GMM-HMM.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from __future__ import division\n",
    "import numpy as np\n",
    "from numpy.random import choice, normal, multivariate_normal\n",
    "from scipy.cluster.vq import kmeans2\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import pickle as pkl\n",
    "import glob\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import csv\n",
    "def read_data(file_path):\n",
    "    # The read-in data should be a N*W matrix,\n",
    "    # where N is the length of the time sequences,\n",
    "    # W is the number of sensors/data features\n",
    "    i = 0\n",
    "    with open(file_path, 'r') as file:\n",
    "        reader = csv.reader(file, delimiter = ',')\n",
    "        for line in reader:\n",
    "            line = np.array(line, dtype = 'float') # str2float\n",
    "            if i == 0:\n",
    "                data = line\n",
    "            else:\n",
    "                data = np.vstack((data, line))\n",
    "            i += 1\n",
    "    return data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "'''\n",
    "Load the segments where a continuous obs car exists\n",
    "'''\n",
    "data_path = \"drive_segments/*.csv\"\n",
    "files = glob.glob(data_path)\n",
    "all_segs = []\n",
    "for name in files:\n",
    "    data = read_data(name)#[frame_id, x_obs,y_obs,dv,a_obs]\n",
    "    data[:,0] += 1149 # change the \n",
    "#     data = data[:,3:7]# [x_obs,y_obs,dv,a_obs]\n",
    "    all_segs.append(data)\n",
    "    n,d = data.shape\n",
    "    \n",
    "\n",
    "# plt.figure(1)\n",
    "# plt.plot(range(data.shape[0]),data)\n",
    "# plt.legend([r'$\\Delta x$', r'$\\Delta y$', r'$\\Delta v$', r'$\\Delta a$'])\n",
    "# plt.savefig('segment_1.png',dpi=200)\n",
    "# plt.show\n",
    "\n",
    "'''\n",
    "load the ego car data\n",
    "'''\n",
    "ego_data_path = \"ego_data_24d.csv\"\n",
    "ego_data = read_data(ego_data_path)\n",
    "ego_lane = ego_data[:,[18,19,15,16,20,21,22,23]]\n",
    "ego_data = ego_data[:,[0,9,17,14,18,19]] # x_dot_ego ,accel_x, psi_T_0, r_rate, left_dist, right_dist\n",
    "'''\n",
    "Load the MRM result\n",
    "'''\n",
    "data_name='MRM_result.pkl'\n",
    "MRM_results = pkl.load(open(data_name, 'rb'))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "num_actions = 3\n",
    "data_segments = []\n",
    "for segment in all_segs:\n",
    "    ego_lane_list = []\n",
    "    ego_car_list = []\n",
    "    ego_prob_list = []\n",
    "    obs_prob_list = []\n",
    "    violation_prob_list = []\n",
    "    for i, MRM in enumerate(MRM_results):\n",
    "        if MRM['frame_id'] < 1150: # skip parking lot\n",
    "            continue\n",
    "        if MRM['frame_id']+1 in segment[:,0]:\n",
    "            # read ego car data\n",
    "            ego_lane_data = ego_lane[MRM['frame_id'],:]\n",
    "            ego_car_data = ego_data[MRM['frame_id'],:]\n",
    "            matched_obs_idx = np.where(MRM['obs_id'] == segment[0,1])[0]\n",
    "            ego_action_prob = np.reshape(MRM['ego_probs'], (num_actions,))\n",
    "            obs_action_prob = np.reshape(MRM['obs_probs'][:,matched_obs_idx], (num_actions,))#matched_obs_idx*3:(matched_obs_idx+1)*3\n",
    "            violation_prob = np.reshape(MRM['violation_prob'][:,matched_obs_idx,:], (num_actions, num_actions))\n",
    "\n",
    "            # save to list\n",
    "            ego_lane_list.append(ego_lane_data)\n",
    "            ego_car_list.append(ego_car_data)\n",
    "            ego_prob_list.append(ego_action_prob)\n",
    "            obs_prob_list.append(obs_action_prob)\n",
    "            violation_prob_list.append(violation_prob)\n",
    "    ego_lane_list = np.array(ego_lane_list)\n",
    "    ego_car_list = np.array(ego_car_list)\n",
    "    ego_prob_list = np.array(ego_prob_list)        \n",
    "    obs_prob_list = np.array(obs_prob_list)\n",
    "    violation_prob_list = np.array(violation_prob_list)\n",
    "    \n",
    "    concat_seg = np.hstack([segment[:,3:7],ego_car_list])\n",
    "    data_segments.append(concat_seg)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "class GMM_HMM():\n",
    "    def __init__(self,num_seq, data, K=10,model='Gaussian',kmeans_init=False):\n",
    "        self.K = K # hmm state space size\n",
    "        \n",
    "        self.state_list = np.arange(self.K)\n",
    "        self.data = data # training data\n",
    "        self.num_seq = num_seq # number of sequences used to train\n",
    "        # initialize parameters\n",
    "        self.pi0 = np.ones(self.K)*1/self.K # first state prior, a (K,) array\n",
    "        \n",
    "#         if num_seq > 1:\n",
    "        self.n,self.d = data[0].shape\n",
    "        len_seq = [self.data[0].shape[0]] # length of each data sequence\n",
    "        ravel_data = self.data[0] # reshape multiple sequences in to a long sequence for GMM only\n",
    "        for i , data in enumerate(self.data):\n",
    "            if i > 0:\n",
    "                ravel_data = np.vstack([ravel_data,data])\n",
    "                len_seq = np.append(len_seq, data.shape[0])\n",
    "\n",
    "        if kmeans_init:\n",
    "            self.state = kmeans2(ravel_data, self.K)[1]\n",
    "        else:\n",
    "            self.state = choice(self.K,ravel_data.shape[0])\n",
    "        self.model = model # observation model\n",
    "        if self.model == 'Gaussian' or 'GMM':\n",
    "            # compute parameter for each cluster\n",
    "            self.mu = []\n",
    "            self.sigma = []\n",
    "            for k in range(self.K):\n",
    "                idx = np.where(self.state==k)[0] # multi variable\n",
    "                nc = len(idx)\n",
    "                if nc:# multi variable\n",
    "                    cluster = np.reshape(ravel_data[idx,:],(nc,self.d))\n",
    "                    self.mu.append(np.mean(cluster,0)) # multi variable\n",
    "                    if nc > 1:\n",
    "                        self.sigma.append(np.cov(cluster,rowvar=False)) # multi variable\n",
    "                    else:\n",
    "                        self.sigma.append(np.eye(self.d))\n",
    "                else:\n",
    "                    cluster = []\n",
    "                    self.mu.append(multivariate_normal(np.zeros(self.d), np.eye(self.d))) # multi variable\n",
    "                    self.sigma.append(np.eye(self.d)) # multi variable\n",
    "        \n",
    "        self.log_like_matrix = np.zeros([self.n, self.K])\n",
    "        \n",
    "        # transition model \n",
    "        self.N = np.zeros((self.K, self.K))\n",
    "        ctr = 0\n",
    "        for i  in range(num_seq):\n",
    "            self.data[i].shape[0]\n",
    "            ctr += 1\n",
    "            for n in range(1, len_seq[i]):\n",
    "                self.N[self.state[ctr-1], self.state[ctr]] += 1\n",
    "                ctr += 1\n",
    "        self.A = (self.N.T/(np.sum(self.N, axis=1)+1e-7)).T\n",
    "        \n",
    "        self.test_states = []\n",
    "        self.test_trans_list = []\n",
    "        \n",
    "        self.pred_hidden_state_list = []\n",
    "        \n",
    "        if self.model == 'GMM':\n",
    "            self.omega = np.ones(self.K) * 1/self.K\n",
    "            \n",
    "        self.xi_list = []\n",
    "        self.gamma_list = []\n",
    "            \n",
    "    def train(self, num_iter,converge_thresh):\n",
    "        prevQ = -1e20\n",
    "        i = 0\n",
    "        while i < num_iter:\n",
    "            gamma_list = []\n",
    "            xi_list = []\n",
    "            p_list = []\n",
    "            log_like_matrix_list = [] \n",
    "            for j, segment in enumerate(self.data):\n",
    "                log_like_matrix = self.log_likelihood_matrix(segment)\n",
    "                \n",
    "                alpha,beta = self.forward_backward_algorithm(segment,log_like_matrix)\n",
    "                '''\n",
    "                run E-step for all data sequences\n",
    "                '''\n",
    "                gamma, xi, P = self.E_step(alpha,beta,segment,log_like_matrix)\n",
    "#                 if j == 0:\n",
    "#                     gamma_list = gamma\n",
    "#                     xi_list = xi\n",
    "#                     P_list = P\n",
    "#                 else:\n",
    "                gamma_list.append(gamma)\n",
    "                xi_list.append(xi)\n",
    "                p_list.append(P)\n",
    "                log_like_matrix_list.append(log_like_matrix)\n",
    "            Q = self.check_log_likelihood(gamma_list, xi_list, log_like_matrix_list)\n",
    "\n",
    "            self.M_step(gamma_list, xi_list, p_list)\n",
    "            \n",
    "            print ('log-likelihood: ', Q)\n",
    "            if Q < prevQ:\n",
    "                print('warning: the log-likelihood is decreasing')\n",
    "            if abs(Q - prevQ) < converge_thresh:\n",
    "                print('THe log-likelihood converges')\n",
    "                return\n",
    "            prevQ = Q\n",
    "            i += 1\n",
    "    \n",
    "    def test(self,test_data):\n",
    "        log_like_matrix = self.log_likelihood_matrix(test_data)\n",
    "        self.test_states =  np.argmax(log_like_matrix,axis=1)\n",
    "        self.test_trans_list = [] #np.zeros(len(self.test_states))\n",
    "        for n in range(1,len(self.test_states)):\n",
    "            p_trans = self.A[self.test_states[n-1], self.test_states[n]]\n",
    "            self.test_trans_list = np.append(self.test_trans_list, p_trans)\n",
    "            print (\"p_trans\", p_trans)\n",
    "            \n",
    "    def predict(self, observable_data, pred_step=20):\n",
    "        '''\n",
    "        Predict the future data based on the initial observation.\n",
    "        Observable data: {x, y, yaw_angle, velocity, d_l, d_r}\n",
    "        Predict data: {acceleration, yaw_rate}\n",
    "        '''\n",
    "        num_observable = 4\n",
    "        num_hidden = 2\n",
    "        pred_start = 10\n",
    "        \n",
    "        observable_matrix = self.log_likelihood_matrix(observable_data)\n",
    "        # init beta\n",
    "        if self.model == 'GMM':\n",
    "            beta = self.omega * observable_matrix[0,:]/np.sum(self.omega * observable_matrix[0,:])\n",
    "        else:\n",
    "            beta = np.ones(self.K) * 1/self.K # init a beta\n",
    "            \n",
    "        for t in range(observable_data.shape[0]):\n",
    "            beta_den = beta * np.sum(self.A, axis=0) * observable_matrix[t,:]\n",
    "            beta_num = np.sum(beta_den)#beta * np.sum((np.sum(self.A, axis=0) * observable_matrix[t,:]))\n",
    "            beta = beta_den/beta_num\n",
    "            pred_hidden_state = np.zeros(num_hidden) \n",
    "            if t > pred_start:\n",
    "                for k in range(self.K):\n",
    "                    diff = observable_data[t] - self.mu[k][0:num_observable]\n",
    "                    tmp_mu = self.mu[k][num_observable:num_observable+num_hidden]\n",
    "                    tmp_sigma_1 = self.sigma[k][0:num_observable,num_observable:num_observable+num_hidden]\n",
    "                    tmp_sigma_2 = self.sigma[k][0:num_observable, 0:num_observable]\n",
    "                    \n",
    "                    pred_hidden_state += beta[k] * (tmp_mu + np.dot(np.dot(tmp_sigma_1.T, np.linalg.pinv(tmp_sigma_2)),diff))\n",
    "            \n",
    "                self.pred_hidden_state_list.append(pred_hidden_state)\n",
    "        \n",
    "#         log_like_init = self.log_likelihood_matrix(np.reshape(test_data[0,:], (1,self.d)))\n",
    "#         init_state =  np.argmax(log_like_init)\n",
    "        \n",
    "#         for n in range(1,tets_data.shape[0]):\n",
    "#             data = test_data[n,:]\n",
    "#             *self.A[init_state,:]\n",
    "        \n",
    "    def E_step(self,alpha,beta,segment,log_like_matrix):\n",
    "        T = segment.shape[0]\n",
    "        pX = np.reshape(np.sum(alpha*beta,axis=1),(T,1)) \n",
    "        \n",
    "        gamma = alpha*beta/pX # n * k matrix\n",
    "        \n",
    "        xi = np.zeros([T-1, self.K, self.K]) # n-1 * k * k matrix\n",
    "        for n in range(1,T):\n",
    "            \n",
    "            xi_num = (np.dot(np.reshape(alpha[n-1,:],(self.K,1)),\n",
    "                             np.reshape(np.exp(log_like_matrix[n,:]) * beta[n,:], (1,self.K))) *\\\n",
    "                                  self.A)\n",
    "            xi_den = np.sum(np.sum(xi_num,axis=1))\n",
    "            xi[n-1,:,:] = xi_num/xi_den\n",
    "        '''\n",
    "        Figure out the difference using MGD and GMM model in E-step\n",
    "        '''\n",
    "        P = self.omega * log_like_matrix / np.reshape(np.sum(self.omega * log_like_matrix, axis=1), (T,1)) # an n*K matrix\n",
    "        return gamma, xi, P\n",
    "    \n",
    "    def M_step(self, gamma_list, xi_list, p_list):\n",
    "        # update first state prior\n",
    "#         self.pi0 = self.gamma[0,:]/np.sum(self.gamma[0,:])\n",
    "        pi0_num = 0\n",
    "        for gamma in gamma_list:\n",
    "            pi0_num += gamma[0,:]\n",
    "        pi0_den = np.sum(pi0_num)\n",
    "        self.pi0 = pi0_num/pi0_den\n",
    "        \n",
    "        # update transition model\n",
    "        A_num = 0\n",
    "        for xi in xi_list:\n",
    "            A_num += np.sum(xi, axis=0)\n",
    "        A_den = np.reshape(np.sum(A_num,axis=1),(self.K,1))\n",
    "        self.A = A_num/A_den\n",
    "        \n",
    "        # update emission model\n",
    "        if self.model == 'Gaussian':\n",
    "            for k in range(self.K):\n",
    "                mu_num = np.sum(np.reshape(self.gamma[:,k],(self.n,1)) * self.data, axis=0) # (d,) matrix\n",
    "                mu_den = np.sum(self.gamma[:,k])# scalar\n",
    "                self.mu[k] = mu_num/mu_den\n",
    "\n",
    "                sigma_num = np.dot((self.data-self.mu[k]).T, np.reshape(self.gamma[:,k],(self.n,1))*(self.data-self.mu[k]))\n",
    "                sigma_den = mu_den\n",
    "                self.sigma[k] = sigma_num/sigma_den\n",
    "        if self.model == 'GMM':\n",
    "#             p_num = self.omega * self.log_like_matrix # n * K matrix\n",
    "#             P_den = np.reshape(np.sum(self.omega * self.log_like_matrix, axis=1),(self.n,1)) # n * 1 array\n",
    "#             self.P = P_num/P_den # n * K matrix\n",
    "#             self.omega = np.sum(self.P, axis=0)/self.n\n",
    "            # update omega\n",
    "#             ctr = 0\n",
    "#             omega = 0\n",
    "#             mu_num = np.zeros(self.K, self.d)\n",
    "#             mu_den = np.zeros(self.K,1)\n",
    "#             for i, gamma in enumerate(gamma_list):\n",
    "#                 T = p_list.shape[0]\n",
    "#                 ctr += T\n",
    "#                 omega += np.sum(p_list[i],axis=0)\n",
    "            P = np.concatenate(p_list)\n",
    "            all_gamma = np.concatenate(gamma_list)\n",
    "            all_data = np.concatenate(self.data)\n",
    "\n",
    "            self.omega = np.sum(P, axis=0)/P.shape[0]\n",
    "            # update mu and sigma]\n",
    "            for  k in range(self.K):\n",
    "#                 for i, p in enumerate(p_list):\n",
    "#                     mu_num[k,:] += np.sum(self.data[i] * np.reshape(P_list[i][:,k], (T,1)), axis=0)\n",
    "#                     mu_den[k] += np.sum(p_list[i][:,k])\n",
    "#                 self.mu[k] = mu_num[k,:]/mu_den[k]\n",
    "                self.mu[k] = np.sum(all_data * np.reshape(P[:,k], (P.shape[0],1)), axis=0) / np.sum(P[:,k]) # (d,) array \n",
    "                self.sigma[k] = np.dot((all_data - self.mu[k]).T, \n",
    "                                np.reshape(P[:,k], (P.shape[0],1))*(all_data-self.mu[k]))/ np.sum(P[:,k])\n",
    "#                 for i, p in enumerate(p_list):\n",
    "#                     sigma_den[k] +=\n",
    "#                 self.sigma[k] = np.dot((self.data - self.mu[k]).T, \n",
    "#                                 np.reshape(self.P[:,k], (self.n,1))*(self.data-self.mu[k]))/ np.sum(self.P[:,k])\n",
    "                \n",
    "#             self.omega = omega/ctr\n",
    "    def forward_backward_algorithm(self, data, log_like_matrix):\n",
    "        # forward\n",
    "        T = data.shape[0]\n",
    "        alpha = np.zeros((T, self.K))\n",
    "        alpha[0, :] = self.pi0*np.exp(log_like_matrix[0,:])#1\n",
    "        for t in range(1, T):\n",
    "#             if self.model == 'Gaussian':\n",
    "            alpha[t,:] = (self.A.T.dot(alpha[t-1,:])) * np.exp(log_like_matrix[t,:])\n",
    "#             if self.model == 'GMM':\n",
    "                \n",
    "#                 alpha[t,:] = (self.A.T.dot(alpha[t-1,:])) * np.exp(self.P[t,:])\n",
    "            alpha[t, :] /= np.max(alpha[t, :])\n",
    "        # backward\n",
    "        beta = np.zeros((T, self.K))\n",
    "        beta[-1, :] = 1\n",
    "        for t in range(T - 1, 0, -1):\n",
    "#             if self.model == 'Gaussian':\n",
    "            beta[t-1, :] = self.A.dot(beta[t, :] * np.exp(log_like_matrix[t,:]))\n",
    "#             if self.model == 'GMM':\n",
    "#                 beta[t-1, :] = self.A.dot(beta[t, :] * np.exp(self.P[t,:]))\n",
    "            beta[t-1, :] /= np.max(beta[t-1, :])            \n",
    "        return alpha, beta\n",
    "    \n",
    "    def log_likelihood_matrix(self,data):\n",
    "        '''\n",
    "        Likelihood matrix of all observations, each row is the likelihood of one data given all possible states.\n",
    "        Output: log-likelihood matrix with size self.n*K\n",
    "        '''\n",
    "#         if self.model == 'Gaussian':\n",
    "        N,d = data.shape\n",
    "        log_like_matrix = np.zeros([N,self.K])\n",
    "        for n in range(N):\n",
    "            for k in range(self.K):\n",
    "                diff = data[n,:]-self.mu[k][0:d]\n",
    "                sigma = self.sigma[k][0:d,0:d]\n",
    "                log_like_matrix[n,k] = (-1/2) * np.dot(np.dot(diff, np.linalg.pinv(sigma)), diff.T) - np.log(np.sqrt(np.linalg.det(2*np.pi*sigma)))\n",
    "                if np.isnan(log_like_matrix[n,k]):\n",
    "                    print (\"nan sigma: \", sigma)\n",
    "                    print (\"nan diff: \", diff)\n",
    "#         if self.mode == 'GMM':\n",
    "            \n",
    "        return log_like_matrix\n",
    "\n",
    "    def check_log_likelihood(self, gamma_list, xi_list, log_like_matrix_list):\n",
    "#         idx = np.where(self.pi0!=0)[0]\n",
    "#         term1 = np.dot(self.gamma[0,idx],np.log(self.pi0[idx]))\n",
    "#         term2 = 0\n",
    "#         idx = np.where(self.A!=0)\n",
    "#         for n in range(1,self.n):\n",
    "#             xi = self.xi[n-1,:,:]\n",
    "#             term2 += np.sum(np.sum(xi[idx]*np.log(self.A[idx])))\n",
    "#         term3 = np.sum(np.sum(self.gamma * self.log_like_matrix))\n",
    "        \n",
    "        \n",
    "        idx = np.where(self.pi0!=0)[0]\n",
    "        term1 = 0\n",
    "        for gamma in gamma_list:\n",
    "            term1 += np.dot(gamma[0,idx],np.log(self.pi0[idx]))\n",
    "        term2 = 0\n",
    "        idx = np.where(self.A!=0)\n",
    "        for xi in xi_list:\n",
    "            for n in range(1,xi.shape[0]):\n",
    "                tmp_xi = xi[n-1,:,:]\n",
    "                term2 += np.sum(np.sum(tmp_xi[idx]*np.log(self.A[idx])))\n",
    "        term3 = 0\n",
    "        for i, log_like_matrix in enumerate(log_like_matrix_list):\n",
    "            term3 += np.sum(np.sum(gamma_list[i] * log_like_matrix))\n",
    "        print (term1)\n",
    "        print (term2)\n",
    "        print (term3)\n",
    "        return term1+term2+term3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "# new_data_try = np.hstack([data,ego_car_list[:,0:3]])\n",
    "num_train = 1\n",
    "train_data = data_segments[0:num_train]\n",
    "test_data = data_segments[num_train:]\n",
    "hmm = GMM_HMM(num_train, train_data,model='GMM')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([  3.94225000e+01,  -3.67187500e+00,   4.98725000e-01,\n",
      "        -5.25000000e-02,   2.77777778e-01,   3.89950000e-01,\n",
      "         6.22558625e-03,  -1.90241325e-02,  -1.15000000e-01,\n",
      "        -6.50000000e-02]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([  3.60465000e+01,  -3.42187500e+00,   5.16725000e-01,\n",
      "        -5.25000000e-02,   2.77777778e-01,   1.00550000e-01,\n",
      "         5.73730525e-03,  -1.80641325e-02,  -1.15000000e-01,\n",
      "        -8.50000000e-02]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([  3.11095000e+01,  -2.85937500e+00,   8.80725000e-01,\n",
      "        -5.25000000e-02,   2.77777778e-01,  -2.20700000e-02,\n",
      "         4.27246125e-03,  -1.93731325e-02,  -1.35000000e-01,\n",
      "        -6.50000000e-02]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([  2.77345000e+01,  -2.60937500e+00,   9.67725000e-01,\n",
      "        -5.25000000e-02,   2.77777778e-01,   3.89950000e-01,\n",
      "         4.27246125e-03,  -1.05592325e-02,  -1.35000000e-01,\n",
      "        -6.50000000e-02]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([  2.25465000e+01,  -2.23437500e+00,   1.01972500e+00,\n",
      "        -5.25000000e-02,   2.77777778e-01,  -3.67900000e-02,\n",
      "         2.31933625e-03,  -1.50971325e-02,  -1.35000000e-01,\n",
      "        -4.50000000e-02]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([  2.03595000e+01,  -1.98437500e+00,   1.14172500e+00,\n",
      "        -5.25000000e-02,   2.77777778e-01,  -2.45000000e-03,\n",
      "         1.83105475e-03,   1.30896750e-03,  -1.15000000e-01,\n",
      "        -6.50000000e-02]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([  1.53595000e+01,  -1.60937500e+00,   1.33272500e+00,\n",
      "        -2.25000000e-02,   2.77777778e-01,  -3.67900000e-02,\n",
      "        -1.22070250e-04,   6.45766750e-03,  -7.50000000e-02,\n",
      "        -6.50000000e-02]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([  1.21725000e+01,  -1.35937500e+00,   1.29772500e+00,\n",
      "        -2.25000000e-02,   2.77777778e-01,   1.71700000e-02,\n",
      "        -6.10351750e-04,   6.63226750e-03,  -5.50000000e-02,\n",
      "        -6.50000000e-02]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([  8.42250000e+00,  -9.84375000e-01,   7.41725000e-01,\n",
      "         7.50000000e-03,  -2.50004462e-11,   2.82040000e-01,\n",
      "        -1.22070250e-04,   9.16296750e-03,  -3.50000000e-02,\n",
      "        -4.50000000e-02]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([  8.17250000e+00,  -7.34375000e-01,   4.29725000e-01,\n",
      "         7.50000000e-03,  -2.50004462e-11,  -3.18800000e-02,\n",
      "        -1.22070250e-04,   8.81386750e-03,  -3.50000000e-02,\n",
      "        -6.50000000e-02]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([  5.67250000e+00,  -2.96875000e-01,   2.38625000e-01,\n",
      "         3.75000000e-02,  -2.50004462e-11,  -1.74120000e-01,\n",
      "        -6.10351750e-04,   1.83256750e-03,  -3.50000000e-02,\n",
      "        -2.50000000e-02]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([  5.04650000e+00,  -4.68750000e-02,  -7.38750000e-02,\n",
      "         3.75000000e-02,  -2.50004462e-11,   7.11200000e-02,\n",
      "         3.66211250e-04,  -2.70523250e-03,  -1.50000000e-02,\n",
      "        -5.00000000e-03]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([  1.85950000e+00,   3.90625000e-01,  -5.94675000e-01,\n",
      "         7.50000000e-03,  -2.50004462e-11,   3.06565000e-01,\n",
      "         3.66210750e-04,  -6.19593250e-03,   2.50000000e-02,\n",
      "        -5.00000000e-03]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([ -2.64050000e+00,   4.53125000e-01,  -2.47475000e-01,\n",
      "         7.50000000e-03,  -2.50004462e-11,  -1.15260000e-01,\n",
      "        -6.10351750e-04,  -4.01423250e-03,   5.00000000e-03,\n",
      "         1.50000000e-02]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([ -8.26550000e+00,   7.03125000e-01,  -4.90475000e-01,\n",
      "         7.50000000e-03,  -2.77777778e-01,   1.44700000e-01,\n",
      "        -1.09863275e-03,  -5.93413250e-03,   6.50000000e-02,\n",
      "         1.50000000e-02]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([ -1.05155000e+01,   8.90625000e-01,  -7.85675000e-01,\n",
      "        -2.25000000e-02,  -2.77777778e-01,   1.83940000e-01,\n",
      "        -1.09863275e-03,  -5.23602500e-04,   6.50000000e-02,\n",
      "         3.50000000e-02]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([ -1.44535000e+01,   1.14062500e+00,  -1.16757500e+00,\n",
      "        -2.25000000e-02,  -2.77777778e-01,  -7.60300000e-02,\n",
      "        -6.10351250e-04,  -3.83973250e-03,   8.50000000e-02,\n",
      "         3.50000000e-02]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([ -1.77655000e+01,   1.20312500e+00,  -1.01137500e+00,\n",
      "        -2.25000000e-02,  -2.77777778e-01,   8.09300000e-02,\n",
      "         3.66210750e-04,   2.61797500e-04,   8.50000000e-02,\n",
      "         5.50000000e-02]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([ -2.33275000e+01,   1.32812500e+00,  -9.07175000e-01,\n",
      "         7.50000000e-03,  -2.77777778e-01,  -9.56400000e-02,\n",
      "         4.27246125e-03,  -1.29151325e-02,   6.50000000e-02,\n",
      "         9.50000000e-02]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([ -2.55775000e+01,   1.45312500e+00,  -8.72475000e-01,\n",
      "         1.27500000e-01,  -5.55555556e-01,  -3.16370000e-01,\n",
      "         4.27246075e-03,  -1.26541325e-02,   6.50000000e-02,\n",
      "         1.35000000e-01]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([ -3.08905000e+01,   1.64062500e+00,  -7.16175000e-01,\n",
      "         1.57500000e-01,  -5.55555556e-01,  -2.69800000e-02,\n",
      "         7.20214875e-03,  -1.35261325e-02,   8.50000000e-02,\n",
      "         1.35000000e-01]))\n",
      "('nan sigma: ', array([[  1.94655544e+02,  -1.58421146e+01,   1.29060732e+01,\n",
      "          7.45250000e-02,   3.06712963e+00,  -4.46528937e-01,\n",
      "         -1.64037874e-03,   3.04944019e-02,  -9.16976667e-01,\n",
      "         -6.80103333e-01],\n",
      "       [ -1.58421146e+01,   1.41503906e+00,  -1.15677969e+00,\n",
      "          3.59375000e-03,  -2.60416667e-01,   3.36722917e-02,\n",
      "          1.14440776e-04,  -3.92332495e-03,   7.53125000e-02,\n",
      "          5.92708333e-02],\n",
      "       [  1.29060732e+01,  -1.15677969e+00,   9.55245569e-01,\n",
      "         -3.95225000e-03,   2.17046296e-01,  -3.58705177e-02,\n",
      "         -1.36356509e-04,   3.17826357e-03,  -6.20168333e-02,\n",
      "         -4.85318333e-02],\n",
      "       [  7.45250000e-02,   3.59375000e-03,  -3.95225000e-03,\n",
      "          8.25000000e-04,  -7.50013477e-13,   2.69800000e-04,\n",
      "          1.22070750e-06,  -9.42469750e-05,  -2.50000000e-04,\n",
      "          5.00000000e-05],\n",
      "       [  3.06712963e+00,  -2.60416667e-01,   2.17046296e-01,\n",
      "         -7.50013477e-13,   5.14403292e-02,  -1.09000000e-02,\n",
      "         -4.52112037e-05,   5.73691667e-04,  -1.48148148e-02,\n",
      "         -1.11111111e-02],\n",
      "       [ -4.46528937e-01,   3.36722917e-02,  -3.58705177e-02,\n",
      "          2.69800000e-04,  -1.09000000e-02,   8.74869700e-03,\n",
      "          4.35074958e-05,   1.79646623e-05,   2.66506667e-03,\n",
      "          1.58600000e-03],\n",
      "       [ -1.64037874e-03,   1.14440776e-04,  -1.36356509e-04,\n",
      "          1.22070750e-06,  -4.52112037e-05,   4.35074958e-05,\n",
      "          2.18550466e-07,   2.55663743e-07,   1.05794183e-05,\n",
      "          5.69660833e-06],\n",
      "       [  3.04944019e-02,  -3.92332495e-03,   3.17826357e-03,\n",
      "         -9.42469750e-05,   5.73691667e-04,   1.79646623e-05,\n",
      "          2.55663743e-07,   2.17341176e-05,  -1.47188317e-04,\n",
      "         -1.50678950e-04],\n",
      "       [ -9.16976667e-01,   7.53125000e-02,  -6.20168333e-02,\n",
      "         -2.50000000e-04,  -1.48148148e-02,   2.66506667e-03,\n",
      "          1.05794183e-05,  -1.47188317e-04,   4.36666667e-03,\n",
      "          3.23333333e-03],\n",
      "       [ -6.80103333e-01,   5.92708333e-02,  -4.85318333e-02,\n",
      "          5.00000000e-05,  -1.11111111e-02,   1.58600000e-03,\n",
      "          5.69660833e-06,  -1.50678950e-04,   3.23333333e-03,\n",
      "          2.50000000e-03]]))\n",
      "('nan diff: ', array([ -3.39535000e+01,   1.70312500e+00,  -9.07175000e-01,\n",
      "         6.75000000e-02,  -5.55555556e-01,   3.99760000e-01,\n",
      "         9.64355475e-03,  -1.38751325e-02,   8.50000000e-02,\n",
      "         1.55000000e-01]))\n",
      "nan\n",
      "nan\n",
      "nan\n",
      "('log-likelihood: ', nan)\n"
     ]
    },
    {
     "ename": "LinAlgError",
     "evalue": "SVD did not converge",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mLinAlgError\u001b[0m                               Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-61-d2f28e9967a8>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0mnum_iter\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m100\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      2\u001b[0m \u001b[0mconverge_thresh\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m1e-5\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 3\u001b[0;31m \u001b[0mhmm\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrain\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnum_iter\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mconverge_thresh\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m<ipython-input-59-a07477e5a3ee>\u001b[0m in \u001b[0;36mtrain\u001b[0;34m(self, num_iter, converge_thresh)\u001b[0m\n\u001b[1;32m     75\u001b[0m             \u001b[0mlog_like_matrix_list\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     76\u001b[0m             \u001b[0;32mfor\u001b[0m \u001b[0mj\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msegment\u001b[0m \u001b[0;32min\u001b[0m \u001b[0menumerate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 77\u001b[0;31m                 \u001b[0mlog_like_matrix\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlog_likelihood_matrix\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msegment\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     78\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     79\u001b[0m                 \u001b[0malpha\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mbeta\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward_backward_algorithm\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msegment\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mlog_like_matrix\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-59-a07477e5a3ee>\u001b[0m in \u001b[0;36mlog_likelihood_matrix\u001b[0;34m(self, data)\u001b[0m\n\u001b[1;32m    268\u001b[0m                 \u001b[0mdiff\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdata\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mn\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmu\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mk\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0md\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    269\u001b[0m                 \u001b[0msigma\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msigma\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mk\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0md\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0md\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 270\u001b[0;31m                 \u001b[0mlog_like_matrix\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mn\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mk\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m/\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m*\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdiff\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlinalg\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpinv\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msigma\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdiff\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mT\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m-\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlog\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msqrt\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlinalg\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdet\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpi\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0msigma\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    271\u001b[0m                 \u001b[0;32mif\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0misnan\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlog_like_matrix\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mn\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mk\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    272\u001b[0m                     \u001b[0;32mprint\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0;34m\"nan sigma: \"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msigma\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/Users/BrianYao/anaconda2/lib/python2.7/site-packages/numpy/linalg/linalg.pyc\u001b[0m in \u001b[0;36mpinv\u001b[0;34m(a, rcond)\u001b[0m\n\u001b[1;32m   1615\u001b[0m     \u001b[0m_assertNoEmpty2d\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ma\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1616\u001b[0m     \u001b[0ma\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0ma\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mconjugate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1617\u001b[0;31m     \u001b[0mu\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0ms\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvt\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msvd\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ma\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1618\u001b[0m     \u001b[0mm\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mu\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1619\u001b[0m     \u001b[0mn\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mvt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/Users/BrianYao/anaconda2/lib/python2.7/site-packages/numpy/linalg/linalg.pyc\u001b[0m in \u001b[0;36msvd\u001b[0;34m(a, full_matrices, compute_uv)\u001b[0m\n\u001b[1;32m   1357\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1358\u001b[0m         \u001b[0msignature\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m'D->DdD'\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0misComplexType\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mt\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32melse\u001b[0m \u001b[0;34m'd->ddd'\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1359\u001b[0;31m         \u001b[0mu\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0ms\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvt\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mgufunc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ma\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msignature\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0msignature\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mextobj\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mextobj\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1360\u001b[0m         \u001b[0mu\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mu\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mastype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mresult_t\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcopy\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mFalse\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1361\u001b[0m         \u001b[0ms\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0ms\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mastype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0m_realType\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mresult_t\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcopy\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mFalse\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/Users/BrianYao/anaconda2/lib/python2.7/site-packages/numpy/linalg/linalg.pyc\u001b[0m in \u001b[0;36m_raise_linalgerror_svd_nonconvergence\u001b[0;34m(err, flag)\u001b[0m\n\u001b[1;32m     97\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     98\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0m_raise_linalgerror_svd_nonconvergence\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0merr\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mflag\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 99\u001b[0;31m     \u001b[0;32mraise\u001b[0m \u001b[0mLinAlgError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"SVD did not converge\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    100\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    101\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mget_linalg_error_extobj\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcallback\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mLinAlgError\u001b[0m: SVD did not converge"
     ]
    }
   ],
   "source": [
    "num_iter = 100\n",
    "converge_thresh = 1e-5\n",
    "hmm.train(num_iter, converge_thresh)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<function matplotlib.pyplot.show>"
      ]
     },
     "execution_count": 138,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiIAAAEYCAYAAABhvZ18AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f72de850160>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "states = np.argmax(hmm.log_like_matrix,axis=1)\n",
    "plt.figure(2)\n",
    "ax1 = plt.subplot(311)\n",
    "plt.plot(range(hmm.n),train_data)\n",
    "plt.xlim([0,hmm.n])\n",
    "ax1.set_title(\"Data\")\n",
    "# lgd = plt.legend([r'$\\Delta x$', r'$\\Delta y$', r'$v_{obs}$', r'$a_{obs}$'],bbox_to_anchor=(1.05, 1), loc=2, ncol=2, borderaxespad=0.)\n",
    "# lgd = plt.legend([r'$\\Delta x$', \n",
    "#                   r'$\\Delta y$', \n",
    "#                   r'$\\Delta v$', \n",
    "#                   r'$\\Delta a$'\n",
    "# #                   r'$v_{ego}$',\n",
    "# #                   r'$a_{ego}$',\n",
    "# #                   r'$\\psi_{ego}$'\n",
    "#                  ],bbox_to_anchor=(1.05, 1), loc=2, ncol=2, borderaxespad=0.)\n",
    "lgd = plt.legend([r'$v$', \n",
    "                  r'$\\psi$', \n",
    "                  r'$d_l$', \n",
    "                  r'$d_r$',\n",
    "                  r'$a$',\n",
    "                  r'$\\dot{\\psi}$'\n",
    "                 ],bbox_to_anchor=(1.05, 1), loc=2, ncol=2, borderaxespad=0.)\n",
    "\n",
    "ax2 = plt.subplot(312)\n",
    "sns.heatmap(np.reshape(states,(1,hmm.n)), ax=ax2, cbar=False)\n",
    "ax2.set_title(\"Hidden states\")\n",
    "\n",
    "ax3 = plt.subplot(337)\n",
    "sns.heatmap(hmm.A, ax=ax3,vmin=0, vmax=0.05, cbar=False)\n",
    "ax3.set_title(\"Transition matrix\")\n",
    "# plt.subplot(133)\n",
    "# plt\n",
    "plt.tight_layout(h_pad=1.0)\n",
    "# plt.savefig('example_1_result.png', bbox_extra_artists=(lgd,),bbox_inches='tight', dpi=200)\n",
    "plt.show"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 1.05099442716e-73\n",
      "p_trans 0.903732455448\n",
      "p_trans 0.903732455448\n",
      "p_trans 0.903732455448\n",
      "p_trans 0.903732455448\n",
      "p_trans 0.0722006875222\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 2.31288410777e-33\n",
      "p_trans 0.952641890648\n",
      "p_trans 1.54197692224e-12\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 2.31288410777e-33\n",
      "p_trans 1.54197692224e-12\n",
      "p_trans 2.31288410777e-33\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 1.54197692224e-12\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 2.31288410777e-33\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 1.54197692224e-12\n",
      "p_trans 2.31288410777e-33\n",
      "p_trans 1.54197692224e-12\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 0.965193852466\n",
      "p_trans 2.31288410777e-33\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "p_trans 0.952641890648\n",
      "[  4   9  47  49  73  74  75 145 342 348 349 350 356]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([  1.05099443e-73,   7.22006875e-02,   2.31288411e-33,\n",
       "         1.54197692e-12,   2.31288411e-33,   1.54197692e-12,\n",
       "         2.31288411e-33,   1.54197692e-12,   2.31288411e-33,\n",
       "         1.54197692e-12,   2.31288411e-33,   1.54197692e-12,\n",
       "         2.31288411e-33])"
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiIAAAEYCAYAAABhvZ18AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f72de80f358>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# new_test_try = np.hstack([test_data,ego_car_list[:,0:3]])\n",
    "hmm.test(test_data)\n",
    "plt.figure(3)\n",
    "ax1 = plt.subplot(211)\n",
    "plt.plot(range(test_data.shape[0]),test_data)\n",
    "plt.xlim([0,test_data.shape[0]])\n",
    "ax1.set_title(\"Data\")\n",
    "# lgd = plt.legend([r'$\\Delta x$', \n",
    "#                   r'$\\Delta y$', \n",
    "#                   r'$\\Delta v$', \n",
    "#                   r'$\\Delta a$'\n",
    "# #                   r'$v_{ego}$',\n",
    "# #                   r'$a_{ego}$',\n",
    "# #                   r'$\\psi_{ego}$'\n",
    "#                  ],bbox_to_anchor=(1.05, 1), loc=2, ncol=2, borderaxespad=0.)\n",
    "lgd = plt.legend([r'$v$', \n",
    "                  r'$\\psi$', \n",
    "                  r'$d_l$', \n",
    "                  r'$d_r$',\n",
    "                  r'$a$',\n",
    "                  r'$\\dot{\\psi}$'\n",
    "                 ],bbox_to_anchor=(1.05, 1), loc=2, ncol=2, borderaxespad=0.)\n",
    "ax2 = plt.subplot(212)\n",
    "sns.heatmap(np.reshape(hmm.test_states,(1,test_data.shape[0])), ax=ax2, cbar=False)\n",
    "ax2.set_title(\"Hidden states\")\n",
    "\n",
    "plt.tight_layout(h_pad=1.0)\n",
    "plt.savefig('example_2_result.png', bbox_extra_artists=(lgd,),bbox_inches='tight', dpi=200)\n",
    "plt.show\n",
    "\n",
    "state_change_idx = np.where(np.array(hmm.test_trans_list)<0.9)[0]\n",
    "print(state_change_idx)\n",
    "np.array(hmm.test_trans_list)[state_change_idx]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "hmm.predict(test_data[:,0:4], 20)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<function matplotlib.pyplot.show>"
      ]
     },
     "execution_count": 142,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXYAAAD8CAYAAABjAo9vAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzsnXl43FW5xz9n9iX71qRJ2qQLbaF7S6G0FBUUECggu4CA\nehEU5YpeUHAHFMUN70UQBUGsgFhlXwQLlNKWbnSjC23TJUnTNNskM5l9fuf+MZnJnkwyv2zD+TxP\nnsxy5pwzk8l3vvOe97xHSClRKBQKRepgGOkJKBQKhUJflLArFApFiqGEXaFQKFIMJewKhUKRYihh\nVygUihRDCbtCoVCkGErYFQqFIsVQwq5QKBQphm7CLoQwCiE+EEK8pFefCoVCoRg4Jh37uhXYDWT0\n1zAvL0+WlZXpOLRCoVCkPps3b66XUub3104XYRdClADnAfcCt/XXvqysjE2bNukxtEKhUHxsEEIc\nTqSdXqGY3wK3A1ofE7pRCLFJCLGprq5Op2EVCoVC0ZWkhV0IcT5wXEq5ua92UspHpJQLpZQL8/P7\n/SahUCgUikGih2NfAiwXQhwCngY+JYT4qw79KhQKhWIQJC3sUsrvSilLpJRlwJXAKinlNUnPTKFQ\nKBSDQuWxKxQKRYqhZ7ojUsq3gbf17FOhUCgUA0M5doVCoUgxlLCnIO6gmxcPvDjS01AoFCOErqEY\nxejgp+//lJcqXmJS1iROyj1ppKejUCiGGeXYU5A6X3QDWKOvcYRnolAoRgIl7GOEwy2HuXXVrfjC\nvn7b2o12AOp99UM9LYVCMQpRwj5G+Mm6n7CqchVbarf029ZitABwtPXoUE9LoVCMQpSwjxFagi0A\n7Gvax+2rb+9zcbQ11ApAjadmWOamUChGF2rxdIzgCXoA+NXmXwFwrPUYZ5SewRMfPsENJ91AmiUt\n3rYp0ARATasSdoXi44hy7GOE5mBzp+uN/kZerXiVR7Y/wrdXf7vTfS6/C+hd2J/c9aRKh1QoUhgl\n7GOAJn8T7qC70221rbXsatwFwHvV77G7YXf8PlcgKuzVnupuC6hSSn6x8RfcuebOIZ61QqEYKZSw\njwEq3ZWdrk/MmIg/4mdN9Rpm58/GZrTx9N6nAQhEAnjDXpZPXo4mNV448AKhSIjHdj5GS7CFKnfV\nSDwFhUIxjKgY+xigztv5YJIF4xZwuOUwx73HuXTqpZSklbC6ajXQHoaZWzCXKncVz+9/HqvRym82\n/wZvyMvkrMnxflx+F1m2rOF7IgqFYlhQjn0MUOut7XR9wbgF8csLCxdS5CzCFXAhpYyHYbKsWZxb\nfi4VzRX8ZvNvAHjxwIvcvvr2+GMPtRwa+skrFIphRwn7GKDOV4dJmPhU6acAmJs/N37fvIJ5ZFgz\nCGthfGEfB1wHABjvHM9ZE88CouEZaM9r//TETwPRTU8KhSL1UKGYMcBx73HyHHnct+w+6n31FDoL\nAci2ZmMymMi0ZALRXPf3jr5HpjWT6TnTMRqM3DTnJjItmTT6G/njjj9y6/xbue6k63jrr28px65Q\npChK2McAdd46CuwF2E12StNLAXjorIeYlj0NgAxrBhDNhll3dB2LixZjNBgB+NrcrwHR9EiJ5JoZ\n12A2mCnLLGNv494ReDYKhWKoUcI+Bqjz1TExY2Kn25YWL41fjjn2fU37qPPVdYrBx8ix5XDr/Fvj\n12fkzGB9zfohmrFCoRhJVIx9lCOl5HDLYfLt+b22iTn2j5o+AqA4rbjffmfkzqDOV6cKhSkUKYgS\n9lHOf73xX4S0EEVpRb22iTn2WGglFoPvi+k50wHYUbdDh1kqFIrRhBL2Uc7uht1MSJ/AFdOu6LVN\nzLHvbUpc2E/KPYk8ex4/3/jzeNEwhUKRGihhH8VoUsMddHNu+bk4zc5e2zlMDkzCRKO/EafZSbol\nvd++HWYHd51yF9WeanY17NJz2gqFYoRRwj6K8YQ8SGS/Qi2EiLv2Qkf/bj3GOMc4gIQO71AoFGMH\nJeyjmJZAtAZ7hiWj37axNomEYWLYTdGTlrxh7yBmp1AoRitK2EcxscM1Ym68L3JsOUBiGTExHGYH\nAL6QcuwKRSqh8thHMXFhT8Cxf+/U77G1bivLipcl3L9y7ApFapK0sAshbMBqwNrW3z+klD9Mtl/F\nwEIxU7OnMjV76oD6jwm7irErFKmFHo49AHxKSukRQpiBNUKIV6WUaltjksQce6Y1c0j6txqtGIQB\nb0g5doUilUha2KWUEvC0XTW3/chk+1UMLBQzGIQQ2E125dgVihRDl8VTIYRRCLEVOA68IaV8X49+\nP+64g26MwhgPmQwFDpNjyIT9w/oP+c+R/wxJ3wqFond0EXYpZURKORcoARYJIWZ2bSOEuFEIsUkI\nsamurq57J4putARayLBkIIQYsjHsJvuQLZ5+d813+e+3/pv7Ntynar8rFMOIrumOUkoX8BZwTg/3\nPSKlXCilXJif33tBK0U7LcGWhFIdk8FhdgxZuqNRREsHr9i9gge2PDAkYygUiu4kLexCiHwhRFbb\nZTvwaWBPsv1+XJFSsrl2My6/iz2Ne+L56UPFUMbYWwItXDTlIj4//fO8U/kO7qB7SMZRKBSd0cOx\nFwFvCSG2AxuJxthf0qHfjyWbajdx/WvXc/ozp3Oo5RBfnvXlIR1vqIQ9okVo8DeQb8/ns5M+S1AL\nsqZ6je7jKBSK7uiRFbMdmKfDXBRAhasCgEtPuJQl45ewrCTxDUeDwWFycNx7XPd+mwJNRGSEPHse\nM3JmIBAcbD6o+zgKhaI7aufpKOOI+wg2o43vn/p9DGLoKz4MlWOPHeBR4CjAYrRQ6Cyk0l2p+zgK\nhaI7qlbMKKPSXUlJesmwiDpEF0+HYoNS7FtAnj0PgAnpEzjiPqL7OAqFojtK2EcZle7K+IHVw8FQ\nO/Z8RzQDqjSjlCp3le7jKBSK7ihhH0VoUqPKXTWswu4wOfBH/ES0iK79xoQ95thL00tp9DfiCXr6\nephCodABJewjSLQaQzsNvgb8ET8l6SXDNofYrlZ/xK9rv03+JpxmJ1ajFSD+YVXlUa5doRhqlLCP\nEJUtlcz+y2y+/c63mfXELJ796Fka/Y0A5Npyh20eWbYsABp9jbr22xRoIsuaFb+eb4+GZGJOXqFQ\nDB1K2EeIDcc2APD6odcB+N8t/0tToAmAbFv2sM2jyFkEwNHWo7r26wq4Ogl77Dk1+Zt0HUehUHRH\nCfsIERPxjtdrPDUAQ77btCPjneMBqGmt0bVfl98V/zYA7cIe+1aiUCiGDiXsI8RRT3eHvLtxN0An\npzvUjHNGD7SOfajoRVfHnm5Ox2QwKceuUAwDSthHiEMth7rdtrthNwIxrMJuMVrIt+fr79i7CLsQ\ngmxrdrdvKgqFQn+UsI8Qh5oPxS/HjrTb1bCLTGsmRoNxWOdS5CzSNcYeioRoDbV2+4DKtmWrUIxC\nMQyokgIjQIOvgTpfe036KVlT2N+0n6AWJIMC7nlpF2/uruX608owGQ089t5Blk7JY96ELF7dcYz3\n9tczb0I2B+o8tPhCg56HyWjgypNL0UJZ7G85wL8+0CcV0R2Kivfh43TqMxxycKChtsdxSrId2ExG\nshxm8tKsrNpznGCke259jtPKsql5Q1qjXpFa7DnWwu6aFkwGA2fNGIfdYmTl5ipCEY0zZ4zjqQ1H\nuGxhCUWZQ3egzXCjhH0Y0KTGtrpt5Nvz2d24O16nPMeWQ6O/kVxbLnn2POp8dRxrMvGnrQcpyrTx\noxd3AVCYYeMv6w7zl3WHsZoMXDBnPC9tP8oJ49I5b1bRoOdV2eTlD6srsI4LY85s4JvPbNPl+Rqs\nx3BOgmc3NPKUu71P23iJ0X68z3Hs5qi41zT3nlc/bVw6abbR+dYVwGdnFXHalKFNWS3KtJNpN3e7\nvbE1yHG3H5vJSFmec0jnMNSEIxouX4i8NGun24NhjYP1rcgeTuD0BSP86d2DHGuJvn80KdlW6UJr\na/q982YwtzSL//nHNiSQ9doemrwhVrx/mFdvXUaO0zLUT2tYGJ3/HSnGfRvu46k9T8WvLy5ajMlg\n4tSiU3nl4CtkWjPJthRT56tjRkE+D119JnlpVqqavEgJJdl2Gr1BvIEI2Q4LmQ4zP7zgRJwWEwZD\ncs71WLOfP+zYxXMVG3j7259I8plG2dGwhTvXw68uOY05eSfHb//Dh5t5q+oAL3cZRwIvbjuK2x+i\nwRPE5Qtx78UzmZSX1q3vdz6q483dtchReqquyxfkJy/tGvJx0m0mzp89HmOHYGooLHl+WzX+kAbA\nJ6blU5I9dl3oB0dc7D3mZvnc8Tgs7eHJtfsbqKhv7fVxaVYTc0vbw4DXnjqRaxeXceOTm1i9r56X\nttdQlGnHaBCkWU388IKT+O9ntvKVJzexbGo+588ZT/kY/1BUwj7EeEPeTqIOsK5mHbPzZlPgKAAg\nzZyBrfVMYCvC5GZchg2Aibntb66CdBukt/eRbuvu1gZDYaaNcenphGWY4hwLZkPy/e5rjbqlaQWF\nlOW0P4eJWQW0HvIwPtuMxdjZGX3jzKkJ9V2W5+S608qSnuNQIaVkXUUDzd7Bh8j6IyIlT204wr8/\nPNbtvtOn5nPxvGL21LTwzKZKdlQ1D9k8hpp0m4lPTS/gnb2dj9LMcVr4xSWzSe/lW9vcCVk9hlVO\nn5LHE+uiRzTefdFMLl9YgsVoQAjBs5sreW9/AxsPNfH0xkpe/+Yy0qwDk8eqJi8tvjAnjh/aU88S\nQQn7EOMKuDpdz7JmMb9gPhdOuZCK5mjt9de3t/DejhLKJy/ne4svHfY5xssKhP2YLckLe21rLdC+\n2zTGCdknALDl+BZOLTo16XFGI0IITpucN+TjnD97fJ/3f3ZWEbd9ZtqQz2MsseyEfJ5Yd5gpBWlc\ntqAEq6n9W8BPL57F6n31TMlP4/N/Ws/lD6+j3hPgBxec2OtrHQxr3LFyO/MnZnPguIfH1x4CYN+9\n52I2jmxeihL2ISZ2HJxJmAjLMF+f93Uun3Y5APV7o9vrPzgUYEpBOn++5C5KcxzDPseOwp5uSe+n\ndf9Ueaqwm+zdNlotHr8Ym9HGqiOrUlbYFaOXT00v4O9fWcy8CVndhHdirpNr274h/+ziWXzvuZ1k\nOy1885mtnD41v9t6xqNrDvL7t/bT0BrkXx9Ud7pv3YEGlp0wsuc6q3THIaY1FI0FLhi3AGh3rQCZ\n1kwAvH4rPzj/xBERdWgXdr3K91a7qylJL+mWuWI32Tl1/KnqiDzFiCCEYFF5Tr9u+spFE9j547P5\nxSWzCUUk+2rbz+qVUvKXdYe4+6VdNLQGyXVaKM9zcu2pEzl9ah4Wk4G/vX+kW4G/4UY59iHGE4qW\nqb1s2mWUZZZxUu5J8ftOGbeENM+lZDqmsHjy8BX+6orNFI3p6yXsVZ7eSw+XZ5SztnotUkqVsqgY\ntdjMRk4ojH573VvrZmFZ9Nvnc1ur+cHzH7LshHy+vLScCTmOTtlHv/vPPn79xkf8Zd3hEV0LUsI+\nxMRCMSdkn8DZZWd3uu+5LXXUVC7kL1+cM6IxOT0du5SSak81i8cv7vH+bFs2QS2IN+zFaR7bmQeK\n1GZ8po10q4m7/rWTQEjji0vLee6Do0zIcfD49Sf3mJH29U9NYcPBRn74wocIAdecMjHpzLXBoEIx\nQ0zsYImuseuIJnnsvUPMn5A14vE4PYW9wd+AL+yjOK24x/tVMTDFWEEIgTsQBuAnL+3ieIuftQfq\nOWdmYa9iLYTgO+dOB+AHz3/I6n11PbYbapSw90Nti58PjjTxUa17UHEzdyjq2NPMnXOy1x1o4Eij\nlxuWlOsyz2SIhWL84eQP2zjSEj3XtLdQTGxBVRUDU4wFvry0/f9z0U//g5Rw0dyeTUuMmcWZvNW2\nV2PjoZExMErY+8AbDHPuA+9y8e/X8pnfrGZrpav/B3XBE/RgMpjiJwnFeGn7UZwWI58+cZxe0x00\nejr2/a79AEzN6jkvPduq6rIrxg53nTeDLd//dPz6yptPSyhPvTzPyZzSLDYeHJn3uRL2Pli5uYrG\n1iA3LCkDwO0PD7gPT8hDujm900Khpkle+/AYZ504Dpt5eAt+9YTdqN/xePua9uE0Oyl0FvZ4vwrF\nKMYSQghynBb+cdNi3r39k8wpTbzy6skTs9la5aLFP3Sb1XpDCXsvaJrk0TUHmVuaFd+goA0mFBN0\nk2bpHIapqG/F5Q2xdMrQb2RJBL0d+5SsKb1mvMRDMap8r2IMsbAsZ8DpyBfNKyYc0bj/tb1DNKve\nSVrYhRClQoi3hBC7hBAfCiFu1WNiI81/9hznUIOXLy0tJ7ZOMpjMVE/I0y2+vqM6GtKZXTJ8ddf7\nQq90RyllXNh7w26yYzVaVShGkfLMLM7kc/NL+MfmKjRtePPa9XDsYeBbUsoTgVOBrwkhTtSh3xHl\n0TUVFGfZOXdmYdx9DmTx1OV3UdFcgSfo6ebYt1c1YzMbmJw/OtL9rEYrApG0sLeGWnEFXEzMmNhr\nGyGEqsuu+NgwtzQLXygSrzY5XCQt7FLKGinllrbLbmA30Pey8ShnZ3Uz6ysaue60iZiMhrhj17TE\n+/jaqq9x4XMX0uhv7ObYt1W6OGl8JqYRricRQwiB3WRPWthjm7EyLH0vLmVbs2nwNSQ1lkIxFpjU\nZt4O1HmGdVxdlUUIUQbMA97Xs9/h5rH3DuKwGLni5AkACNoc+wD62F63HYgegdcxh33/cTdbjrg4\nY4Rz17tiM9mSTneMbcbq+g2lK3Py57Dx2EYVjlGkPFPyo/8LFXW9lxkeCnQTdiFEGrAS+G8pZUsP\n998ohNgkhNhUVzcySfuJUOcO8NK2Gi5dUBIv/BNbBxzI4mnHOPPCcQvjlx9dcxCLycDVp0zQZ8I6\noadj7/oNpStXTLuCoBbk+f3PJzWeQjHayU+3kmY1UTEWHbsQwkxU1FdIKf/ZUxsp5SNSyoVSyoX5\n+aPLrXbkb+8fIRjROtV5MMRj7In3ExNJu8nO8snLgehmp5Wbq7l0QQm5XU6FGWnsJvuwOfYp2VOY\nmj2V9TXrkxpPoRjtCCGYnO9kb4dCYsOBHlkxAngU2C2l/HXyUxo5gmGNv75/mDNOyGdyfrs4xRx7\nIounvrCP32/9PTWtNVwy9RLeuvyt+OHUj645SFjT+MqySUMy/2TQxbHHyieY+y/9Oyd/Djvqd6DJ\nASxcKBRjkFMn5bL5cNOw5rPr4diXANcCnxJCbG37+awO/Q47r+yooc4diG9IihF37An08Y+P/sFD\n2x5Ckxql6aXxQlfN3hAr1h/m/NnjO52MNFqwmWz6hWL6cewAs/Nm0xJs4XDL4aTGVChGO2edOI5Q\nRLL6o+ELQeuRFbNGSimklLOllHPbfl7RY3LDzeNrDzEpz8myqZ1DRQOJscc2+wDk2ttL8f5jSxWt\nwQg3nTFZn8nqTOwgkGSIh2L6ibEDzM6fDcDy55ZT2VKZ1LgKxWhm/oRsshxm1uyrH7YxR0e+3SjA\nEwiztdLFRfOKu1Vui6c7JmDZO56cnmtrF/aXth/lxKKMUXEeYk8YDUYiWiSpPjwhD0Zh7PTh1huT\nMifx6YnRGhwHWw4mNa5CMZoxGgQnFKT3eQC33ihhb+NQ24s+taAnt5n4BqXWYPsfL+bYKxu9fHDE\nxXmzi5Kf6BBhMpgIa8k59thmrEQO0BBCcNOcmwAIRAJJjatQjHbK8hwcVMI+/MQ2EEzK7y7s8ZIC\nCTj21nD7Hy9WyfDJ9YcxGgQXzRu9+7bMBnPywt5D+YS+iFW8VMKuSHXK89KocwdwD9MCqhL2Nirq\nWhECJuZ2L/TTvniagGNvO+P0fxb+D4XOQq7/8wYeWV3BuTMLKc7qP0QxUhiFkYhMMhQTHKSwh5Ww\nK1Kb8rbj8w7Ve4dlPCXsbRysb6U4y95jGV0xgJICraFWCuwFfOGkL/BRrYe399YxvTCd28+ervOM\n9cVkMBHSknMT7lD3SpZ9YTO2HfChQ7lghWI0ExP2ivrh2aikhL2NQw2t8Re/KwNJd2wNteIwR13/\ny9uPYhDw5JdOYUIP3wRGEyaDSRfHnkgOewyL0QKoUIwi9ZmY6yDdZsITSC7cmSjqMOs2XN5Qp01J\nPZFIumPHOPO7++uZPyGb/PTRtcu0J4zCmHSM3R10MzW755OTeiJWLlgJuyLVsZmNbP/hZxJKLNAD\n5djb8ATCOK09n2YUS39MJCvGG/LGNyUdbvAydVzioYmRJNmsGH/YzzHvsV7POu0JgzBgNphVjF3x\nsWC4RB2UsMfx+MOkWc093hf7cySSFeMJeXCanbT4QzS2BkflLtOeMBlMSeWxH2o5hCY1JmUNrFyC\n1WhVjl2h0Bkl7EAgHCEY0Ui39RyZisXYE9mgFHPsRxqiq98TB3ic1kiR7M7T2CHWUzJ7Pz2pJ5Sw\nKxT6o4SdqFsHSLP2JuzR34mkO8Yc++GYsI8hx55MVkyFqwKTMPV5elJP2Ew2JewKhc4oYQdaA9EQ\nRG/CToIlBaSUtIZaiYQtfO1vW4Ce8+JHI8mUFHh6z9P8cccfmZAxAbOx53BWb1iMlqTLBSsUis6k\nrLDf9vetzP3JvxNq6w5EnaqzV8ee2NbToBYkrIXZeDDq1ifnO3vtc7RhMpiQyEGJ+6sHXwXgjkV3\nDPixNqNy7AqF3owN1RkE/9xSnXDbWCimtxh7bPG0P8ce23X6UU2IG5aU8Z1zR/empI6YDVGnHZER\njPScHdQTwUiQnfU7ue7E6zht/GkDHlfF2BUK/UlZxz4QYpsGeo+xxxZPe1f2sBbmxT0bARCRbG5c\nNgmrKXGBHGmMIjrXgaY87qzfSVALMn/c/EGNq4RdodCflHXsAyEu7L059gQiMTe/+VXW16xDakau\nnn0WRZmjty5MT5gM0ec+0AXUvU17AZiVN2tQ41pNVlq83Y7IVSgUSaAcO+COhWJ6ceyiH8e+t3Ev\n62vWAZBjPInbz545BLMcWmLCPtCyAq6AC4AsW9agxrUarQQjwUE9VqFQ9Ixy7EBrP47d0M+GsRcO\nvIBJmHEfuZJfXXLJmArBxBhsKMYddOMwOeIx+oFiNVpVETCFQmeUYycaijEIsPdQ2RH6d+zratZR\nYp9J2HMSJxaO3sM0+iImzAMV9pZACxnWwZ8KpWLsCoX+KGEnGopxWk291nLo66CN5kAz+5v2ky6m\nYjEaxlxsPYbREP1QG2i6Y0uwhQzL4IXdZrKpWjEKhc4oYSfq2HuLrwMIei8psPX4ViQS6S2nJMeO\nsb+4zSjFJNoWT+XAFk+TFXaL0aIcu0KhM0rYAW8wjN3Se1xc9FFSYH3NeiwGC01NRWOmLkxPxBdP\nB+HY0y2J12Dvis1oI6gF0WQCp5goFIqEUMIOhCMSs7H3l6K3dEcpJW9VvsWpRadS2RAaM3VheiIW\nihlUjD0Jx67OPVUo9CclhT2RuukdCWsSk7H3EEp8g1KXWMxHTR9R7alm0bhltAYjY6YuTE8MdvHU\nHXQnvXgKqJRHhUJHUlLYg5GBfa0PaxKjofeXorej8WKlarONJwBjp+BXT8Ri7APJYw9pIbxhb3KO\n3RQVdlUITKHQj5QUdn9oYMIe0TTMfSx6tteK6Szt3nC02FejO9oiFUIxA9l56g66AZIS9sF8oCgU\nir7RRdiFEI8JIY4LIXbq0V+yBEIDE4lQRPaZzdJbjN0bigr739YfA6Ake2ymOkL74ulAQjEtgWgp\ngGRCMQYRfQsqYVco9EMvx/44cI5OfSVNIDxQx97f4mnPZ57uPd4AwEc1QSbkOMbkjtMYsZ2nAxHY\nlmCbsCfh2GPCrrJiFAr90KWkgJRytRCiTI++9MA/QMcejmgY+6mbbhDdY+wuvwepmfj5JXNZMDFn\ngLMcXQxm8TRWpjh2ePdgUMKuUOjPsMXYhRA3CiE2CSE21dXVDelYHWPsiWTIhDWJqZ+NRUKIbjH2\nQMSH1CwsKs9lSkHa4CY7ShhMHnssHm8xWAY9buybghJ2hUI/hk3YpZSPSCkXSikX5ufnD+lY/nC7\nOCVyAHWkn3RHiDr2rn0FIj7QLBh7KUUwlogJ7EB2noYibcJuHLywqxi7QqE/KZoV01HY+1f2UETD\n1Ee6I0Qde9euoo7dirGfD4WxwGAWT4NaNPd8oOecdkQ5doVCf1JS2AMdQjGJCHsijl3QPawT0Pwg\nU8OxDyYUE9tUlEwoRjl2hUJ/9Ep3fApYB0wTQlQJIb6kR7+DpWMoJpFNqP2lO0J0k1LXroIxxz5G\nC391JCnHPsha7NCeP69pyrErFHqhV1bMVXr0oxf+wTj2fhdPu5cUCGo+pDZ2Kzp2ZFDCHnPsKsau\nUIwqUjIU0znG3n/7aK2Yvl+KHh275ocUcezxE5Rk4sKu5+KpirErFPrxMRD2RNIdtcQce5e+QtKP\n1CwpIezJhGL0SHdUjl2h0I+UE/atlS7uf31v/HoiRjASkf1nxdA9Xh917Cm2eDrAImAdHzsYlGNX\nKPQn5YT94bcPYLcY+eKSciDBdEdN6z+P3SA6ZcVoUiMs/amzeCoGF2O3GCy9HimYCMqxKxT6k1LC\nXtXk5d+7jnHlyRMoy4uW0NVt8ZTO8fpYmdmPdSgmEkwqhx2UY1cohoKUEva/rj8CwLWLJ3Yotdv3\nY6SUhCL9C3t08bS9s1jJXjQrKaDrCCEwCuOAhD2khZKKr4PaoKRQDAUpI+wRTfLMxiN85sRCirPs\nvVZk7EpM+Ps6aANii6ft12MlewXWpEIRowmjMA4oK0YXx972ug/0rFWFQtE7KSPse4+5afKGOGdm\nIdD7qUddCbdtjOl352mXkgKxyoYGaR3chEchJoNpwFkxyTp2AyoUo1DoTcoI+5YjTQDMn5ANEA+P\n9BdjD0ei9/cfiuns/j0hT/R2OXYP1+iK0WAccEmBZHLYoUOMHSXsCoVepJSw56VZKM2JCm38AOp+\nLHu4rUF/G5QEncv2xg6ZMMqxe85pVwzCwN/2/I1qT3VC7UNaKGlhV1kxCoX+pISwhyMaa/c3MG9C\ndjzeHQt7dy0D0NNjIVHH3n49dt6nIHUc+/yC+QC8fuj1hNqHIqGk6sRAe4xd1YpRKPQjJYT9tQ+P\ncazFz2V3SwxzAAAgAElEQVQLSuK3xWPs/Tj2SJvw95eyGD1oo/26JxgNxZhSSNgf+OQDCER8/aA/\nglowaWFXjl2h0J+UEPY/vXuQslwHZ84YF78tluTSU4z9trdvY8XuFUB7KMbc7+IpndIdY47dkELC\nLoTAYXbEM376Q9cYu1o8VSh0Y8wL++bDTWytdPHFpeWdXHd7jL27sL9x+A3u23AfEFs8lfH2vSG6\nhGJagi0YsWFOYjv9aMRhcrTn6PeDHsKu8tgVCv0Zk8K+YvcKNtRsAODRNRVk2ExcMr+kUxvRy+Jp\nVzca1jTsJU/w1JEf9DmmQXQuKeAOujELZ78fCGMNp9mZsGPXY4NS17K9iZxRq1Ao+mZMCvt9G+7j\nS//+EpWNXl7beYzPnzIRp7Wzc46Z965C0eBv6HR9d+OHmNL3sM+9qc8xDV1j7CEPRlKjFntH7CZ7\n4jH2iH4xdk1quPwu5j45l1cqXkmqT4Xi486YFPYYj689hEEIrjttYrf7ekt3bPB1FvYNte/GL/cV\nDgiZP2Jd8LZ4bN0ddGPC0W82zVjDaXYmHIoJaSHdasVEZIQ/bP8DmtR488ibSfWpUHzcGXPCHju1\nB+CZjUc4b3YRRZndFzB726BU76vvdN0X8ccvuwKuXscNm/fjp549jXuAqLAbsWNIMWEf7sXTjo79\nrcq3AChyFiXVp0LxcWfMCXvHMEFrpJEvLS2PX28ONHPUcxQpZTzG7g/7O+2mrPPVAe3VDIORQPy+\nRl9jr+NGjNH79rv2A22LpzIFHbspcceuS0mBDrViYhUzY78VCsXgGHPCHtvKD3DihCCzS7KAqDB8\n5h+f4eyVZ/NO1TttoRjJF1adyV3v3RV/TMyx24w2AAIdvgE0+nsXds0YfdwB14H4PAw4Um7xNObY\na1trWV21usc2/z70b96tepdQRL+dp5rUCLR9yPojStgVimQYc8K+s6Y2fvmI9Vcc9RwFoouiMae5\nvW47BgHCGHX3L1e8HH9MLMYeE49ABxGJCXtlSyW/2/I76rx18fvCxujjDrgO8JvNv6E50IxB2vot\nHjbWiC2e/nDtD/naf75GZUtltzbfeudbfPU/X9Vlg1LHGHvsb+IL+5LqU6H4uDOmhP3nG37O7etu\nACDNnA7Au1XRxc+a1pp4u/2u/RiEwGCp79ZHzLGHtTChSIhgJIgWjvYVy5j584d/5o87/sj1r13P\nxmMbOdR8CM0Qjb8fcB3g2b3PApAemZ9yjt1pduIL++Lle//+0d873d/kb4pf1qSW9OJpzLGHtFC8\nsqQKxSgUyTGmhN1mssUvP/zph8iwZLC7cTdSyrhzL8soY79rP0KAwRJ13B1dZUdh8oa9BCMBZCgD\ngYFGfyNSSt6rfg+AI+4jfPH1L3LVy1cB4KCUpkAT7pCb2xbchk1OSLkYu8PsQCLjIvv8/ucJRUI8\nv/95ntz1JA9ufbBT+2Rj7LG1kI4uXYViFIrkGFPbJs1kxC+nmdOYkTODPY17+Np/vsa71VHnvrR4\nKSt2ryCk+TFYo8JuMVpYW72WH6/7cfwAZohuVgpqQaRmId2cya6GXRxsPsjR1qN8Y943+N0HvwPa\n4/rpchpeEQ1NFKUVEdFkymXFOE1OAKrcVQA0BZq49/17WblvZY/t9YqxdxL2Do799tW3MzlzMl+Z\n85WkxlEoPk6MKcfucrcfauE0O5mRO4N9Tfviog4wf9x8JJL36/6NJTe6+NcaauUHa3/A0daj1Pnq\nyLXlAlHHHtKCIE3k2wtZU72G296+DYDzJp3HkvFLOo3vpD0DZ7xzfEJnpY41HOZoGeJaby1nl51N\npjWTlftWUuQs4t0r3uWUolM6tbcakztoJBZj7yjsscu1rbW8evBV/m/r/yU1hkLxcUMXxy6EOAd4\nADACf5JS3qdHv11pqA5hDksKXOD71g/45PFKnAYffguU1knSfVDy2N18MzvCh+W/45NCI5KRzv4M\nDxN31bCkKdpualMzdlcY/9q7OcNcQyk2vlx8M/ebH2VX8wEmZ05mfNp4HjrrIW5840bW16wHwC4n\nEjtMdXzaeCLaXmzm1BR2gAJHAb8+49f8evOvue6k68iyZfHHT/+RVw++yh3v3gFAobMw6TGNwtgp\ndz4QCXCk5Qi3rLqlU7uXKl7iqT1P8eez/5z0NwWFIpVJWtiFEEbgQeDTQBWwUQjxgpRyV7J9d+XC\nDZv54usRDBKCmduxTyrhpP0SSxga06EmRzCtfC4z3n2DxXtiIZfmTn0cy4LwOCcbCwJ8YtsuLnBF\nd5Ly5l38wGpm5ckRJp4yEf+uXQizmSxDWvyxJplFvj2flmALubbcaChGCDSfD887qwk31KO53Vin\nTEFYbchQEGNmJhG3G6PTiXnCBEz5+Yh+zlcdSRymdmHPseWwqGgRT5//dPw2IQR59rz49fKMcpLF\nIAzxjKYsaxb+sJ+V+1ZysPkgEP1WUNFcwXff/S4AFc0VTM+Z3qmPWOkI0VbT52jrUYrTivsdu8ZT\nw592/Ilb5t1Cti076eei6BlNarxX/R4Lxi3oZB4GSkSLYDQYdZlTWAtjFMYhPbPYH/bTHGhmnHNc\n/411RA/HvgjYL6WsABBCPA1cCOgu7CdcciGPeN7EZxHc/eMXCWWncfmKRQBMyZrCjbNvZGL5uVz3\n1DIMxxsJNJ3CHSeewZvrf86RfMHxLHA7BF+ceQmP7XyMEz/5c+5e9WPk4XE8tvhmDC+t5LJ/vwlr\n/s3BX/0bgCuz7ThnauwvMnBK7SocZmgxWGmK/IUFO6sZH25l/x//m0hTU19TjyMsFozZ2VgmTsRc\nVIR12jTMxcWYiwox5eVhHj9e75dtQGRY2tcxsqxZPbbJseXEL49PG/h8ZShEuL6eSIsbGfBT3Cgw\n0xgf0+Wpp66pmuK0Yq6afhW/3Hg//9n8LNagxBaCQ0d2MCF9Av/98lcodRbzzVNv52cf/JKq/R9w\n2wk3sq15Ny9vXsGiBcv5+unfQXq9bKnbSnFGCeOyJ7Cu+QOCQR9nFZ3Big8e5fk9z7Dh4GoumHwB\nFdU7WTjtk1x6UnTB/Jcbf0mjv5Hvzv0WB5r28+B7v+TqyZdziuNEzOOLMOXkEGlpIXTsGKYJpQRE\nBLvJjsFoJBIJo9U3IP1+XMLPQ3se46aFt5Cr2VndsIFsZx6zbJM4Vn+ItQ2buWjR9RgDoejr43Jh\ndDg4HD7Oqrq1LHHM5ARnGTXeWv5S8QwZuUVcV34F1owsLJqBcG0tms+HweHA4HAgLWa2Nu3k7aOr\nOdRwgPlFCxmfVUqZrZip1hIM6ekIou/HcCTM4ZrdGJxOnt79NwxNLZyf/wmmTT4Fg9OJwRZNWpCa\nRlNrAx80bifPkYfP5+alAy9xfslnWFR8KsJqBYOB8PE6THm5SIOg2d/MK4deYf3R9bxd+Rbz8ufx\n0Gm/wtwaxGC1EvL7eM+7g1ZvM9lpeSwwTsJWUITB6SDicmFMTwejkQpXBU9teYznj77OnafexWeL\nP43FbCXiciGDQUyFhQijkS3VG3nxw2cpNOeClMyaspRTMmchNQ1jVhZCCNZWr+W3W37LPtc+ipxF\n3H/6L5huKSXc2IiltJSmcAsv7fkXRdkTqGk5SlaVi0/knoJ9Qhmm3FyE2YwWDHKgagc//eB+6oSb\n/1nwbZZkzEX6/Wh+PwarFX9jPd/deg8bAh9xQ/4FXLzwC2T7DJiLi+Ov6VAhkq2mJ4S4FDhHSvnl\ntuvXAqdIKW/p7TELFy6Umzb1XXSrJ7QPVjBnezTKs2PcBWBxMKvyGQCemPR55mdPA4uTq3c+yHb3\nQc6uLeTM83/P7Rs/16mf75/6fe5efzf3nX4fd6+9j6b6abx1/f9SlGnDv/NDZChIuL4e6fOx87Hf\nkrP3WPyx0iAQXQrQOJedTu4NN0Sdut1O4KN90dfGYiFcX4fR6UQLBAlVVRI8UkmkqQn/3j1E6hsI\nHz/eqS9zcTFZl12Kff58jFlZWEpKEHZ7j65C8/kQNlvcpQoh0LxehM2GDIWINDYSbmgk0tSEKS8X\nrbWVcFMTQggMaWn4d+5EhsMYHE5MRYUYMzIJayEu3vAV6jLhgcW/4JSmnOg/d1oaWnMzkeZmWnwu\n7nr3TrI98C3TuYQbGjAXFqK1thI6XkukyRX9pz1+HBkKdap3bLBakeEwmsfT7flU50C+x4AlGK3Z\n43MYsWImEvBj7nIOR8RuwegLduujPzQB9RmQ2wLGXt76GiAsZgITx+GqqyLdB85Az229WTYczQGQ\nEq3tT2SQEDEAsvcxIiIa1TPI9jGDdiM238AOHAkZ6fba9IZG+6KaFCCEAWG34zdLrC5vdP6y88Kb\nFAIy0sAfQASCRAzRb73pPkjzt8+/27wyHASDXpyDSHAKGyFkNmD3R98HfosgIiTOAHht0feHQQPN\nbMAUaq/vFLQa0DQNW6jnfiNGgRAGNBnBZzNgstiQPh+WoMQU68ZqwS9D2IKS+nSwhCGjy7YKYTZH\n39ex52oShAwSR4JvR8/Pb+PkC/8r0Zej89hCbJZSLuyv3bBlxQghbgRuBJgwYcKg+jA0HWq/svlx\nCHmhvBSA8W/fD5HoO7w0P5ftaU6ujmzFJw1MzpzMgeYD8YfGvqJ7w15CMki6DJL11ncQvlrsRjMY\nLdEfk4UPr8nm8fo6FlTDb8LVmGwaMgJaRKCFBdXWSUw40QOVD0G1GYxmHAYTGM1gMIPJApoNjFYo\ntUJ5JpjywTQbTFbCnhChBg9ht5/QcRee97dR99sHOj/vNCe2qWWYsjKJeH34d+9DC4aQ/gCmvGyE\n1Uq4oQnL+HEEDlVGTxkJD/5EogeBoBHM8n840suRdbe3/fbmvocpP5fWin0YbFZMedlYJo8n4vXj\nOHERwmJuP6cQ0FqjIRfbtEkY05wIs4kfb/w9ZS4TacdaqZuZxx7qSTPamRzOYHJGKa/Wb6Q+QzDX\nWsoRXHg9brI8Iex5+fhEGJevGac0ceWMC9hvcRPx+1g4cQFrPnybTUe347HBuQWLQGpE6lykNfrZ\nl21kY/AwAGflz+eE9FL+dOhlLHYHWV7wtjRTUl9NsMhE7rhiNkYqiRhgSv4knFYnO0QtxU0GDIeP\nEViYxdFMifO4h3lpk2kliAhHaAi52WNtJmAWWMISawjSpJnTSxay+sBaQOByQsAMJ7Q6MTR5qM0z\nEzZInJlZ1Lc2UiLTmWkazyr/XqxOJ82hVs50nEipzOC4aCXY5GJLqJKGdAhZjJiCEWyhqCAtTZ/O\n7LQybFY79b4mDKEwuwM1bG6twBrQEEgcAS+ZreBamM4iWxnFzgIcuXm86d/Nnuo92Dxh0n0e/GYI\nmA0USDuLfQVo2Wn40+3kOLPZHqjkiKcWY0jDH/Jz3Bxg0jEf6Rk5OLNzmGovIs+SAQbBetdeNnoP\nkp6TS627jrARzjBOYm7eCdT7mthnbsFS7aLJ3cDBjCCWEJwQzqTImkNByQmk17dQZfBQI1s4fPww\n7jQzLock1w3jIw5yLOksmnoymsmIEcHf9rxKiylM2AB2dxDQEFJwgXMmOQYneyN1rPMdwGMTuB0w\nuUESiQjOKppDVm0zaRYrx8sdPBP+CBq8FPiM5IctHDZFKEnL5BxTEcId4CNvE5szQ+wRrXitYA1B\nxGbg8nAOJ7YaqTcHqHb7ODw+n0/O6leXk0YPx74Y+JGU8uy2698FkFL+rLfHDNaxA8x6YhYAO67b\nAVLypddvYEPtZrZe8DzGcACCXh7a/3d+f/gVVh+uonrpQ2QvOp2fvf8zKporqHRX8sJFL7D8ueV8\ne+G3+fWm33BJk5/ve1yIvCkQCUMkAJEQRII8ZZX8NMPG9ECYz7Rexn9dch4EPRDy8vBzb3K2eJ/y\nTGO0vRZqe1zschgiQQj7gcRf56DHSMhjJOwzEvJHL/tdZiJ+A1pEkF7sx2CSGC0agRYTEb8BszNC\n2G/EmhmKui6zxGjTMFk1jBaNsN+AwSwx2SJIKYgEDVgzQphsGpGggZDXiBaKCvBfRRZNrWYu9noo\nzvETCQqQAqMl2hcCvjUul8Waj2sizSQbolw6oRirlBw3mVju9vBCenRd4wZXCxd7PCwviYZ77qpv\n5E9ZGdSaTJzV6uUbTS7+mZbG41kZ3FnfyFXuzt8CJPBAdiYZmsYXm93d7nvZ6aDJaORStwe7lGyx\nWknTNDI1jccz05kTCHJCMEhhOMKNhQVss1l5/GgtCwLt9v0Nh50f5uWSqUW4r66BOYF226YBh80m\n/pCVyZRgiFN9fq4rGkfQIHBoGt9qdHF3Xg4zAwEerTnOG04HG2xWLvS0MicQYLfFwqxAEA24rSCP\nSrOJVoOBvx6tZVwk0mmcLxSN4yOLmQdq65gXCBIG0nr535bAL3KycBmNFIbD7LZY+FF9I4WR7mbA\nIwQr09PI0DQmhMLMCgToa9m6zmjgvwoLON3r59YmVzfnKAG/EJik5Md5OUwOhbi+2U3Xt9CHFjO/\nzMlmTiDALU3NPTrQ3RYz/0pLQxNwo6uFgh7mv89s5pGsDEzAF5pb2GOxMCkUiv+dvELw58wMlnta\n+Ve6kzV2O5e6PVze5b0UAd6z23glzUmNycgiX4CbWlqj6boGU9uPgXVWC7ssJuZEDMyLCIwGM5jt\nYLJFf5/8ZTjpoj5ewb5J1LHrIewm4CPgTKAa2Ah8Xkr5YW+PSUbYb111K5rU+N8z/xeI5qIf9x6n\nLLMs3qbR38hzO97m6pdupKnsXAo/+10w27ht8y954+i7bLxiDSc/s5SbZn6Jh3c+ys2NzVx3wVM4\npyzpNt6LB17kzjV34ozMoCz8TZ6+cXH8vk/c/xZzSrN44Mp5fU9aStDCUYEPtwl92N8u+uFAh58B\nfHdNSFETVN0OfXkjQd5u2sVn8+YOrr+ElT7a7oxNPyIoI3gifq4tPJ0nj0XTV79TdiEX5i1k8abv\nA/Duwh+zseUA7zbt4UeTL8UgDHjCflY17eT8vAXx1Ek959YRd9hHusneV5N++9rYvJ+VtetYlDmV\niwtOYXdrFVMdRQPYwdvzoK1hP41hD6W2toXtIXpvJN2fnn0l3F+C84p9WzeY2y8bLW3fvk0dfowg\nDAN4LvoxbKEYKWVYCHEL8DrRdMfH+hL1ZHngU53DFA6zo5OoQ3Rxb0Hhp3hVO5mLDr8AD70AQHF2\nFjnpTmz3TcA2sYSmtb+FjHTWROZy/YTO+dkxYouJRhzdartHpMSYyB83/qYxQ3Jp38OCA/gsFw/b\neIZtP8fbVqcnq2gutAl7YfkncZZ+CjZ9n9OLTyfrpM/xaaLpVzHSgOVcOCzzTNehj5M5k5Np32x1\nog59AjjbfhQK0CnGLqV8BRhVx94YhOCboa9S/OlvcHKuH0I+/svfxOcCLphixXH4aVyF5eA9zMbI\nTEy9pCCmW6L/zkac3aIpkYhMuROURgKDMMQPOemYiVPoKEQIwbqr1mE3pc6h4QrFUDOmSgoMBCFA\nYqApdx6cFN1Ek9H2A2Bf+SautELwHgZp7nUHaZolGu81Yu92aEdEKmHXg1hZAYAMa3u65aSsSUD7\n30ChUCRGygp7b0fjxXCYHTQG2uqvS1OvNV9ioRgTzm7LnxFNCbsedIyNd3TsyqWnBqFQiKqqKvx+\nVdwtUWw2GyUlJZjNg6uemrrC3qYVvS0OO0yOeEVIo+h9nT/XnsuykmXUHJrW3bErYdeFmGM3GUzx\n2jMTM7qfY6sYm1RVVZGenk5ZWdmQ7vJMFaSUNDQ0UFVVRXn54HZ2j9697UnSr2M3OeIlfE19CLvZ\nYObBMx8kTUzp1ldYCbsuxBy7zWgj05oJwJXTrhzJKSl0xO/3k5ubq0Q9QYQQ5ObmJvUNJ3Udey+H\nWcdwmB3xwySMov+vOwZBpx2UAJqWYFaMok9iwm41WpmUOYk3Ln1Dl+JiitGDEvWBkezrlbKOXcQd\ne++hmBh9OfZ4f3R3/8qx60PcsbcdpKJEXaFIjpQV9lgoprf9Vx0rzCUi7AYhkF2WTzWVFaMLsRi7\nKsWrGAukpY3+LK0UFvbo74QcewLHuwkBXcumKMeuDx1j7ArFSBDpoRzBWCZlhV3Q9+Kp3dyeSmdJ\nYE+hEKKTX5dSIiVK2HUg5tg7nmmrUOjFoUOHmD59OldffTUzZszg0ksvxev1UlZWxh133MH8+fN5\n9tlnOXDgAOeccw4LFizg9NNPZ8+ePQAcPHiQxYsXM2vWLL73ve/1OZbH4+HMM89k/vz5zJo1i+ef\nf344nmI3UnbxVAzAsVsMGT226YhBdE6djLR9YqjF0+QxGJRj/7jw4xc/ZNfRFl37PHF8Bj+84KQ+\n2+zdu5dHH32UJUuW8MUvfpHf//73AOTm5rJlyxYAzjzzTB5++GGmTp3K+++/z1e/+lVWrVrFrbfe\nys0338wXvvAFHnzwwb6GwWaz8a9//YuMjAzq6+s59dRTWb58+bAvHqessMc3HPUTYzdIGxZj/yey\nCESnD4lwTNiNStiTRTl2xVBTWlrKkiXRIn/XXHMNv/td9KD6K664Aog67bVr13LZZZfFHxNoq+L5\n3nvvsXJl9DD3a6+9ljvuuKPXcaSU3HnnnaxevRqDwUB1dTW1tbUUFg5vQkDqCnuCjt1EGuYExNlg\n6LwQG+tXOfbk6ZoVo0hd+nPWQ0VXxxy77nRGS6dpmkZWVhZbt25N6PG9sWLFCurq6ti8eTNms5my\nsrIR2XGbsjH2RDYoARikA5Ox/5ehV8euYuxJE3PsqoSAYqg4cuQI69atA+Bvf/sbS5cu7XR/RkYG\n5eXlPPvss0DUeW/btg2AJUuW8PTT0XN/V6xY0ec4zc3NFBQUYDabeeuttzh8+LDeTyUhUlbY+4ux\nxw7ENUgnlkSEXXSO6mhK2HVDZcUohppp06bx4IMPMmPGDJqamrj55pu7tVmxYgWPPvooc+bM4aST\nToovfD7wwAM8+OCDzJo1i+rq6j7Hufrqq9m0aROzZs3iL3/5C9OnT++z/VCRwqGYWB57z8IeiETj\nZwYtE1MCoZjouaLt12OOvbeqkIrEUTF2xVBjMpn461//2um2Q4cOdbpeXl7Oa6+91u2x5eXlcbcP\ncM899/Q6Tl5eXqe2I0XKOvb+QjFnlJzBNTOuIa31YswJOPauWTExx95bVUjFwFHCrlDoQwo79ujv\n3kIxFqOFOxbdwRtr3kkoFGMQotOHhHLs+hHUoudP2o0qxq7Qn7KyMnbu3Klrnzt27ODaa6/tdJvV\nauX999/XdZzBkrLCLvpx7DFCES2xUAydPyRieewGlRWTNMFIVNiVY1eMFWbNmtVrBs1oIIVDMdHf\n/R3WHY7IhEIxXWPs8XRH5diTRgm7QqEvKSzsfVd3jBGMaAnlsYsuMXaV7qgfStgVCn35GAh73+3C\nES3xxdMO1yNK2HVDxdgVCn1JWWHvL489RigiMRkGvkFpW6ULgNJsR28PUSSIcuwKhb6krLD3V489\nRiiiYTYNvKTAyztqKMm2M7skM5lpKoCQFgKUsCvGBnrWY//BD37Am2++qVt/MVI2Kyae7thPLCYU\n0TAn4tg7pDsGwxpr9tVz/WnqcF49iDl2VVJAMVJEIhGMCRQD1Juf/OQnQ9Jvyjv2vnQ9okk0SWJZ\nMbQvnh53+wlrkikFo/8klbFAXNhVjF0xBAxnPfa3336bZcuWcd555zFt2jRuuukmNE0jEolw/fXX\nM3PmTGbNmsVvfvMbAK6//nr+8Y9/6P6ck3LsQojLgB8BM4BFUspNekxKDxKJsYci0SOREsljN3Q4\naKO2JVqOYFyGCh3oQWzxVIViPga8+h04tkPfPgtnwbn39dlkuOqxA2zYsIFdu3YxceJEzjnnHP75\nz39SXl5OdXV1fKOUy+VK8kn3TbKOfSfwOWC1DnPRFdFPrRhoF/ZEi4DFPiSOt0TLcBZkWJOdpqID\nStgVQ0XXeuxr1qwBeq7HPnfuXL7yla9QU1MDROuxX3XVVQDddpv2xKJFi5g0aRJGo5GrrrqKNWvW\nMGnSJCoqKvj617/Oa6+9RkZG/4f7JENSjl1KuRsSr1U83BhE36GYcKStLECijr2tr9o2YVeOXV+U\nsH8M6MdZDxXDVY+9t7Gys7PZtm0br7/+Og8//DB///vfeeyxxwbyFAZEysbYIVbfpX/HnkiMHdod\ne607gMkgyHH0fwi2InHMBvNIT0GRogxXPXaIhmIOHjyIpmk888wzLF26lPr6ejRN45JLLuGee+6J\nh3+Gin4VTQjxphBiZw8/Fw5kICHEjUKITUKITXV1dYOf8QDoGBfviVCbnU/oBKUOBdmPtwQoSLeq\nyo468ZXZX4nXZFcohoLhqscOcPLJJ3PLLbcwY8YMysvLufjii6muruYTn/gEc+fO5ZprruFnP/uZ\n7s+xI/2GYqSUZ+kxkJTyEeARgIULF/aTXa4PHePiPREKJ+7YDR1j7G4/BSoMoxu3zLuFW+bdMtLT\nUKQww1WPHaLu/6WXXup025w5c3p06Y8//ng/Mx8cKW2TDF0Kd3UlrMWyYhJdPI1erm3xM04tnCoU\nilFKsumOFwP/C+QDLwshtkopz9ZlZjpgEH1vUAqGo/dZEk53bIuxtwQ4dVKuPpNUKBRDynDXY//E\nJz6h61iDIdmsmH8B/9JpLrrT9XCMrsQdewI7T2lz7P5QhGZfSGXEKBQfY1Q99hGk3xh7LCvGlNgJ\nSsjowilAQboKxSgUitFJSgu7wSD62aDUlhWTQHZL7ASlWrfKYVcoFKOb1Bb2fkIxA3XsErU5SaFQ\njH5SXNj7DsXEdp4OJN2xvU6MCsUoFIrRSUoLu+jHsQdjRcAS2WjUljp5vMWPxWQg0652SSoUY5Ef\n/ehH/PKXv+z1/ueee45du3YN44z0J6WF3SD6LgIWc+yWhEIx0d/H2nLYR2t9HIVCkRypIOwpe9AG\n9F8rxh+KAAlWdyQq5MdbAuSlqTCMQjEYfr7h5+xp3KNrn9NzpnPHojv6bHPvvffyxBNPUFBQQGlp\nKUMyEdkAAAvWSURBVAsWLOCPf/wjjzzyCMFgkClTpvDkk0+ydetWXnjhBd555x3uueceVq5cyapV\nq7q1czh6PhLzxRdf5J577iEYDJKbm8uKFSsYN26crs83EVLcsfcdinH5okeyZTn6D6vEHHtja1AV\n/1IoxhCbN2/m6aefZuvWrbzyyits3LgRgM997nNs3LiRbdu2MWPGDB599FFOO+00li9fzv3338/W\nrVuZPHlyj+16Y+nSpaxfv54PPviAK6+8kl/84hfD9TQ7kdKOvb889mZvECEgw9a/sMciL43eIDOL\n1TmnCsVg6M9ZDwXvvvsuF198cdxlL1++HICdO3fyve99D5fLhcfj4eyze940n2g7gKqqKq644gpq\namoIBoOUl5fr/4QSIOUde1+1Yly+EJl2c0JVGmMx9cbWINkJOHyFQjG6uf766/m///s/duzYwQ9/\n+EP8fn9S7QC+/vWvc8stt7Bjxw7+8Ic/9Nl2KElpYe/PsTd5Q2QlmN0SO0M1okmynSoUo1CMFZYt\nW8Zzzz2Hz+fD7Xbz4osvAuB2uykqKiIUCnWqs56eno7b7Y5f761dTzQ3N1NcXAzAE088MQTPJjFS\nWtj7jbF7g2QmGC/vmASTrWLsCsWYYf78+VxxxRXMmTOHc889l5NPPhmAu+++m1NOOYUlS5Ywffr0\nePsrr7yS+++/n3nz5nHgwIFe2/XEj370Iy677DIWLFhAXl7ekD6vvvh4x9h9oYRF2tBJ2FUoRqEY\nS9x1113cdddd3W7v6cCNJUuWdEp3vPnmm3ts1xMXXnghF144oDOIhoSUd+x9HaHk8oYSFulYuiNA\nlnLsCoViFJPSjr2/kgJN3mDCIt0xFJOjYuwKxceae++9N34+aozLLrusx28FI0GKC3vvG5TCEQ23\nP5xwaYCOO01VKEah+HjTW2hntJDSoZi+asW0+MNAYpuToHOMXYViFArFaCalhb2vWjEubxBIPMPF\n0MGxJ1JbRqFQKEaKlFaovtIdd9W0AFCaY0+or5iuF2cl1l6hUChGihQX9t4XT9/cVUu2w8zc0uyE\n+orF2Mvyei7+o1AoFKOFlBZ2IQSRHiy7PxThrb11fGr6OIyJ1GIHguFo7faJuU5d56hQKIaX0VSP\n/bTTThuSflNa2LMdZpraYukdeWR1Bc2+EFecXJpwX5WNXgDKcpVjVyhSmeEU9rVr1w5Jvymd7jgu\nw8aO6pZOt22tdPF/q/Zz3qwiFpXnJNxXVVNU2JVjVygGz7Gf/pTAbn3rsVtnTKfwzjv7bDNc9div\nv/56bDYbmzZtoqWlhV//+tecf/75fPjhh9xwww0Eg0E0TWPlypVMnTqVtLQ0PB6Prq8HpLhjH5dh\no6E1ED+0usET4Kt/3Ux+upV7Lpo5oL6+c+50zp1ZyBkn5A/FVBUKxRAxnPXYAQ4dOsSGDRt4+eWX\nuemmm/D7/Tz88MPceuutbN26lU2bNlFSUjKkzznlHbuUUOcOUJBu5etPfUB9a5B/3nzagCs0TilI\n56FrFgzRTBWKjwf9OeuhYDjrsQNcfvnlGAwGpk6dyqRJk9izZw+LFy/m3nvvpaqqis997nNMnTpV\n3yfZhZR27IWZ0SPsalv8/G3DEdYeaOCei2aqgzIUCsWQ1GMHup2HLITg85//PC+88AJ2u53Pfvaz\nrFq1Srfn0RNJCbsQ4n4hxB4hxHYhxL+EEFl6TUwPCtJtABxr9vP42kPMLc3i8oWJL5gqFIqxz3DW\nYwd49tln0TSNAwcOUFFRwbRp06ioqGDSpEl84xvf4MILL2T79u36P9EOJOvY3wBmSilnAx8B301+\nSvpRmBkV9vtf30tFXStfWDxxhGekUCiGm+Gsx87/t3euIVaUYRz//bPVEyWVXcXN2q0gRWKzzYrE\noBuuGBbsB6MwKxC6QH2IsoSyD30oqCCKFqPabpR2oz4UZCX0KWsrta3N3G6kWNpGN4LKfPow7+p4\n3DmrnuO8c8bnB4fzzvvOOee3z848Z+aZOTPA5MmTmTFjBl1dXfT09FCpVFixYgXTpk2jo6OD/v5+\nFixYsN/+XgBl/eR+r99IuhzoNrMrR5u3s7PT+vr6GvK5tdi+3Wi/800ALppyLD1XncnBY0pdfXKc\nwjEwMMCUKVNia+TCwoULmTt3Lt3d3XW/10hxk/SxmXWO9tpGHjy9FlieNShpEbAIkm+0PDjoIHH3\npVMRcMXZkz2pO45zQDBqYpf0DnD8CENLzOz1MM8SYBuQWYAys2XAMki22PfJdh+45rw4dwl3HKe8\nZF2Pvbe3N45QFaMmdjO7qNa4pIXAXOBCa1Rdx3Ecp8AU/XrsdZViJM0GbgPON7O/GqPkOE7ZMLPd\nTgN0sql3G7neovMjwHhgpaQ1knrqfD/HcUpGpVJhaGio7mR1oGBmDA0NUalU9vk96tpiN7NT6nm9\n4zjlp7W1lY0bN7J169bYKk1DpVKp67IDpb6kgOM48WlpaaGtzU9iyBM//89xHKdkeGJ3HMcpGZ7Y\nHcdxSkbDLimwVx8qbQW+38eXHw383ECd/YV7Np5mcXXPxuKeOznRzEa9KUSUxF4Pkvr25FoJsXHP\nxtMsru7ZWNxz7/FSjOM4TsnwxO44jlMymjGxL4stsIe4Z+NpFlf3bCzuuZc0XY3dcRzHqU0zbrE7\njuM4NWiqxC5ptqT1kgYlLY7tk0bSd5I+CxdD6wt9EyStlLQhPB8ZwetJSVsk9af6RvRSwsMhvusk\nTY/suVTSphDTNZLmpMbuCJ7rJdW+bXxjPU+QtErSF5I+l3Rz6C9UTGt4FiqmkiqSPpS0NnjeE/rb\nJK0OPssljQ3948L0YBg/KQ/PUVx7JX2bimlH6I+2PmFmTfEAxgBfA+3AWGAtMDW2V8rvO+Doqr77\ngcWhvRi4L4LXLGA60D+aFzAHeAsQcA6wOrLnUuDWEeadGv7/44C2sFyMyclzIjA9tMeT3Ot3atFi\nWsOzUDENcTkstFuA1SFOK4D5ob8HuD60bwB6Qns+sDzHZTTLtZfktqDV80dbn5ppi30GMGhm35jZ\nP8CLwLzITqMxD3g6tJ8GLstbwMzeB36p6s7ymgc8YwkfAEdImhjRM4t5wItm9reZfQsMkiwf+x0z\n22xmn4T2H8AAMImCxbSGZxZRYhri8meYbAkPAy4AXg791fEcjvPLwIXK6ULvNVyziLY+NVNinwT8\nkJreSO0FNW8MeFvSx0ru7wpwnJltDu0fgePiqO1GllcRY3xT2I19MlXKKoRnKAOcQbLlVtiYVnlC\nwWIqaYykNcAWYCXJ3sKvZrZtBJcdnmH8N+CoPDxHcjWz4ZjeG2L6kKRx1a6B3GLaTIm96Mw0s+lA\nF3CjpFnpQUv2zQp3ClJRvQKPAScDHcBm4IG4OjuRdBjwCnCLmf2eHitSTEfwLFxMzew/M+sAWkn2\nEk6LrJRJtaukacAdJM5nAROA2yMqAs2V2DcBJ6SmW0NfITCzTeF5C/AayQL60/CuV3jeEs9wF7K8\nChVjM/sprEjbgcfZWRqI6imphSRZPm9mr4buwsV0JM+ixjS4/QqsAs4lKVsM3y8i7bLDM4wfDgzl\n6Qm7uM4OZS8zs7+BpyhATJspsX8EnBqOlo8lOXDyRmQnACQdKmn8cBu4BOgn8bs6zHY18Hocw93I\n8noDWBCO5p8D/JYqL+ROVT3ycpKYQuI5P5wh0QacCnyYk5OAJ4ABM3swNVSomGZ5Fi2mko6RdERo\nHwJcTHI8YBXQHWarjudwnLuB98Ie0n4nw/XL1Be6SI4FpGMaZ33K6yhtIx4kR5m/IqnBLYntk/Jq\nJzmjYC3w+bAbSe3vXWAD8A4wIYLbCyS73P+S1Piuy/IiOXr/aIjvZ0BnZM9ng8c6kpVkYmr+JcFz\nPdCVo+dMkjLLOmBNeMwpWkxreBYqpsDpwKfBpx+4K/S3k3yxDAIvAeNCfyVMD4bx9hz/91mu74WY\n9gPPsfPMmWjrk//y1HEcp2Q0UynGcRzH2QM8sTuO45QMT+yO4zglwxO74zhOyfDE7jiOUzI8sTuO\n45QMT+yO4zglwxO74zhOyfgfbyldISLMzQkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f72dced2a20>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "prediction = np.array(hmm.pred_hidden_state_list)\n",
    "plt.plot(prediction)\n",
    "plt.plot(test_data[10:,[4,5]])\n",
    "plt.legend(['pred_a','pred_psi','data_a','data_psi'])\n",
    "plt.show\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "\n",
    "'''\n",
    "load one example segment of the first channel obstacle data\n",
    "'''\n",
    "exp_data_path = \"test_data.csv\"\n",
    "exp_data = read_data(exp_data_path)\n",
    "exp_data[:,0] += 1149\n",
    "'''\n",
    "load the ego car data\n",
    "'''\n",
    "ego_data_path = \"ego_data_24d.csv\"\n",
    "ego_data = read_data(ego_data_path)\n",
    "ego_lane = ego_data[:,[18,19,15,16,20,21,22,23]]\n",
    "ego_data = ego_data[:,[0,17,18,19,9,14]] # x_dot_ego, psi_T_0, left_dist, right_dist, accel_x, r_rate,\n",
    "\n",
    "# other wise the left and right dist is 1.88 and -1.88\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1.,  1.])"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "a = np.ones([5,3])\n",
    "a[1,0:2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "\n",
    "'''\n",
    "Use kinematics to estimate distances from obs cars to their lane edges and obs cars' yaw angle and rate\n",
    "'''\n",
    "x = exp_data[:,3]\n",
    "y = exp_data[:,4]\n",
    "left_at_obs = -ego_lane_exp_list[:,6] * x**3 - ego_lane_exp_list[:,4] * x**2 - ego_lane_exp_list[:,2] * x + ego_lane_exp_list[:,0]\n",
    "right_at_obs = -ego_lane_exp_list[:,7] * x**3 - ego_lane_exp_list[:,5] * x**2 - ego_lane_exp_list[:,3] * x + ego_lane_exp_list[:,1]\n",
    "center_at_obs = (left_at_obs + right_at_obs)/2\n",
    "\n",
    "obs_left_dist = np.ones(exp_data.shape[0]) * 1.88\n",
    "obs_right_dist = np.ones(exp_data.shape[0]) * -1.88\n",
    "\n",
    "dist_to_left = y-left_at_obs\n",
    "dist_to_right = y-right_at_obs\n",
    "# when the obs car is within ego lane\n",
    "within_ego_lane_idx = (dist_to_left < 0) & (dist_to_right > 0)\n",
    "obs_left_dist[within_ego_lane_idx] = -dist_to_left[within_ego_lane_idx]\n",
    "obs_right_dist[within_ego_lane_idx] = -dist_to_right[within_ego_lane_idx]\n",
    "# when the obs car is within the left/right next lanes\n",
    "within_left_lane_idx = (dist_to_left > 0) & (dist_to_left < 3.7)\n",
    "within_right_lane_idx = (dist_to_right < 0) & (dist_to_right > -3.7)\n",
    "obs_right_dist[within_left_lane_idx] = -dist_to_left[within_left_lane_idx]\n",
    "obs_left_dist[within_left_lane_idx] = 3.7 -dist_to_left[within_left_lane_idx]\n",
    "obs_right_dist[within_right_lane_idx] = -dist_to_right[within_right_lane_idx]-3.7\n",
    "obs_left_dist[within_right_lane_idx] = -dist_to_right[within_right_lane_idx]\n",
    "\n",
    "'''\n",
    "Use kinematics to estimate obs car yaw angle;\n",
    "assume the heading angle of obs car is parallel to its lane edge.\n",
    "'''\n",
    "heading_at_obs = ((-3*ego_lane_exp_list[:,6] * x**2 - 2*ego_lane_exp_list[:,4] * x - ego_lane_exp_list[:,2]) + \n",
    "                 (-3*ego_lane_exp_list[:,7] * x**2 - 2*ego_lane_exp_list[:,5] * x - ego_lane_exp_list[:,3]))/2\n",
    "\n",
    "psi_dot = np.zeros(exp_data.shape[0])\n",
    "psi_dot[1:exp_data.shape[0]] = heading_at_obs[1:exp_data.shape[0]]-heading_at_obs[0:exp_data.shape[0]-1]\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.37049304  0.48727663  0.08956849]\n",
      " [ 0.13063079  0.77840375  0.42856613]\n",
      " [ 0.24579774  0.8063904   0.4388607 ]\n",
      " [ 0.70256156  0.73732384  0.95859231]\n",
      " [ 0.16382847  0.14731958  0.77359607]]\n",
      "[ 0.12377799  0.04667647  0.15624804]\n",
      "[[ 0.04585888  0.02274435  0.0139949 ]\n",
      " [ 0.01616922  0.03633314  0.06696262]\n",
      " [ 0.03042435  0.03763945  0.06857112]\n",
      " [ 0.08696166  0.03441567  0.14977817]\n",
      " [ 0.02027836  0.00687636  0.12087287]]\n",
      "[ 0.08259813  0.11946497  0.13663493  0.2711555   0.14802759]\n",
      "[[ 0.55520483  0.27536155  0.16943362]\n",
      " [ 0.13534693  0.30413213  0.56052094]\n",
      " [ 0.22266891  0.27547461  0.50185648]\n",
      " [ 0.32070771  0.12692227  0.55237003]\n",
      " [ 0.1369904   0.04645322  0.81655638]]\n"
     ]
    }
   ],
   "source": [
    "# print(ego_lane_exp_list[0:50,0])\n",
    "# print(ego_lane_exp_list[0:50,1])\n",
    "# print(left_at_obs[0:50])\n",
    "# print(right_at_obs[0:50])\n",
    "# print(dist_to_left[0:50])\n",
    "# print(dist_to_right[0:50])\n",
    "# print(obs_left_dist[1100:-1])\n",
    "# print(obs_right_dist[1100:-1])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<function matplotlib.pyplot.show>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiEAAAEYCAYAAACKiiZ/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nGxubaeH/A56O5h+hyhviAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f8757f786a0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(1)\n",
    "ax1 = plt.subplot(411)\n",
    "ax1.plot(range(ego_prob_list.shape[0]),ego_prob_list)\n",
    "ax1.legend(['prob_maintain','prob_left_turn', 'prob_right_turn'], bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)\n",
    "ax1.set_title('Ego car maneuver probability')\n",
    "\n",
    "ax2 = plt.subplot(412)\n",
    "ax2.plot(range(obs_prob_list.shape[0]),obs_prob_list)\n",
    "ax2.legend(['prob_maintain','prob_left_turn', 'prob_right_turn'], bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)\n",
    "ax2.set_title('Participant car maneuver probability')\n",
    "\n",
    "ax3 = plt.subplot(413)\n",
    "ax3.plot(range(ego_car_list.shape[0]),ego_car_list)\n",
    "ax3.legend([r'$v$', r'$a$', r'$\\dot{\\psi}$', r'$\\psi$', r'$d_l$', r'$d_r$'], bbox_to_anchor=(1.05, 1), loc=2, ncol=2, borderaxespad=0.)\n",
    "ax3.set_title('Ego car data')\n",
    "\n",
    "ax4 = plt.subplot(414)\n",
    "ax4.plot(range(exp_data.shape[0]),exp_data[:,3:7])\n",
    "lgd = ax4.legend([r'$\\Delta x$', r'$\\Delta y$', r'$\\Delta v}$', r'$\\Delta a$'], bbox_to_anchor=(1.05, 1), loc=2, ncol=2, borderaxespad=0.)\n",
    "ax4.set_title('Participant car data')\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.savefig('action_data.png', bbox_extra_artists=(lgd,),bbox_inches='tight', dpi=200)\n",
    "plt.show"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model name:  GMMHMM_models/GMMHMM_model_10_5.pkl\n",
      "<bound method _BaseHMM.score of GMMHMM(algorithm='viterbi', covariance_type='diag', covars_prior=0.01,\n",
      "    init_params='stmcw', n_components=10, n_iter=10000, n_mix=5,\n",
      "    params='stmcw', random_state=None, startprob_prior=1.0, tol=0.0001,\n",
      "    transmat_prior=1.0, verbose=True)>\n",
      "Model name:  GMMHMM_models/GMMHMM_model_5_5.pkl\n",
      "<bound method _BaseHMM.score of GMMHMM(algorithm='viterbi', covariance_type='diag', covars_prior=0.01,\n",
      "    init_params='stmcw', n_components=5, n_iter=10000, n_mix=5,\n",
      "    params='stmcw', random_state=None, startprob_prior=1.0, tol=0.0001,\n",
      "    transmat_prior=1.0, verbose=True)>\n",
      "Model name:  GMMHMM_models/GMMHMM_model_10_8.pkl\n",
      "<bound method _BaseHMM.score of GMMHMM(algorithm='viterbi', covariance_type='diag', covars_prior=0.01,\n",
      "    init_params='stmcw', n_components=10, n_iter=10000, n_mix=8,\n",
      "    params='stmcw', random_state=None, startprob_prior=1.0, tol=0.0001,\n",
      "    transmat_prior=1.0, verbose=True)>\n",
      "Model name:  GMMHMM_models/GMMHMM_model_5_2.pkl\n",
      "<bound method _BaseHMM.score of GMMHMM(algorithm='viterbi', covariance_type='diag', covars_prior=0.01,\n",
      "    init_params='stmcw', n_components=5, n_iter=10000, n_mix=2,\n",
      "    params='stmcw', random_state=None, startprob_prior=1.0, tol=0.0001,\n",
      "    transmat_prior=1.0, verbose=True)>\n",
      "Model name:  GMMHMM_models/GMMHMM_model_5_8.pkl\n",
      "<bound method _BaseHMM.score of GMMHMM(algorithm='viterbi', covariance_type='diag', covars_prior=0.01,\n",
      "    init_params='stmcw', n_components=5, n_iter=10000, n_mix=8,\n",
      "    params='stmcw', random_state=None, startprob_prior=1.0, tol=0.0001,\n",
      "    transmat_prior=1.0, verbose=True)>\n",
      "Model name:  GMMHMM_models/GMMHMM_model_10_2.pkl\n",
      "<bound method _BaseHMM.score of GMMHMM(algorithm='viterbi', covariance_type='diag', covars_prior=0.01,\n",
      "    init_params='stmcw', n_components=10, n_iter=10000, n_mix=2,\n",
      "    params='stmcw', random_state=None, startprob_prior=1.0, tol=0.0001,\n",
      "    transmat_prior=1.0, verbose=True)>\n",
      "Model name:  GMMHMM_models/GMMHMM_model_15_2.pkl\n",
      "<bound method _BaseHMM.score of GMMHMM(algorithm='viterbi', covariance_type='diag', covars_prior=0.01,\n",
      "    init_params='stmcw', n_components=15, n_iter=10000, n_mix=2,\n",
      "    params='stmcw', random_state=None, startprob_prior=1.0, tol=0.0001,\n",
      "    transmat_prior=1.0, verbose=True)>\n",
      "Model name:  GMMHMM_models/GMMHMM_model_15_5.pkl\n",
      "<bound method _BaseHMM.score of GMMHMM(algorithm='viterbi', covariance_type='diag', covars_prior=0.01,\n",
      "    init_params='stmcw', n_components=15, n_iter=10000, n_mix=5,\n",
      "    params='stmcw', random_state=None, startprob_prior=1.0, tol=0.0001,\n",
      "    transmat_prior=1.0, verbose=True)>\n"
     ]
    }
   ],
   "source": []
  },
  {
   "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}