]\n",
      " [ 0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.    0.\n",
      "   0.    0.    0.    0.1 ]]\n"
     ]
    }
   ],
   "source": [
    "# Creates a 100 by 100 positive semidefinite variable.\n",
    "np.set_printoptions(suppress=True, precision = 2)\n",
    "\n",
    "d = 16\n",
    "n = d+1\n",
    "\n",
    "AA = A[:n, :n]\n",
    "\n",
    "X = Semidef(n)\n",
    "\n",
    "# # Simple A\n",
    "# A = np.random.randn(n,n)\n",
    "# A = A + A.T\n",
    "# lamb,Q = al.eig(A)\n",
    "# A = A + np.identity(n) * (- min(min(lamb),0) + 0.001)\n",
    "\n",
    "lamb,Q = al.eig(A)\n",
    "\n",
    "\n",
    "Q = np.matrix(Q)\n",
    "print 'A = ', AA\n",
    "# print 'lamb = ',np.sort(lamb), '\\n Q = ', Q\n",
    "\n",
    "# print 'Q.T*Q = ', Q.T*Q\n",
    "\n",
    "# print 'Q*Q.T = ', Q*Q.T\n",
    "# print 'A = QlambQ.T = ', Q*np.diag(lamb)*Q.T\n",
    "print 'inv(A) = ', al.inv(AA)\n",
    "# # You can use X anywhere you would use\n",
    "# # a normal CVXPY variable.\n",
    "\n",
    "alpha = 0\n",
    "obj =  Minimize(- log_det(X) + trace(X*AA) +  alpha* norm(X,1) )# Minimize(norm(A*X) + sum_entries(A*X)) #\n",
    "constr = []\n",
    "\n",
    "prob = Problem(obj, constr)\n",
    "prob.solve()\n",
    "print \"status:\", prob.status\n",
    "print \"optimal value\", prob.value\n",
    "\n",
    "print prob.value,X.value\n",
    "\n",
    "alpha = 10\n",
    "obj =  Minimize(- log_det(X) + trace(X*AA) + alpha* norm(X,1))# Minimize(norm(A*X) + sum_entries(A*X)) #\n",
    "constr = []\n",
    "\n",
    "prob = Problem(obj, constr)\n",
    "prob.solve()\n",
    "print \"status:\", prob.status\n",
    "print \"optimal value\", prob.value\n",
    "\n",
    "print prob.value,X.value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 1  4  9 16]\n",
      "[ 13.  16.  21.  28.]\n",
      "[[ 4.606  0.     0.     0.   ]\n",
      " [ 0.     6.     0.     0.   ]\n",
      " [ 0.     0.     7.583  0.   ]\n",
      " [ 0.     0.     0.     9.292]]\n"
     ]
    }
   ],
   "source": [
    "lambvec = [1 ,2, 3, 4 ]\n",
    "eta1 = 3\n",
    "# lambvec= np.matrix(lambvec)\n",
    "print np.power(lambvec,2)\n",
    "print np.power(lambvec,2) + 4*eta1 * np.ones(len(lambvec))\n",
    "print np.diag(lambvec + np.sqrt( np.power(lambvec,2) + 4*eta1 * np.ones(len(lambvec)) )  )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 1.  1.  1.]\n"
     ]
    }
   ],
   "source": [
    "print np.ones(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "m1 = 8\n",
    "m2 = 8\n",
    "m3 = 8\n",
    "m4 = 8\n",
    "\n",
    "dim1 = 1\n",
    "dim2 = 1\n",
    "dim3 = 2\n",
    "dim4 = 3\n",
    "\n",
    "m = np.array([1, m1, m2, m3, m4])\n",
    "dim = np.array([1, dim1, dim2, dim3, dim4])\n",
    "dist_num = np.sum(m)\n",
    "\n",
    "Mvec = np.zeros([33,1])\n",
    "for i in range(0,dist_num):\n",
    "    if i < 17:\n",
    "        Mvec[i] = 1\n",
    "    elif i < 25:\n",
    "        Mvec[i] = 2\n",
    "    else:\n",
    "        Mvec[i] = 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0 1]]\n"
     ]
    }
   ],
   "source": [
    "print 0 + np.matrix([0,1])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEwCAYAAABbv6HjAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10e676bd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "\n",
    "import matplotlib as mpl\n",
    "#mpl.use('Agg')\n",
    "import matplotlib.pyplot as pl\n",
    "import numpy as np\n",
    "mpl.rcParams['ps.useafm'] = True\n",
    "mpl.rcParams['pdf.use14corefonts'] = True\n",
    "mpl.rcParams['text.usetex'] = True\n",
    "\n",
    "TPR_H1 =  [1, 0.955,  0.955,  0.955,  0.955,  0.955,  0.955,  0.952 , 0.948,  0.952 , 0.925,  0.889,  0.818 , 0.745,  0.704 , 0.702 , 0.702,  0.636 , 0.563,  0.563 , 0.563, 0]\n",
    "\n",
    "FPR_H1 =  [1 ,0.916,  0.915,  0.915 , 0.914 , 0.913 , 0.909,  0.906 , 0.906,  0.9   , 0.872,  0.798,  0.674 , 0.553,  0.51  , 0.504,  0.494,  0.371,  0.248,  0.248,  0.248 ,0]\n",
    "\n",
    "\n",
    "TPR_H =  [1, 0.955 , 0.955 , 0.955 , 0.955 , 0.955  ,0.955 , 0.952  ,0.948,  0.952 , 0.925,  0.889 , 0.818,  0.745,  0.704 , 0.702 , 0.702,  0.636 , 0.563 , 0.563 , 0.563, 0]\n",
    "\n",
    "FPR_H =  [1 ,0.916 , 0.915,  0.915 , 0.914 , 0.913 , 0.909 , 0.906 , 0.906 , 0.9   , 0.872,  0.798 , 0.674 , 0.553 , 0.51 ,  0.504  ,0.494 , 0.371  ,0.248  ,0.248 , 0.248 ,0]\n",
    "\n",
    "\n",
    "TPR_L =  [1 ,0.926  ,0.926,  0.926 , 0.925 , 0.924 , 0.919  ,0.918 , 0.914  ,0.908 , 0.885,  0.821 , 0.71,   0.607 , 0.572 , 0.568 , 0.573 , 0.451 , 0.328 , 0.328,  0.328, 0]\n",
    "\n",
    "FPR_L =  [1 ,0.909  ,0.909,  0.908,  0.908  ,0.907 , 0.903 , 0.901 , 0.899  ,0.896  ,0.87  , 0.79, 0.669 , 0.544 , 0.514  ,0.511  ,0.503  ,0.371  ,0.244  ,0.244  ,0.244 ,0]\n",
    "  \n",
    "\n",
    "FPR_H_W = [0 , 0.01, 0.1, 0.41, 0.58, 0.66, 0.71, 0.8, 0.95, 1]\n",
    "TPR_H_W = [0, 0.21, 0.42, 0.7, 0.81, 0.86, 0.89, 0.93, 0.98, 1]\n",
    "\n",
    "FPR_L_W = [0, 0.07, 0.41, 0.58, 0.68 ,0.73 , 0.8 , 0.9, 1]\n",
    "TPR_L_W= [0, 0.1,  0.46, 0.63, 0.71, 0.76, 0.82, 0.9, 1]\n",
    "\n",
    "\n",
    "TPRData = np.genfromtxt('TPR.csv', delimiter=\",\")\n",
    "FPRData = np.genfromtxt('FPR.csv', delimiter=\",\")\n",
    "\n",
    "pl.figure(figsize=(4,7))\n",
    "pl.subplot(2, 1, 1)\n",
    "pl.plot(FPR_H,TPR_H,'ro-', label = 'PE-MRF')\n",
    "pl.plot(FPR_H_W,TPR_H_W,'gs-', label = 'VS-MRF')\n",
    "pl.plot(FPRData, TPRData, 'o-')\n",
    "pl.plot([0,1],[0,1],'--')\n",
    "# pl.xlabel('FPR')\n",
    "pl.ylabel('TPR')\n",
    "pl.title('High Sparsity Structure')\n",
    "pl.legend(loc = 'lower right')\n",
    "# pl.show()\n",
    "# pl.rcParams.update({'font.size': 12})\n",
    "# pl.savefig('highSparse.png', bbox_inches = 'tight', dpi = 1000)\n",
    "\n",
    "\n",
    "# pl.subplot(2, 1, 2)\n",
    "# pl.plot(FPR_L,TPR_L,'ro-', label = 'PE-MRF')\n",
    "# pl.plot(FPR_L_W,TPR_L_W,'gs-', label = 'VS-MRF')\n",
    "# pl.plot([0,1],[0,1],'--')\n",
    "# pl.xlabel('FPR')\n",
    "# pl.ylabel('TPR')\n",
    "# pl.title('Low Sparsity Structure')\n",
    "\n",
    "\n",
    "\n",
    "# #pl.plot(FPR_H,TPR_H,'bo-')\n",
    "# #pl.plot(FPR_H1,TPR_H1,'ro-')\n",
    "# #pl.plot(FPR_H_W,TPR_H_W,'ws-')\n",
    "# pl.legend(loc = 'lower right')\n",
    "# # pl.savefig('lowSparse.png',  bbox_inches = 'tight', dpi = 1000)\n",
    "    \n",
    "# pl.savefig('error_plots.eps', format = 'eps', bbox_inches = 'tight', dpi = 1000)\n",
    "pl.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWoAAAEjCAYAAAAbj0tHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10e1fc050>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.97560976  0.97560976  0.97560976  0.97560976  0.97560976  0.97560976\n",
      "  0.92682927  0.87804878  0.70731707  0.46341463] \n",
      "[ 0.9044586   0.9044586   0.89171975  0.87473461  0.83864119  0.76433121\n",
      "  0.59023355  0.33545648  0.14437367  0.03609342]\n"
     ]
    }
   ],
   "source": [
    "TPRData = np.genfromtxt('TPR.csv', delimiter=\",\")\n",
    "FPRData = np.genfromtxt('FPR.csv', delimiter=\",\")\n",
    "\n",
    "pl.figure(figsize=(4,7))\n",
    "pl.subplot(2, 1, 1)\n",
    "pl.plot(FPR_H,TPR_H,'ro-', label = 'PE-MRF')\n",
    "pl.plot(FPR_H_W,TPR_H_W,'gs-', label = 'VS-MRF')\n",
    "pl.plot(FPRData, TPRData, 'o-')\n",
    "pl.plot([0,1],[0,1],'--')\n",
    "pl.show()\n",
    "print TPRData, '\\n', FPRData\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgUAAAFiCAYAAABmhatyAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJBEKAACAj1AAAAAkEQoAAIDv/wOQavNZ6FqYtQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10eef2550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.98507463  0.98134328  0.95522388  0.94402985  0.74626866] \n",
      "[ 0.98015873  0.97751323  0.9484127   0.91931217  0.71428571]\n"
     ]
    }
   ],
   "source": [
    "TPRData = np.genfromtxt('TPR.csv', delimiter=\",\")\n",
    "FPRData = np.genfromtxt('FPR.csv', delimiter=\",\")\n",
    "pl.plot(FPRData, TPRData, 'o-')\n",
    "\n",
    "pl.plot([0,1],[0,1],'--')\n",
    "pl.show()\n",
    "print TPRData, '\\n', FPRData\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n"
     ]
    }
   ],
   "source": [
    "print 10%2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 2.  2.  2.]\n",
      " [ 2.  2.  2.]\n",
      " [ 2.  2.  2.]] [[ 4.  4.  4.]\n",
      " [ 4.  4.  4.]\n",
      " [ 4.  4.  4.]]\n"
     ]
    }
   ],
   "source": [
    "a = np.ones([3,3])*2\n",
    "aa = np.power(a,2)\n",
    "print a, aa"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda root]",
   "language": "python",
   "name": "conda-root-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}