infiltration.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Populating the interactive namespace from numpy and matplotlib\n"
     ]
    }
   ],
   "source": [
    "from SimPEG import (\n",
    "    maps,\n",
    "    utils,\n",
    "    data_misfit,\n",
    "    regularization,\n",
    "    optimization,\n",
    "    inverse_problem,\n",
    "    directives,\n",
    "    inversion,\n",
    ")\n",
    "from SimPEG.electromagnetics.static import resistivity as DC, utils as DCutils\n",
    "import discretize\n",
    "import numpy as np\n",
    "import time\n",
    "import matplotlib.pyplot as plt\n",
    "from pymatsolver import Pardiso as Solver\n",
    "\n",
    "%pylab inline\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Create a mesh"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number of cells = 21700\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "n_elec = 48 # number of electrodes\n",
    "e_spacing = 0.5 # interelectrode spacing\n",
    "n_between = 4. # how many cells between each electrode\n",
    "\n",
    "csx =e_spacing / n_between \n",
    "csz = 0.0625\n",
    "\n",
    "# number of core cells \n",
    "ncx = n_elec * e_spacing /csx +1\n",
    "ncz = 88\n",
    "\n",
    "# number of padding cells \n",
    "npadx = 12\n",
    "npadz = 12\n",
    "\n",
    "# create mesh\n",
    "hx = [(csx, npadx, -1.5), (csx, ncx), (csx, npadx, 1.5)]\n",
    "hz = [(csz, npadz, -1.5), (csz, ncz)]\n",
    "mesh = discretize.TensorMesh([hx, hz],x0=\"CN\")\n",
    "\n",
    "print(\"number of cells =\", mesh.nC)\n",
    "\n",
    "mesh.plotGrid();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Create a survey\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/teddiherringnew/anaconda/envs/simpeg14/lib/python3.7/site-packages/SimPEG/electromagnetics/static/resistivity/IODC.py:190: UserWarning: code under construction - API might change in the future\n",
      "  warnings.warn(\"code under construction - API might change in the future\")\n"
     ]
    }
   ],
   "source": [
    "IO = DC.IO()\n",
    "xmin, xmax = -11.75,11.75 # survey end points\n",
    "zmin, zmax = -0.59375,-0.59375 # put electrodes below water layer at a cell centre\n",
    "endl = np.array([[xmin, zmin], [xmax, zmax]])\n",
    "survey = DCutils.generate_dcip_survey(endl, \"dipole-dipole\", dim=mesh.dim, a=0.5, b=0.5, n=10)\n",
    "survey = IO.from_ambn_locations_to_survey(\n",
    "    survey.locations_a,\n",
    "    survey.locations_b,\n",
    "    survey.locations_m,\n",
    "    survey.locations_n,\n",
    "    survey_type='dipole-dipole',\n",
    "    data_dc_type=\"volt\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Create a conductivity model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Water saturation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# load water content profile\n",
    "theta = np.load('theta2.npy',encoding = 'latin1')\n",
    "z_theta = np.load('z_theta2.npy',encoding = 'latin1')\n",
    "\n",
    "# interpolate onto mesh\n",
    "theta_mesh = np.interp(mesh.vectorCCy,z_theta,theta)\n",
    "theta2D = np.repeat(theta_mesh,mesh.nCx)\n",
    "\n",
    "# saturation model\n",
    "\n",
    "# background\n",
    "Sw_bg = 0.2\n",
    "Sw = ones(mesh.nC) * Sw_bg\n",
    "\n",
    "# infiltration x location\n",
    "Ix = np.where(np.logical_and(-2.5 < mesh.gridCC[:,0], mesh.gridCC[:,0]< 2.5))\n",
    "\n",
    "# infiltration with depth\n",
    "for i in range(mesh.nCy):\n",
    "    Iz = (np.where(mesh.gridCC[:,-1]==mesh.vectorCCy[i])) # infiltration x location\n",
    "    Sw[np.intersect1d(Ix, Iz)] = (theta_mesh[i]*0.4)\n",
    "\n",
    "# add 0.5 m water layer\n",
    "z_layer = -0.5\n",
    "Sw [mesh.gridCC[:, -1] > z_layer] = 2 # 2 is just a placeholder, b/c we'll say Sw>1.0 is just water\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Temperature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# load temperature curve from 1D conductive heat transport model, January profile\n",
    "z = np.load('z_infiltration.npy',encoding = 'latin1')\n",
    "T = np.load('T_infiltration.npy',encoding = 'latin1')\n",
    "\n",
    "# interpolate onto mesh\n",
    "T_mesh = np.interp(-mesh.vectorCCy,-z,T)\n",
    "T2D = np.repeat(T_mesh,mesh.nCx)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Petrophysical relationship (from Herring et al. 2019)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def ref2earth(T,a,por,m,n,Sw0,C0,d,popt=None):\n",
    "    \n",
    "    rho_f25 = (0.16*C0)**-1   \n",
    "\n",
    "    rho_0 = rho_f25 * (d*(0-25)+1)**-1\n",
    "    \n",
    "    if Sw0>1:  # shortcut so we can just get resistivity of water layer\n",
    "        return rho_f25 * (d*(T-25)+1)**-1\n",
    "    \n",
    "    else:\n",
    "        \n",
    "        if T > 0:\n",
    "            rho_f = rho_f25 * (d*(T-25)+1)**-1\n",
    "            R_earth = a * por**-m * Sw0**-n *  rho_f\n",
    "\n",
    "        else:\n",
    "            Sw = func_exp(T, *popt)\n",
    "            Sr = Sw/Sw0\n",
    "            rho_f = (Sr*(rho_0))\n",
    "            R_earth = a * por**-m * Sw**-n * rho_f\n",
    "\n",
    "        return R_earth\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Putting it all together..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# fitting parameters from Herring et al. 2019\n",
    "d = 0.0215\n",
    "por = 0.40\n",
    "Sw0 = 1.0\n",
    "a = 0.363\n",
    "m = 1.32 \n",
    "n = 2.29\n",
    "\n",
    "rho = []\n",
    "for i in range(mesh.nC):\n",
    "    rho = np.append(rho,ref2earth(T2D[i],a,por,m,n,Sw[i],1./0.5,d))\n",
    "    \n",
    "sig = 1./rho.copy()    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAswAAAIWCAYAAABDdNtTAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzde7RlZXnn+++vdoFcQoKkuAgUKVTUEDSY2kHTaryhYI6BJsYR7RzxkpMKZ4jpJJ2ohGGw2/aEhKSN8RJTGiKeVox90ghROlwyNHSng6FAVFBQVIQSTKW8JVEEqvZz/lirks1mrbXXdc+11v5+xpij9pzzne98dq3a8NRbz/u+qSokSZIkdbah6QAkSZKkaWbCLEmSJPVgwixJkiT1YMIsSZIk9WDCLEmSJPVgwixJkiT1YMIsSZKkxiS5OMmuJLd0uZ8kf5jkjiSfTvJjax2jCbMkSZKa9F7g9B73XwCc0D62AX+0BjE9hAmzJEmSGlNV1wHf6NHkTOB91XI9cGiSR61NdC0mzJIkSZpmxwB3Lzvf2b62Zjau5cskSZI0f3L66cXu3Z1v3njjrcD3ll3ZXlXbB+m+w7Ua4PmRmTBLkiRpNLt3s+HvdnS8tbSQ71XV4gi97wQ2Lzs/FrhnhP4GZkmGJEmSRlOwYanzMQZXAGe3V8t4KvDtqrp3LD33yRFmSZIkjSQMnxwnuRR4FrApyU7gAmA/gKp6F3Al8FPAHcB3gVeOHPCAZiph3rRpU23ZsqXpMCRJkhpz44037q6qw5uOY6UMWVVcVS9d5X4Brx6u9/GYqYR5y5Yt7NjRuT5GkiRpPUjylaZjeJgaW/nFVJqphFmSJEnTZ5SSjFlgwixJkqSRmTBLkiRJ3dTwNcyzwIRZkiRJI7EkQ5IkSerFSX+SJElSd44wS5IkSauY54S50a2xk5ye5PYkdyR5fZOxSJIkaUjtSX+djnnQWMKcZAF4B/AC4ETgpUlObCoeSZIkDWdfSUanYx40WZJxCnBHVX0JIMkHgTOBz3Z74EZaH4gkSZKmy7wkx500WZJxDHD3svOd7WsPkWRbkh1JdvAP/7BmwUmSJKlPNd8jzE0mzJ0Gix9W6VJV26tqsaoWOfzwNQhLkiRJgwjzXcPcZEnGTmDzsvNjgXsaikWSJEkjmJfR5E6aHGG+ATghyfFJ9gdeAlzRYDySJEkaxpyXZDQ2wlxVe5KcC1wFLAAXV9WtTcUjSZKk4bhxyQRV1ZXAlU3GIEmSpBG5NbYkSZLU27xM8OvEhFmSJEkjsSRDkiRJ6sWSDEmSJKk7R5glSZKkVVjDLEmSJHVjSYYkSZLU3byXZDS509/AthbUrVCPhkqP49Htdnv6bD/MM5Nub0zG1PQ7+mi/0vHALcCD7V+PX+VnetLtjWn0mNb7n3FjmtKY1uv33W6/devWrX38WK+5ed7pj6qamWPr1q1Ve6rqllVa3lKtdlX9tR/mmUm3NyZjavodfbSfwpCMacwxTWVQxmRM6/X7brffunVr9RHNmh4HPWFrnfKJ6ngAO5qOb9QjVbNTob24uFg7duyAPcB+PRo+yEOLTVZrP8wzk25vTMbU9Dv6aL9ygscUhLTm75j3mCpTGFRT7Y1pemJai3dMcUyLi4vs2LGj209nI77vhxfrpPfu6HjvE0/NjVW1uMYhjdVMlWQAsBe4fZU2t7fb9dt+mGcm3d6YjKnpdwwR0xSGZEwTimkqgzKm9RPTWrxj2mOaQsOWZCQ5PcntSe5I8voO938gyV8k+VSSW5O8chLx99T0EPcgx9Yf29r654jjV2l5fLXaPVj9tR/mmUm3NyZjavodfbSfwpCMacwxTWVQxmRM6/X7brefxpKMgx+/tZ72v6rjQY+SDGAB+CLwaGB/4FPAiSva/CbwO+2vDwe+Aey/lt/fbJVkZLF20Hm4X9Lam+c1N9XStSRDUmPC9JU4HPKExXryuzvnaP/zJ7vHm+QngDdW1Wnt8/MAquq3l7U5D9gMvBrYAlwDPK6q1mxKocvKSZIkaWRDDqIcA9y97Hwn8JQVbd4OXAHcAxwC/NxaJsswizXMkiRJmi7Vs4Z5U5Idy45ty57s9O9YK1Pv04CbgaOBk4G3J/n+SXwb3TSSMCd5cbtoeynJVP2TgiRJkgYTeibMu6tqcdmxfdmjO2mVW+xzLK2R5OVeCfz3arkD+DLwhMl9Nw/X1AjzLcDPANc19H5JkiSNS+8R5l5uAE5IcnyS/YGX0Cq/WO4u4LkASY4EHg98abzfQG+N1DBX1ecAEmeTSJIkzbp9I8yDqqo9Sc4FrqK1YsbFVXVrknPa998FvAl4b5LPtF/1uqraPa7Y++GkP0mSJI1s2JWTqupK4MoV19617Ot7gOePEtuoJpYwJ7kWOKrDrfOr6vIB+tkGbAM4juPGFJ0kSZLGpoYbYZ4VE0uYq+rUMfWzHdgOrXWYx9GnJEmSxmfYkoxZYUmGJEmSRjbPCXNTy8qdlWQn8BPAR5Nc1UQckiRJGoNq1TB3OuZBU6tkXAZc1sS7JUmSNF6WZEiSJEmrMGGWJEmSuoirZEiSJEm9mTBLkiRJPczLBL9OTJglSZI0EksyJEmSpFWYMEuSJEndOMIsSZIkdec6zJIkSdIqnPQnSZIkdWNJhiRJktSdJRmSJElSL44wS5IkSb1ZwyxJkiR1YUkGkGQReAZwNHAfcAtwbVV9Y4KxSZIkaRbMeUnGhl43k7wiyU3AecCBwO3ALuDpwDVJLkly3OTDlCRJ0rTaN8Lc6ZgHq40wHww8raru63QzycnACcBdg7w0yUXATwMPAF8EXllV3xqkD0mSJE2PeUmOO+k5wlxV7+iWLLfv31xVfzXEe68BTqqqJwGfpzWCLUmSpFlUrUl/nY550G8N8/HAa4Aty5+pqjOGeWlVXb3s9HrgZ4fpR5IkSc1z0l/Lh4E/Af4CGPdvx6uAP+t2M8k2YBvAcVguLUmSNI1MmOF7VfWHg3Sc5FrgqA63zq+qy9ttzgf2AO/v1k9VbQe2AyxmcU4G9iVJkubInK+S0W/C/NYkFwBXA/fvu1hVN3V7oKpO7dVhkpcDLwSeW1UmwpIkSTMqTH+9cpIfA/65qj7fPj8YeD5wb1Vd3+vZfhPmJwIvA57Dv5ZkVPt8mIBPB14HPLOqvjtMH5IkSZoew44wt/PCtwILwHuq6sIObZ4F/AGwH7C7qp45xKveBry83V+A/0VrpbdNSa6sqjd3e7DfhPks4NFV9cAQwXXyduARtNZyBri+qs4ZU9+SJElaS0OWZCRZAN4BPA/YCdyQ5Iqq+uyyNocC7wROr6q7khwxZJQ/UFV3tL9+BnBgVZ2ZZCNwMzBywvwp4FBam5aMrKoeO45+JEmS1LwRVsk4Bbijqr4EkOSDwJnAZ5e1+XfAf6+quwCqath8dM+yr58D/FW7vz1J9vZ6sN+E+UjgtiQ38NAa5qGWlZMkSdIcGX7S3zHA3cvOdwJPWdHmccB+ST4OHAK8tareN8S77kjyJlqjyduAlwAkOQwYS8J8wRBBSZpz0z7BQ5K0dnr8P2FTkh3Lzre3V0GD1uD0Sit72ghsBZ4LHAj8bZLr903eG8A24I206phfX1XXta/vD/xSrwd7JsxJUi1/vVqbweKVJEnSvFilJGN3VS12ubcT2Lzs/Fjgng5tdlfVd4DvJLkO+FFau0X3F19yEvBIWrtNP3L5CHVVfQ34Wq/ne26NDXwsyWuSPGTHkCT7J3lOkktozzaUtP502wbVY34OSepLtRLmTscqbgBOSHJ8kv1plUlcsaLN5cAzkmxMchCtko3PDRjhn9IaYd4LfGnAZ1ctyTid1k58l7a3x/4WcACtZT+uBt5SVTcP+lJJ82GeF6mXJPVv2El/7Ql35wJX0covL66qW5Oc077/rqr6XJK/BD5Na3nj91TVLQO+58fbS9O9EPgCrSXl+tYzYa6q79FaxuOdSfYDNgH3VdW3BnmJJEmS5tuw/ypVVVcCV6649q4V5xcBFw0dW6uM40bgO8BjBn2+30l/VNWDwL2DvkDS/DrIbYckSTALW2P/e1rL0+0B3rVK24fpO2GWpJW+/QNNRyBJmgYjrMO8Jqrqk8AnkxxJa0WM/zTI8ybMkiRJGtk0J8z7VNXfM2CyDKuskpHktB73XjzoyyRJkjSHhl8lYyastqzclUk+luSYDvfOm0RAkiRJmi1hvpenXC1h/jTwAeD6DiPKnXZmkSRJ0jq0nkeYq6reTWsrwtcm+dP2gtHw8G0LJUmStB6t85IMANp7df8E8Pe0Zhg+ZaJRSZIkaWbsWyVjXhPm1VbJ+Jeyi6raA7y+vdPKpcDhkwxMkiRJs2Ne6pU7WS1h/o8rL1TVx5NspbWGnSRJkta76d+4ZCQ9SzKq6sNdrn+zqi4c9qVJ3pTk00luTnJ1kqOH7UuSJEnNmveSjL5qmCfgoqp6UlWdDHwE+K2G4pAkSdKo5nzSXyM7/VXVPy47PRhX3JAkSZpZ07419qga2xo7yZuBs4FvA8/u0W4bsA3gOI5bm+AkSZI0kHme9Dexkowk1ya5pcNxJkBVnV9Vm4H3A+d266eqtlfVYlUtHu7CHJIkSdPHkozhVNWpfTb9APBR4IJJxSJJkqTJmfeSjEYm/SU5YdnpGcBtTcQhSZKk8XCEefwuTPJ4YAn4CnBOQ3FIkiRpVDXfNcxNrZLxoibeK0mSpPGb95KMxlbJkCRJ0vwwYZYkSZK6yJxvjW3CLEmSpJGZMEuSJEndOOlPkiRJ6s5Jf5IkSdIqTJglSZKkbpz0J0mSJHUX5ruGuZGtsSVJkjRfht0aO8npSW5PckeS1/do9+NJ9ib52XHG3Q8TZkmSJI2mhkuYkywA7wBeAJwIvDTJiV3a/Q5w1fiDX50JsyRJkkayb5WMIUaYTwHuqKovVdUDwAeBMzu0ew3w58CuccbdLxNmSZIkjWbIEWbgGODuZec729f+RZJjgLOAd40z5EE46U+SJEkj6zHpb1OSHcvOt1fV9n2PdWi/sqc/AF5XVXuTTs0nz4RZkiRJI1ll45LdVbXY5d5OYPOy82OBe1a0WQQ+2E6WNwE/lWRPVX146IAHZMIsSZKk0Qy/DvMNwAlJjge+CrwE+HcP6brq+H1fJ3kv8JG1TJah4RrmJL+epJJsajIOSZIkDW/YSX9VtQc4l9bqF58DPlRVtyY5J8k5Ew+8T42NMCfZDDwPuKupGCRJkjQew25cUlVXAleuuNZxgl9VvWK4t4ymyRHmtwCv5eGF3ZIkSZolw6+SMRMaGWFOcgbw1ar6VFOzHSVJkjQeq0z6m3kTS5iTXAsc1eHW+cBvAs/vs59twDaA4zhubPFJkiRpfEyYh1BVp3a6nuSJwPHAvtHlY4GbkpxSVV/r0M92YDvAYhYt35AkSZo2w6+SMRPWvCSjqj4DHLHvPMmdwGJV7V7rWCRJkjS6MPykv1ngOsySJEkamSPME1RVW5qOQZIkSSOwJEOSJEnqzlUyJEmSpFVYwyxJkiR1Y0mGJEmS1J0lGZIkSVIvjjBLkiRJ3TnCLEmSJK3CSX+SJElSN5ZkSJIkSd1ZkiFJkiStwoRZkiRJ6saSDEmSJKm74KQ/SZIkqSdHmCVJkqRuLMmQJEmSunOVDEmSJGkV81zDvKGJlyZ5Y5KvJrm5ffxUE3FIkiRpdGmXZHQ65kGTI8xvqarfa/D9kiRJGpN5SY47sSRDkiRJo3HS38Scm+RsYAfwH6rqm50aJdkGbGuf/nPI7WsV4Ig2AbubDkJrxs97/fCzXl/8vNePWfqsf6jpAFaa90l/qZpMhXaSa4GjOtw6H7ie1h/KAt4EPKqqXjWRQBqSZEdVLTYdh9aGn/f64We9vvh5rx9+1qPZumGx/na/HR3vPeKB3Njr9zbJ6cBbgQXgPVV14Yr7Pw+8rn36z8D/XVWfGkvgfZrYCHNVndpPuyTvBj4yqTgkSZI0YUOWZCRZAN4BPA/YCdyQ5Iqq+uyyZl8GnllV30zyAmA78JTRg+5fU6tkPGrZ6VnALU3EIUmSpNHtK8kYYpWMU4A7qupLVfUA8EHgzOUNqup/LyvdvR44dszhr6qpGubfTXIyrZKMO4FfaiiOSdredABaU37e64ef9fri571++FmP4EZuvGphKZu63D4gyfJ6je1Vte/3+xjg7mX3dtJ79PgXgP8xfKTDmVgNsyRJktRLkhcDp1XV/9U+fxlwSlW9pkPbZwPvBJ5eVV9fyzhdVk6SJElN2QlsXnZ+LHDPykZJngS8B3jBWifL0FANsyRJkgTcAJyQ5Pgk+wMvAa5Y3iDJccB/B15WVZ9vIEYT5nFL8uIktyZZSrK44t55Se5IcnuS05qKUePndu/rQ5LT2z+/dyR5fdPxaHKS3JnkM+2f585rZWlmJbk4ya4ktyy7dliSa5J8of3rI5uMcb2oqj3AucBVwOeAD1XVrUnOSXJOu9lvAT8IvLOpn0lrmMcsyQ8DS8AfA79eVTva108ELqU1G/Ro4FrgcVW1t6lYNT5J3gj8s9u9z6/20kefZ9nSR8BLVyx9pDmR5E5gsapmZSMLDSDJT9Jaz/d9VXVS+9rvAt+oqgvbfyF+ZFW9rlc/Wj8cYR6zqvpcVXXajfBM4INVdX9VfRm4g1byLGk2rLr0kaTZUFXXAd9YcflM4JL215cA/3ZNg9JUM2FeO52WTTmmoVg0Gecm+XT7n/r8p7z548/w+lLA1UluTLKt6WC0Jo6sqnsB2r8e0XA8miKukjGEXtt+V9Xl3R7rcM16mBmyynbvf0Rrm/d9273/PjBX273Ln+F15mlVdU+SI4BrktzWHpWUtA6ZMA+h322/V+hr2RRNL7d7X/f8GV5Hquqe9q+7klxGqyTHhHm+/X2SR1XVve0diXc1HZCmhyUZa+cK4CVJHpHkeOAE4O8ajklj4nbv68KqSx9pPiQ5OMkh+74Gno8/0+vBFcDL21+/HOj2L8ZahxxhHrMkZwFvAw4HPprk5qo6rb1EyoeAzwJ7gFe7QsZcWQ/bva9rVbUnyb6ljxaAi6vq1obD0mQcCVyWBFr/n/xAVf1lsyFpnJJcCjwL2JRkJ3ABcCHwoSS/ANwFvLi5CDVtXFZOkiRJ6sGSDEmSJKkHE2ZJkiSpBxNmSZIkqQcTZkmSJKkHE2ZJkiSpBxNmSZIkqQcTZknrUpLNSb6c5LD2+SPb5z/Uoe2BSf46ycIA/Z+b5JXjjFmS1AzXYZa0biV5LfDYqtqW5I+BO6vqtzu0ezWwsareOkDfBwF/U1VPHl/EkqQmOMIsaT17C/DUJL8CPB34/S7tfp72NrlJntUebf5Qks8nuTDJzyf5uySfSfIYgKr6LnBnklPW4huRJE2OCbOkdauqHgR+g1bi/CtV9cDKNkn2Bx5dVXcuu/yjwL8Hngi8DHhcVZ0CvAd4zbJ2O4BnTCZ6SdJaMWGWtN69ALgXOKnL/U3At1Zcu6Gq7q2q+4EvAle3r38G2LKs3S7g6PGFKklqggmzpHUrycnA84CnAr+a5FEdmt0HHLDi2v3Lvl5adr4EbFx274D285KkGWbCLGldShLgj2iVYtwFXAT83sp2VfVNYCHJyqS5H48DbhkpUElS40yYJa1XvwjcVVXXtM/fCTwhyTM7tL2a1qTAQT0NuHbI+CRJU8Jl5SRpFUmeDPxaVb1sks9IkqaTI8yStIqq+iTwsUE2LqE1WfANEwpJkrSGHGGWJEnSSJ5/2un19d27O9676aYbr6qq09c4pLHauHoTSZIkqbvdu3fzN9fv6HjvoP2zaY3DGTsTZkmSJI1snqsWTJglSZI0svlNl2csYd60aVNt2bKl6TAkSZIac+ONN+6uqsObjmOlJUeYp8OWLVvYsaNzfYwkSdJ6kOQrTcewUgHD5stJLgZeCOyqqpNW3Pt1WhtLHV5VnWcVrgGXlZMkSdJoqjXC3Onow3uBh62ikWQz8DzgrvEGOzgTZkmSJI2sqvOx+nN1HfCNDrfeAryWKSiPnqmSDEmSJE2fUUoyOklyBvDVqvpUkvF1PCQTZkmSJI2oZ/nFpiTLJ6Ftr6rt3RonOQg4H3j+GAMciQmzJEmSRtZjhHl3VS0O0NVjgOOBfaPLxwI3JTmlqr42UpBDMmGWJEnSSKrGt6xcVX0GOGLfeZI7gUVXyZAkSdJMqy7HapJcCvwt8PgkO5P8wuSiHE6jI8xJTgfeCiwA76mqC5uMR5IkScMZdmvsqnrpKve3DNXxGDU2wpxkAXgH8ALgROClSU5sKh5JkiQNb6k6H/OgyZKMU4A7qupLVfUA8EHgzAbjkSRJ0hBay8pVx2MeNFmScQxw97LzncBTej1w02fv4sAnnzvRoCRJkjSgPjcpmVVNJsydVqF+2G91km3ANgD2+74JhyRJkqRhzEv5RSdNJsw7gc3Lzo8F7lnZqL2w9XaADQcdMccfhSRJ0uyal/KLTppMmG8ATkhyPPBV4CXAv+v5xMb94LCj1yA0SZIk9atwhHkiqmpPknOBq2gtK3dxVd3a65ls3I8Nhz1qTeKTJElSv+Zngl8nja7DXFVXAlc2GYMkSZJGN7/p8oxtjb1hYYHvf+T3Nx2GJElSY77bdAAdjHNr7Gk0UwnzwsIChxx6cNNhSJIkNeZrTQfQxRzny7OVMEuSJGk6OcI8JRY2hkc+8sCmw5AkSdJK85svz1bCvHFhA4cdekDTYUiSJGkZl5WTJEmSeik3LpkaGxc28IOHOMIsSZI0bRxhnhLHH3YgF53xBLZfv5Ovf/fBru1+8KD92PbUYzni+/Zn1z8/sGr7YZ6ZdHtjMqam32FMxmRMxjStMa3X73tf+9u2bt3aM5CGzPMI84amAxjUkd+3P9ueemzPNtueeixHft/+LGxIX+2HeWbS7Y3JmJp+hzEZkzEZ07TGtBbvmOaYplH1OObBTI0wA2zYEI48ZH+OPKR76Ecesj8bkr7bD/PMpNsbkzE1/Q5jMiZjMqZpjWkt3jHtMU2fmutl5WZuhHlpqfj6dx7o2ebr33mApXYhTT/th3lm0u2NyZiafocxGZMxGdO0xrQW75j2mKZRVedjHmSW6k0WFxfr6uv+N//jtl380/17urY75BEbecETjuDQA/fjW/c9uGr7YZ6ZdHtjMqam32FMxmRMxjStMa3X73tf++f/5L9hx44dUzXU/MNPfHK974qPd7x3yqMPvbGqFtc2ovGaqYT5MSf+aP0/77+y6TAkSZIa85IfO3bqEtAffuKT65LLP97x3lMeM/sJ88zVMEuSJGn6zNAY7MBmKmHekPB9++3XdBiSJElaYZaqFgY1UwnzQuCQ/WcqZEmSpLlXwN45TpgbWSUjyYuT3JpkKclM17RIkiSte9Xa6a/TMQ+aGq69BfgZ4I8HeSgJB25cmExEkiRJGlJRc7NNycM1kjBX1eeglQAPYkPC/htnbuloSZKkuTbvJRlTXxCcZBuwDeBRx2xuOBpJkiR1Ms87/U0sYU5yLXBUh1vnV9Xl/fZTVduB7QAn/eiPlSPMkiRJ08UR5iFV1anj7jPAxgUTZkmSpKlSsLTUdBCTM/UlGZIkSZpuBSw56W+8kpwFvA04HPhokpur6rQ+nmO/hanaOl2SJEnU0DXMSS4GXgjsqqqT2tfeBJwJLAG7gFdU1T1jCnZgTa2ScRlw2cAPxpIMSZKkaTNiDfN7gbcD71t27aKqegNAkl8Gfgs4Z4QQR2L2KUmSpJENu3FJVV0HfGPFtX9cdnowNFvvMVM1zAE2brAkQ5IkadrUmFfJSPJm4Gzg28Czx9r5gBxhliRJ0kiqWiUZnQ5gU5Idy45t/fVZ51fVZuD9wLmTjH81szXCHFhwhFmSJGnq9Jj0t7uqFkfo+gPAR4ELRuhjJDOVMENMmCVJkqZMUWPduCTJCVX1hfbpGcBtY+t8CDOWMEuSJGka9TPBr5MklwLPolW6sZPWSPJPJXk8rWXlvkKDK2TAjCXMATZYdS1JkjRVip4lGb2frXpph8t/MlJAYzZbCXNgIZZkSJIkTZUaPmGeBTOVMEuSJGn6tDYuaTqKyZm5hDmOMEuSJE2VAvYOW8Q8A2YuYXaRDEmSpOljSYYkSZLURRXsXWo6ismZvYTZEWZJkqQpM951mCctyWFV9Y1+289cwhwzZkmSpKlSwJ7ZqmH+K+DJ/TbuK2FOsgg8AzgauA+4Bbh2kMxckiRJ82kGSzIGyu57JsxJXgH8MvBl4EbgduAA4OnA65LcAryhqu4aKtQBhdZazJIkSZouNWUlGUku6HKrgEcN0tdqI8wHA0+rqvu6BHIycAIwUMKc5CLgp4EHgC8Cr6yqbw3ShyRJkqZDqySj6Sge5p963Ns7SEc9E+aqescq928e5GXLXAOcV1V7kvwOcB7wuiH7kiRJUpOKqZv0V1X/pdu9JC8bpK9+a5iPB14DbFn+TFWdMcjLlj139bLT64GfHaYfSZIkNW8GNy4ZXw3zMh8G/gT4C2DcA+6vAv6s280k24BtAJuPO27Mr5YkSdI4zFa+zCmDNO43Yf5eVf3hIB0nuRY4qsOt86vq8nab84E9wPu79VNV24HtAFu3Ls7WRyFJkrQOVNVMjTBX1Z5B2vebML+1PdPwauD+ZS+7qUcgp/bqMMnLgRcCz61pm1YpSZKkgeyd42yu34T5icDLgOfwryUZ1T4fWJLTaU3ye2ZVfXeYPiRJkjQdprmGOcmhwIHA7qp6cJg++k2YzwIeXVUPDPOSDt4OPAK4Jq2Fla+vqnPG1LckSZLWUGvjkulJmNtLH78aeD6tJeS+Bxya5Mu0SoEvrqrv9dtfvwnzp4BDgV2DhdtZVT12HP1IkiRpOkxLvtzeeO/ZwCXAq5cP+CY5DjgT+EvgWf322W/CfCRwW5IbeGgN81DLykmSJGl+TFlJxiVV9d5ON9q7U78tydsH6bDfhLnb1oJrqmgN+UuSJGm6TMvGJf0sJjHoghM9E+YkqZa/Xq3NIC8dRQ22zrQkSZImbBaWlUtyNq15eQAfrKqu+4CstGGV+x9L8pp2vcfyF+6f5DlJLgFePli4kiRJmjdVnY8p8ivAi4CfAV47yIOrlWScTmsnvkvb22N/CzgAWKC1JvNbqurmgcMdxXT9xkuSJAlYmmNVnaIAACAASURBVPIRZuB9wP9La8D4kkEe7Jkwt5fbeCfwziT7AZuA+6rqW0MGKkmSpDlTNT01zN1U1R8kObj15WD7gPQ76Y/2Qs/3DhqcJEmS5tuUrZLxEEmeA/xI+/SWqvrYoH30nTBPi+n8KCRJkta3aSvJSHIUcBnwALCjfflFSd4MnFVVf99vX7OXME/XZyFJkqQq1nDRtH69FXhPVf3J8otJfhH4A+Cl/XbUc5WMJKf1uPfifl8iSZKk+bWvJKPT0aCtK5NlgKp6N/Djg3S02gjzlUmuA/7PqvrqinvnAf9tkJeNg+swS5IkTZ8pnPR3X4973xuko9XWYf408AHg+g4jyhnkRZIkSZpPVa0a5k5Hg+5MsrjyYpInAXcO0tFqI8xVVe9O8tfA+5P8FPDq9lIczfwOTN1fXiRJkjR9A8z8CnB0h+uHtO/1ra9Jf1X1+SQ/Afxn4JPtrQXXXGG+LEmSNG1GWVYuycXAC4FdVXVS+9pFwE/TWuHii8Ar+90HJMl+VfVgVX2x/exDY636m/b+In1brSTjX8ouqmpPVb0e+CXgUuCEQV4kSZKkOVWdJ/z1mUS/l9bu0stdA5xUVU8CPk9r7ly/fi7Jnyd5UZLD9l1MsiHJk5K8Abh2gP5WHWH+jysvVNXHk2yllTgPJcmbgDOBJWAX8IqqumfVB6dvT3JJkqR1r4ClIZO0qrouyZYV165edno98LMD9Pdfk3wC2Ab8pySHA/cDjwBuAN4PvHmQGFfbGvvDXa5/E7hwkBetcFFVvQEgyS8DvwWc08+Dw34YkiRJmpweo8mbkuxYdr69qrYP0PWrgD8bJJaq+gLwG8BvJFkADqiq7wzSx3KNbFxSVf+47PRgLE2WJEmaWVX02rhkd1U9bLWKfiQ5H9hDa1R4yNhqLzB0sgwN7vTX3pbwbODbwLObikOSJEmjG/cmJUleTmsy4HOr4W0EJ5YwJ7kWOKrDrfOr6vKqOh84P8l5wLnABV362UarBoVjNx9nDbMkSdKUKca75nKS04HXAc9sL2fcqIklzFV1ap9NPwB8lC4Jc7vGZTvAyT+2aLosSZI0bdoblwwjyaXAs2jVOu+klROeR2uS3jVJAK6vqr7mu01CIyUZSU5oF2MDnAHc1t+T5dbYkiRJU2jYKoCqemmHy38yUjBj1lQN84VJHk9rWbmv0OcKGZIkSZo+BSwtLTUdxsQ0tUrGi5p4ryRJkiZghJKMWdDYKhnDctKfJEnStBnvpL9pM3sJc9MBSJIk6SHKEWZJkiSpt3muApi9hHmOPwxJkqRZ5QjzFHFZOUmSpOliSYYkSZK0ChPmKVE13/UxkiRJs6moOU7SZiphliRJ0vSxJEOSJElahQnzFJnfj0KSJGk2VcHevfObpc1cwixJkqTp4wjzNJnfz0KSJGlGOelPkiRJ6spJf5IkSdIqyoR5erjTnyRJ0vRZWlpqOoSJmbmEWZIkSdOlquY6Yd7Q5MuT/HqSSrKpyTgkSZI0oupyzIHGRpiTbAaeB9w1yHNzPAFTkiRpZjnCPBlvAV7L3PzdQ5IkaZ2q1qS/Tsc8aGSEOckZwFer6lNJVmu7DdgGcMyxx61BdJIkSRpEMd81zBNLmJNcCxzV4db5wG8Cz++nn6raDmwH+NEnb7UiQ5IkaQq5cckQqurUTteTPBE4Htg3unwscFOSU6rqa5OKR5IkSRNSrsM8VlX1GeCIfedJ7gQWq2r3WsciSZKk0VmSIUmSJPXiCPNkVdWWpmOQJEnSaBxhnhKF6zBLkiRNIyf9SZIkSV1Uzc+ay52YMEuSJGlklmRIkiRJXblKxpSZ3+F+SZKkmVQMnaIluRh4IbCrqk5qX3sx8Ebgh4FTqmrHWOIc0oYmXy5JkqTZV7RKMjodfXgvcPqKa7cAPwNcN95IhzODI8ySJEmaKlWwtHfIR+u6JFtWXPscQHtX6MaZMEuSJGl03RPmTUmWl1Rsr6rtaxDR2JgwS5IkaUQ9R5h3V9XiWkYzbibMkiRJGk0B5SoZkiRJUhfD1zDPAlfJkCRJ0uiW9nY+VpHkUuBvgccn2ZnkF5KclWQn8BPAR5NcNeHoe3KEWZIkSaMZbZWMl3a5ddnwAY2XCbMkSZJGV/O7uZwJsyRJkkY03zXMJsySJEkaTQF75zdhbmTSX5I3Jvlqkpvbx081EYckSZLGoWBpT+djDjQ5wvyWqvq9Bt8vSZKkcRhh0t8ssCRDkiRJo3PS30Scm+RsYAfwH6rqm50aJdkGbGuf/vPmww64fa0CHNEmYHfTQWjN+HmvH37W64uf9/oxS5/1DzUdwMM5wjyUJNcCR3W4dT7wR8CbaJWIvwn4feBVnfqpqu3A9gmFOTFJdsz6vunqn5/3+uFnvb74ea8fftYjKkyYh1FVp/bTLsm7gY9MKg5JkiRN2nyPMDe1Ssajlp2eBdzSRBySJEkak1rqfMyBpmqYfzfJybQG8O8EfqmhOCZp5spINBI/7/XDz3p98fNeP/ysR1Df3XXV9276w01dbs9KbXhXqTme0ShJkiSNqpGSDEmSJGlWmDBLkiRJPZgwj1mSFye5NclSksUV985LckeS25Oc1lSMGj+3e18fkpze/vm9I8nrm45Hk5PkziSfaf8872g6Ho1XkouT7Epyy7JrhyW5JskX2r8+sskYNV1MmMfvFuBngOuWX0xyIvAS4EeA04F3JllY+/A0QW+pqpPbx5VNB6Pxav+8vgN4AXAi8NL2z7Xm17PbP8+uzTt/3kvr/8XLvR74q6o6Afir9rkEmDCPXVV9rqo67UZ4JvDBqrq/qr4M3AGcsrbRSRrBKcAdVfWlqnoA+CCtn2tJM6aqrgO+seLymcAl7a8vAf7tmgalqWbCvHaOAe5edr6zfU3z49wkn27/U5//lDd//BleXwq4OsmNSbY1HYzWxJFVdS9A+9cjGo5HU6SpdZhnWq9tv6vq8m6Pdbjmmn4zZFzbvWtm+TO8vjytqu5JcgRwTZLb2qOSktYhE+Yh9Lvt9wo7gc3Lzo8F7hlPRFoLbve+7vkzvI5U1T3tX3cluYxWSY4J83z7+ySPqqp72zsS72o6IE0PSzLWzhXAS5I8IsnxwAnA3zUck8bE7d7XhRuAE5Icn2R/WpN4r2g4Jk1AkoOTHLLva+D5+DO9HlwBvLz99cuBbv9irHXIEeYxS3IW8DbgcOCjSW6uqtOq6tYkHwI+C+wBXl1Ve5uMVWO1HrZ7X9eqak+Sc4GrgAXg4qq6teGwNBlHApclgdb/Jz9QVX/ZbEgapySXAs8CNiXZCVwAXAh8KMkvAHcBL24uQk0bt8aWJEmSerAkQ5IkSerBhFmSJEnqwYRZkiRJ6sGEWZIkSerBhFmSJEnqwYRZ0rqUZHOSLyc5rH3+yPb5D3Voe2CSv06yMED/5yZ55ThjliQ1w2XlJK1bSV4LPLaqtiX5Y+DOqvrtDu1eDWysqrcO0PdBwN9U1ZPHF7EkqQmOMEtaz94CPDXJrwBPB36/S7ufp73rV5JntUebP5Tk80kuTPLzSf4uyWeSPAagqr4L3JnklLX4RiRJk2PCLGndqqoHgd+glTj/SlU9sLJNexvsR1fVncsu/yjw74EnAi8DHldVpwDvAV6zrN0O4BmTiV6StFZMmCWtdy8A7gVO6nJ/E/CtFdduqKp7q+p+4IvA1e3rnwG2LGu3Czh6fKFKkppgwixp3UpyMvA84KnAryZ5VIdm9wEHrLh2/7Kvl5adLwEbl907oP28JGmGmTBLWpeSBPgjWqUYdwEXAb+3sl1VfRNYSLIyae7H44BbRgpUktQ4E2ZJ69UvAndV1TXt83cCT0jyzA5tr6Y1KXBQTwOuHTI+SdKUcFk5SVpFkicDv1ZVL5vkM5Kk6eQIsyStoqo+CXxskI1LaE0WfMOEQpIkrSFHmCVJkqQeNq7eRJIkSerutGcfXF//xt6O92789P1XVdXpaxzSWJkwS5IkaSS7v7GXT1x1bMd7+z3qi5vWOJyxM2GWJEnSSIriweo8wjwPTJglSZI0EhPmKbJp06basmVL02FIkiQ15sYbb9xdVYc3HcdyBTzIUtNhTMxMJcxbtmxhx44dTYchSZLUmCRfaTqGlQp4sEyYJUmSpK7mN102YZYkSdKIqooH5nhvDxNmSZIkjaQID5Kmw5gYE2ZJkiSNpFXDbMIsSZIkdVTAXkeYJUmSpM5aI8wbmg5jYkyYJUmSNJIiPFgLTYcxMY3+VSDJ6UluT3JHktc3GYskSZKGU4QHaqHjMQ8aG2FOsgC8A3gesBO4IckVVfXZpmKSJEnS4Fo7/c1HctxJkyPMpwB3VNWXquoB4IPAmQ3GI0mSpCFUhb21oeMxD5qsYT4GuHvZ+U7gKSsbJdkGbAPYfMwC9917/NpEJ0mSpL44wjw5ndYeedgWMVW1vaoWq2px0w/O7wchSZI0q1qT/jZ2PEaR5KIktyX5dJLLkhw6ppAH0uQI805g87LzY4F7ej0QYMMcr/EnSZI0i/ZN+puAa4DzqmpPkt8BzgNeN4kX9dLkCPMNwAlJjk+yP/AS4IoG45EkSdKQlmpDx2MUVXV1Ve1pn15Pa4B1zTU2wtz+m8K5wFXAAnBxVd3aVDySJEkazlLvEeZNSXYsO99eVduHeM2rgD8b4rmRNbpxSVVdCVzZ/xNhIfMx21KSJGleVNGrXnl3VS12u5nkWuCoDrfOr6rL223OB/YA7x811mG4058kSZJGMspOf1V1aq/7SV4OvBB4blU9bIGItWDCLEmSpJFMamvsJKfTmuT3zKr67thf0CcTZkmSJI1s72TWkng78AjgmiQA11fVOZN4US8zlTC3lpWzhlmSJGmaTGqEuaoeO/ZOhzBTCbMkSZKmT8FEEuZpYcIsSZKkkVSFB5fmN62c3+9MkiRJa6JorcU8r2YrYd54End9/TLecMaFfO3Lu7o2O+r4I3jTFa9n8+OP5u7b71m1/TDPTLq9MRlT0+/op/277/pfDznfuHAcR2x6H/ttfAwP7vkiu3afzZ69d3V9x6TbG9PoMf3icU/v2H69/Bk3pumMab1+3/vab9161taegTSgCA8uzW9JRhpazm4oi4uL9YlPfIK7b7uHX3zir3Vt9+7P/Bc2P+FoFhYW2Lt376rth3lm0u2NyZiafkc/7VcmzEcf+XH22/hYkgWq9vLgnju45++f1fUdk25vTKPH1C1hXi9/xo1pOmNar9/3vvZPecpT2LFjx1QN5x71I4fV2R94bsd7F538/93Ya+OSWTBbI8zAwsICmx9/dM82mx/f+sPXb/thnpl0e2MypqbfMUxM+218DEnrmWSB/TY+ptH2xjS5mNbrn3Fjmo6Y1uId0x7TtGnt9DedsY3DzK3RtnfvXu6+/Z6ebe6+/R727t3bd/thnpl0e2Mypqbf0U/7hTz02LPni1S1nqnay549X3xYm7Vsb0yjx7SWf57W+h3GNLsxrcU7pj2maVOEPUsLHY95MFsJc8Hdt7VqfHp5wxkXcvdt97B3z96+2g/zzKTbG5MxNf2OYWL6h6+fzZ49d1C1hz177uAfvn52o+2NaXIxrdc/48Y0HTGtxTumOaZptZd0PObBTNUwf38Oq6dseF7TYUhqu/iu65oOQRP2quN+sukQJK1w7dKHpq4m+PATN9VZ7/s/Ot5794+/b+riHdTM1TBLkiRpurRqmGercGEQJsySJEkayb4a5nllwixpaAtzUpsmSRrdPG9c0sjYeZIXJ7k1yVKSma5pkSRJWu8KXCVjAm4BfgZwxpAkSdKMqwp7akPHYx40UpJRVZ8DSAYfus+G+R3ul6Rp439zpSm01HQAD9caYZ6P5LiTqa9hTrIN2AZwAAc1HI0kSZJWak36M2EeWJJrgaM63Dq/qi7vt5+q2g5sh9Y6zGMKT5IkSWM0z5P+JpYwV9Wpk+pbkiRJ06PKkgxJkiSpK0syJiDJWcDbgMOBjya5uapOayIWScOb3/80SpIGtXdOVsTopKlVMi4DLmvi3ZIkSRqvKlgqa5glSZKkLsJeSzIkSZKkzgpMmCVJkqSuCvZakiFJkiR1VpZkTJEEMr8fhiRNHf+bK6lPNcfby81WwixJkqSpUwVLjjBLkiRJ3e1dsoZZkiRJ6qjIREaYk7wJOBNYAnYBr6iqe8b+olXM79i5JEmS1kx1OUZ0UVU9qapOBj4C/NboXQ7OEWZJkiSNpqAmUJJRVf+47PRgxpKDD86EWZIkSSNb6p4wb0qyY9n59qra3m+/Sd4MnA18G3j28BEOz4RZkiRJI6mC6l7DvLuqFrvdTHItcFSHW+dX1eVVdT5wfpLzgHOBC0YOeEAmzJIkSRpZLQ35XNWpfTb9APBRpjVhTrIIPAM4GrgPuAW4tqq+McHYJEmSNBNCTWBr7CQnVNUX2qdnALcN0cc/AsuDW14Hnao6ZLU+eibMSV4B/DLwZeBG4HbgAODpwOuS3AK8oaruGix0SZIkzY0JTfoDLkzyeFrLyn0FOGfQDqrq+5OcDPwkrWT5b6rqpkH6WG2E+WDgaVV1X6eb7ZefAAyUMCe5CPhp4AHgi8Arq+pbg/QhSZKkKTKBEeaqetGofST5VeAVwIdpJcx/muSSqvov/fbRcx3mqnpHt2S5ff/mqvqrfl+2zDXASVX1JODzwHlD9CFJkqRpsdTlaN6rgKdU1QVV9UbgKcArB+mg3xrm44HXAFuWP1NVZwzysmXPXb3s9HrgZ4fpR5IkSVOgmMgI85gUsLDsfIEB13Pud5WMDwN/AvwF4/+7wquAPxtzn5IkSVpDw66SsQbeDVyf5DJaifKL2tf61m/C/L2q+sNBOl5tTb12m/OBPcD7e/SzDdgGcAAHDRKCJEmS1kgmM+lvZFX1tiTX0VrxLcDPV9WnBumj34T5rUkuAK4G7l8WQNcZhqutqZfk5cALgedWVddh8fZOMNsBvn/DDzayHaIkSZJ6qMCUJsxt+9HKe4sh9iHp94EnAi8DnsO/lmRU+3xgSU4HXgc8s6q+O0wfkiRJmiJTVpKR5P20qhT+A601nD9CK399d5LLqupN/fbVb8J8FvDoqnpg0GC7eDvwCOCaJADXV9XA6+pJkiRpSkxfHcCJVfWdJC8FTq6q+wGS/DbwKWDsCfOngEOBXYNG2klVPXYc/UiSJGkK1FTWMD8iyYG0NjzZj38tK96fAfcQ6TdhPhK4LckNPLSGeahl5SRJkjRnpm+E+b3AdbS2096R5KPt6z8N7Biko34T5gsG6VSSJEnry7SNMFfV7ya5GtgKfHLZrXcN2lfPhDlJquWvV2sz6IslSZI0J4ppHGGmqm4Gbh61n9VGmD+W5M+By6vqX2o9kuwPPB14OfAxWkPektaZKZsQLUlqUKb8fwpJDgUOBHZX1YODPLtawnw6rZ34Lm1vj/0t4ABaWwpeDbylnblLWof2TuNwgiSpGVOYMCc5GXg18HxaEd4HHJrky7Q2zru4qr63Wj89E+Z2B+8E3plkP2ATcF9VfWvE+CVJkjQnMoWrZCR5BfBs4BLg1cuXR05yHHAm8JfAs1brq++dTtpD1/cOGKukOfZPSxuaDkGSNCWmsCTjkqp6b6cb7VLjtyV5ez8dDbw1oCRJkvQwU1alt29RiiQHAy8FfhC4Bbh23yYm/S5cYcIsaWi/esKzmg5BEzdl/weUNJ1qKkeY97kU+Cfg07QS599L8pKq+lS/Hay2rNxpVXVVl3svrqr/Nki0kuZLLZlMSZJapjhhPr6qnrjvJMlW4A+BZ/bbwWoFiFcm+ViSYzrcO6/fl0iSJGnOVZejef+Q5HH7TqrqRloLWfRttZKMTwMfAK5P8msrRpSnayqkJEmSmlGtlTKm1K8Clye5DLgJeCLwhEE6WG2Euarq3cBzgdcm+dMkB+27N2i0kiRJmlNLXY6GtWuV/w1wN3AarT1FBhr47WvSX1V9PslPAP8Z+GSSsweMVZIkSXMqTF8Nc5L99u3oV1XfBP5o2b3vrmzTy2ojzP+SfVfVnqp6PfBLtGYbnjBE7JIkSZo37VUyOh0N+rkkf57kRUkO23cxyQbgsiRvAK7tp6PVEub/uPJCVX0c2Aq8uf94JUmSNM+mLWGuqv8KvB54KvA/k+xKcjfwNeC3gS/S2glwVattjf3hLte/CVw4SNDLJXkTre0Il4BdwCuq6p5h+5MkSVLDpnB2W1V9AfgN4DeSLAAHVNV3Bu2nqX1tL6qqJ1XVycBHgN9qKA5JkiSNajpLMh6iqvYOkyxDQzv9VdU/Ljs9mKn8O4kkSZL6NU3J8bg1tjV2kjcDZwPfpkf9SJJtwDaAAzioWzNJkiQ1pZiKJeQmZWIlGUmuTXJLh+NMgKo6v6o2A+8Hzu3WT1Vtr6rFqlrcLwdMKlxJkiQNKbQ2Lul0jKX/5NeTVJKBdugbl4mNMFfVqX02/QDwUeCCScUiSZKkyZrUTn9JNgPPA+6azBtW18ikvyTL13A+A7itiTgkSZI0JpPb6e8twGtpcM5bUzXMFyZ5PK3fxq8A5/T1VBXUHBfISJIkzaLqOelvU5Idy863V9X2frpNcgbw1ar6VDLQbtZj1dQqGS9q4r2SJEmajB4J8+6qWuz6XHItcFSHW+cDvwk8f+TgRtTYKhmSJEmaH8PWMHeb95bkicDxwL7R5WOBm5KcUlVfGzLMoZgwS5IkaTQTWFauqj4DHLHvPMmdwGJV7R7vm1ZnwixJkqSRBDcukSRJkroryNJkF7Goqi0TfUEPJsySJEkamSPMkiRJUg+T2rhkGpgwS5IkaTS912GeeSbMkiRJGomT/iRJkqRVTHrSX5NMmCVJkjSaah9zyoRZkiRJI8vepiOYHBPm/7+9+wu1rCzjOP79OWJmFCjT6JSiVkpYkoIMgkUG/r/IDARDRCqcLjSyKDElErxIShkiUppC8qZsbiQpSWckErrRqawZK2vKYXK0JsmgKDRnni7OOrE57b1nzllr73XO3t8PLPZea7177efwss55eM+7nleSJEntTKEOc59MmCVJktSaD/1JkiRJI6TKEWZJkiRpHBcukSRJksaY5SkZR/X55Uk+m6SSrO8zDkmSJLVQwMEavs2A3kaYk5wCXAzs6ysGSZIkdWOW5zD3OcK8BbiFmS5zLUmSNB9Sw7dZ0MsIc5IPAvur6pdJDtd2M7AZ4FiOm0J0kiRJWo5Yh3llkuwAThpy6nbgNuCSI7lOVW0FtgK8KSfMbk9IkiStYZmR+crDTCxhrqqLhh1PcjZwOrA4unwy8PMkm6rqz5OKR5IkSRNSBY4wd6eqdgEbFveT7AXOq6qXph2LJEmSuuGUDEmSJGmUmu06zL0nzFV1Wt8xSJIkqSVHmCVJkqTRcmh2h5hNmCVJktRKqqySIUmSJI1VJsySJEnScAU4wixJkiSN5hxmSZIkaZQqMGGWJEmSRpvlh/6O6jsASZIkzYCq4VsLSe5Isj/J0812RUfRLosjzJIkSWqnCg5ObErGlqq6e1IXPxImzJIkSWpvhucwOyVDkiRJ7VTBwYPDN1ifZOfAtnmZV78pya+S3J/k+AlEf1iOMEuSJKm90fOVX6qq80adTLIDOGnIqduB+4A7Waj0fCdwD/CxdoEunwmzJEmS2ilWPIe5qi46knZJvgn8YEVf0pJTMiRJktRSU4d52NZCko0Du1cBu1tdcIUcYZYkSVI7xeJ85a59Ock5zTfsBT4xiS85nF4S5iR3ADcAf20O3VZVj/QRiyRJktqaTFm5qrqu84uuQJ8jzL3X1JMkSVIHCqpmt6ycUzIkSZLU3uQWLuldnw/99V5TT5IkSR0YX4d5zZvYCHNXNfWa4taLBa7/uf217z3bfbQTsR54qe8gNDX29/ywr+eL/T0/1lJfn9p3AP+vqBlJjoeZWMLcVU29qtoKbO0qrmlJsnNckW7NFvt7ftjX88X+nh/2dUsFHBq5cMma11eVjI1V9WKz21tNPUmSJLVX4AjzBKyKmnqSJEnqQDklo3OrpabehK25aSRqxf6eH/b1fLG/54d93cI/ePnRHYe2rR9xeq3MDR8pVbM730SSJElqq8+ycpIkSdKqZ8LcsSRXJ3kmyaEk5y059/kke5I8m+TSvmJU95LckWR/kqeb7Yq+Y1L3klzW3L97ktzadzyanCR7k+xq7uedfcejbjVrQBxIsnvg2AlJtif5ffPqGhH6HxPm7u0GPgw8MXgwyVnANcC7gMuAe5Osm354mqAtVXVOsz3SdzDqVnO/fh24HDgL+EhzX2t2faC5ny01Nnu+zcLf4kG3Ao9X1RnA482+BJgwd66qflNVwxZXuRJ4sKpeqarngD3ApulGJ6mFTcCeqvpjVb0KPMjCfS1pjamqJ4C/LTl8JfBA8/4B4ENTDUqrmgnz9LwV+NPA/vPNMc0Ol3ufbd7D86WAx5L8rFlxVrPvxMU1IprXDT3Ho1WkrzrMa9q4Zb+r6vujPjbkmCVK1pCulnvXmuU9PF8uqKoXkmwAtif5bTMqKWkOmTCvwJEu+73E88ApA/snAy90E5Gmoavl3rVmeQ/Pkap6oXk9kOQhFqbkmDDPtr8srkScZCNwoO+AtHo4JWN6HgauSfK6JKcDZwBP9hyTOtL8cl3kcu+z6SngjCSnJzmGhYd4H+45Jk1AkjckeePie+ASvKfnwcPA9c3764FR/zHWHHKEuWNJrgK+BrwZ+GGSp6vq0qp6Jsk24NfAa8CNVTW7a0jOH5d7n3FV9VqSm4BHgXXA/VX1TM9haTJOBB5KAgt/J79TVT/qNyR1Kcl3gQuB9UmeB74I3AVsS/JxYB9wdX8RarVxpT9JkiRpDKdkSJIkSWOYMEuSJEljmDBLkiRJY5gwS5IkSWOYMEuSJEljmDBLkiRJY5gwS5pLSU5J8lySE5r945v9U4e0fX2SnyRZt4zr35Tko13GLEnqh3WYJc2tJLcA76iqzUm+Aeytqi8NaXcjcHRVfXUZ1z4OdGpcSgAAAXhJREFU+GlVndtdxJKkPjjCLGmebQHOT3Iz8F7gnhHtrqVZJjfJhc1o87Ykv0tyV5JrkzyZZFeStwNU1b+AvUk2TeMHkSRNjgmzpLlVVf8BPsdC4nxzVb26tE2SY4C3VdXegcPvAT4FnA1cB5xZVZuAbwGfHGi3E3jfZKKXJE2LCbOkeXc58CLw7hHn1wN/X3Lsqap6sapeAf4APNYc3wWcNtDuAPCW7kKVJPXBhFnS3EpyDnAxcD7w6SQbhzT7N3DskmOvDLw/NLB/CDh64NyxzeclSWuYCbOkuZQkwH0sTMXYB3wFuHtpu6p6GViXZGnSfCTOBHa3ClSS1DsTZknz6gZgX1Vtb/bvBd6Z5P1D2j7GwkOBy3UBsGOF8UmSVgnLyknSYSQ5F/hMVV03yc9IklYnR5gl6TCq6hfAj5ezcAkLDwt+YUIhSZKmyBFmSZIkaQxHmCVJkqQxTJglSZKkMUyYJUmSpDFMmCVJkqQxTJglSZKkMf4L8R1sNBY1KFIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 936x648 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot model inputs and output\n",
    "fig, ax = plt.subplots(3,1, figsize=[13,9])\n",
    "\n",
    "dat = mesh.plotImage(Sw,ax=ax[0],pcolorOpts={\"cmap\": \"cool_r\"},clim=[min(Sw),1])\n",
    "cbar = colorbar(dat[0],ax=ax[0])\n",
    "cbar.set_label(r'S$_w$', rotation=270,labelpad=20)\n",
    "\n",
    "dat = mesh.plotImage(T2D,ax=ax[1],pcolorOpts={\"cmap\": \"Blues_r\"})\n",
    "cbar = colorbar(dat[0],ax=ax[1])\n",
    "cbar.set_label(r'T ($\\degree$C)', rotation=270,labelpad=20)\n",
    "\n",
    "dat = mesh.plotImage(log(sig),ax=ax[2],pcolorOpts={\"cmap\": \"viridis\"})\n",
    "cbar = colorbar(dat[0],ax=ax[2])\n",
    "cbar.set_label(r'log($\\sigma$)', rotation=270,labelpad=20)\n",
    "\n",
    "for a in ax:\n",
    "    a.set_ylim([-5,0.5])\n",
    "    a.set_xlim([min(survey.electrode_locations.T[0]),max(survey.electrode_locations.T[0])])\n",
    "    a.plot(survey.electrode_locations.T[0],survey.electrode_locations.T[1],'wo',markersize=4)\n",
    "    a.set_xlabel('X (m)')\n",
    "    a.set_ylabel('Z (m)')\n",
    "    a.set_aspect(1)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Simulate data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# calculate geometric factor for buried electrodes using source images\n",
    "\n",
    "def get_d(loc1,loc2):\n",
    "    d = np.sqrt((loc2[0]-loc1[0])**2 + (loc2[1]-loc1[1])**2 )\n",
    "    return d\n",
    "\n",
    "A = survey.a_locations.copy()\n",
    "B = survey.b_locations.copy()\n",
    "M = survey.m_locations.copy()\n",
    "N = survey.n_locations.copy()\n",
    "\n",
    "AP = survey.a_locations.copy()\n",
    "BP = survey.b_locations.copy()\n",
    "MP = survey.m_locations.copy()\n",
    "NP = survey.n_locations.copy()\n",
    "\n",
    "AP.T[1] = -AP.T[1]\n",
    "BP.T[1] = -BP.T[1]\n",
    "MP.T[1] = -MP.T[1]\n",
    "NP.T[1] = -NP.T[1]\n",
    "\n",
    "geo_fact = []\n",
    "\n",
    "for i in range(survey.nD):\n",
    "    ramP = get_d(AP[i],M[i])\n",
    "    ranP = get_d(AP[i],N[i])\n",
    "    rbmP = get_d(BP[i],M[i])\n",
    "    rbnP = get_d(BP[i],N[i])\n",
    "    ram = get_d(A[i],M[i])\n",
    "    ran = get_d(A[i],N[i])\n",
    "    rbm = get_d(B[i],M[i])\n",
    "    rbn = get_d(B[i],N[i])    \n",
    "    \n",
    "    geo_fact.append((+ 1./ram + 1./ramP\n",
    "                     - 1./rbm - 1./rbmP \n",
    "                     - 1./ran - 1./ranP\n",
    "                     + 1./rbn + 1./rbnP)\n",
    "                   /(4. * pi) )\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "expmap = maps.ExpMap(mesh)\n",
    "mapping = expmap \n",
    "simulation = DC.Simulation2DCellCentered(\n",
    "    mesh, survey=survey, sigmaMap=mapping, solver=Solver, bc_type=\"Neumann\"\n",
    ")\n",
    "\n",
    "# overwrite geometric factor\n",
    "IO.G = geo_fact\n",
    "\n",
    "rel_err = 0.01\n",
    "np.random.seed(1)\n",
    "data = simulation.make_synthetic_data(log(sig), relative_error=rel_err, add_noise=True)\n",
    "IO.plotPseudoSection(data=1./(data.dobs/geo_fact), data_type=\"apparent_conductivity\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Smooth inversion"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "SimPEG.SaveOutputEveryIteration will save your inversion progress as: '###-InversionModel-2021-07-12-18-59.txt'\n",
      "SimPEG.SaveModelEveryIteration will save your models as: './###-InversionModel-2021-07-12-18-59.npy'\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  2.93e+06  0.00e+00  2.93e+06    2.59e+05      0              \n",
      "   1  1.00e+00  3.56e+05  5.79e-01  3.56e+05    4.04e+04      0              \n",
      "   2  1.00e+00  1.21e+05  2.32e+00  1.21e+05    7.62e+03      0   Skip BFGS  \n",
      "   3  1.00e+00  9.94e+04  4.76e+00  9.94e+04    5.94e+03      0   Skip BFGS  \n",
      "   4  1.00e+00  9.11e+04  6.31e+00  9.11e+04    3.49e+03      0              \n",
      "   5  1.00e+00  8.49e+04  1.06e+01  8.49e+04    4.01e+03      0   Skip BFGS  \n",
      "   6  1.00e+00  8.21e+04  1.05e+01  8.21e+04    3.22e+03      0              \n",
      "   7  1.00e+00  8.00e+04  1.27e+01  8.00e+04    3.31e+03      0              \n",
      "   8  1.00e+00  7.45e+04  2.07e+01  7.45e+04    3.84e+03      0              \n",
      "   9  1.00e+00  6.93e+04  3.72e+01  6.93e+04    4.03e+03      0   Skip BFGS  \n",
      "  10  1.00e+00  6.62e+04  3.45e+01  6.62e+04    2.98e+03      0              \n",
      "  11  1.00e+00  6.39e+04  3.94e+01  6.39e+04    3.41e+03      0              \n",
      "  12  1.00e+00  6.14e+04  5.27e+01  6.15e+04    3.55e+03      0              \n",
      "  13  1.00e+00  5.91e+04  5.33e+01  5.92e+04    3.66e+03      0              \n",
      "  14  1.00e+00  5.74e+04  6.06e+01  5.74e+04    3.87e+03      0              \n",
      "  15  1.00e+00  5.56e+04  6.56e+01  5.56e+04    3.69e+03      0              \n",
      "  16  1.00e+00  5.38e+04  7.72e+01  5.39e+04    4.46e+03      0              \n",
      "  17  1.00e+00  5.03e+04  9.04e+01  5.04e+04    4.32e+03      0   Skip BFGS  \n",
      "  18  1.00e+00  4.80e+04  1.00e+02  4.81e+04    4.70e+03      0              \n",
      "  19  1.00e+00  4.09e+04  1.27e+02  4.10e+04    4.68e+03      0   Skip BFGS  \n",
      "  20  1.00e+00  3.73e+04  1.37e+02  3.74e+04    3.62e+03      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 3.5558e+03 <= tolF*(1+|f0|) = 2.9317e+05\n",
      "1 : |xc-x_last| = 1.5399e+01 <= tolX*(1+|x0|) = 4.1226e+01\n",
      "0 : |proj(x-g)-x|    = 3.6150e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 3.6150e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =      20    <= iter          =     20\n",
      "------------------------- DONE! -------------------------\n",
      "Mean percent data error =  9.6276840408851\n",
      " Time: 20 mins\n"
     ]
    }
   ],
   "source": [
    "from datetime import datetime\n",
    "np.random.seed(2)\n",
    "\n",
    "# simulation\n",
    "sim = DC.Simulation2DCellCentered(\n",
    "    mesh, survey=survey, sigmaMap=mapping, solver=Solver, bc_type=\"Neumann\",miniaturize=True)\n",
    "\n",
    "# reference model \n",
    "m0 = np.ones(mapping.nP)*np.log(1./mean(data.dobs/IO.G))\n",
    "\n",
    "# data misfit\n",
    "dmisfit = data_misfit.L2DataMisfit(simulation=sim, data=data)\n",
    "\n",
    "# regularization mapping\n",
    "regmap = maps.IdentityMap(nP=mesh.nC)\n",
    "\n",
    "# inversion components\n",
    "reg = regularization.Tikhonov(mesh, mapping=regmap)\n",
    "reg.alpha_x = 1.\n",
    "opt = optimization.InexactGaussNewton(maxIter=20)\n",
    "invProb = inverse_problem.BaseInvProblem(dmisfit, reg, opt) \n",
    "invProb.beta = 1.\n",
    "target = directives.TargetMisfit()\n",
    "saveIter = directives.SaveOutputEveryIteration()\n",
    "saveModel = directives.SaveModelEveryIteration()\n",
    "inv = inversion.BaseInversion(invProb, directiveList=[saveIter,saveModel,target])\n",
    "\n",
    "# run\n",
    "start_time = time.time()\n",
    "dtn_standard = datetime.now().strftime(\"%Y-%m-%d-%H-%M\")\n",
    "mopt_smooth = inv.run(m0)\n",
    "dpred = sim.dpred(mopt_smooth)\n",
    "mpe = (mean(abs(data.dobs-dpred)/abs(data.dobs))*100)\n",
    "\n",
    "print('Mean percent data error = ',mpe)\n",
    "print ('\\r Time: {:.0f} mins'.format((time.time() - start_time)/60.))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x432 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# estimated model \n",
    "\n",
    "fig, ax = plt.subplots(2,1, figsize=[13,6])\n",
    "\n",
    "dat = mesh.plotImage(log(sig),ax=ax[0],clim=[min(log(sig)),max(log(sig))])\n",
    "ax[0].set_title(r'm$_{true}$')\n",
    "\n",
    "dat = mesh.plotImage(log(mapping*mopt_smooth),ax=ax[1],clim=[min(log(sig)),max(log(sig))])\n",
    "ax[1].set_title(r'm$_{est}$, standard inversion')\n",
    "\n",
    "\n",
    "for a in ax:\n",
    "    cbar = colorbar(dat[0],ax=a)\n",
    "    cbar.set_label(r'log($\\sigma$)', rotation=270,labelpad=20,fontsize=14)\n",
    "    a.set_aspect(1)\n",
    "    a.set_ylim([-5,0.5])\n",
    "    a.set_xlim([min(survey.electrode_locations.T[0]),max(survey.electrode_locations.T[0])])\n",
    "    a.plot(survey.electrode_locations.T[0],survey.electrode_locations.T[1],'wo',markersize=4)\n",
    "    a.set_xlabel('')\n",
    "    a.set_ylabel('Z (m)')     \n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Hybrid inversion"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Set up mapping function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "import scipy.sparse as sp\n",
    "from SimPEG.maps import BaseParametric \n",
    "\n",
    "class ParametricLayer(BaseParametric):\n",
    "    \"\"\"\n",
    "        Parametric Layer Space\n",
    "\n",
    "        .. code:: python\n",
    "\n",
    "            m = [\n",
    "                val_background,\n",
    "                layer_z,\n",
    "                func_coeffs \n",
    "            ]\n",
    "\n",
    "        **Required**\n",
    "\n",
    "        :param discretize.base.BaseMesh mesh: SimPEG Mesh, 2D or 3D\n",
    "\n",
    "        **Optional**\n",
    "\n",
    "        :param float slopeFact: arctan slope factor - divided by the minimum h\n",
    "                                spacing to give the slope of the arctan\n",
    "                                functions\n",
    "        :param float slope: slope of the arctan function\n",
    "        :param numpy.ndarray indActive: bool vector with\n",
    "\n",
    "    \"\"\"\n",
    "\n",
    "    def __init__(self, mesh, **kwargs):\n",
    "        super(ParametricLayer, self).__init__(mesh, **kwargs)\n",
    "\n",
    "    @property\n",
    "    def nP(self):\n",
    "        return 4\n",
    "\n",
    "    @property\n",
    "    def shape(self):\n",
    "        if self.indActive is not None:\n",
    "            return (sum(self.indActive), self.nP)\n",
    "        return (self.mesh.nC, self.nP)\n",
    "\n",
    "    def mDict(self, m):\n",
    "        return {\n",
    "            \"val_background\": m[0],\n",
    "            \"layer_z\": m[1],\n",
    "            \"func_coeffs\": [m[2],m[3]]\n",
    "        }\n",
    "\n",
    "    def _atanLayer(self, mDict):\n",
    "        if self.mesh.dim == 2:\n",
    "            z = self.y\n",
    "        elif self.mesh.dim == 3:\n",
    "            z = self.z\n",
    "        layer_z = mDict[\"layer_z\"]\n",
    "        return self._atanfct(z - layer_z, self.slope)\n",
    "\n",
    "    def f0(self, mDict, z):\n",
    "        return mDict[\"func_coeffs\"][0] +  mDict[\"func_coeffs\"][1] * z\n",
    "    \n",
    "    def _transform(self, m):\n",
    "        mDict = self.mDict(m)\n",
    "        \n",
    "        if self.mesh.dim == 2:\n",
    "            z = self.y\n",
    "        elif self.mesh.dim == 3:\n",
    "            z = self.z\n",
    "        \n",
    "        return (\n",
    "            self.f0(mDict, z) * self._atanLayer(mDict) + \n",
    "            mDict[\"val_background\"] * (1 - self._atanLayer(mDict))\n",
    "        )\n",
    "    \n",
    "    def _deriv_val_background(self, mDict):\n",
    "        return np.ones_like(self.x) - self._atanLayer(mDict)\n",
    "    \n",
    "    def _deriv_layer_z(self, mDict):\n",
    "        if self.mesh.dim == 2:\n",
    "            z = self.y\n",
    "        elif self.mesh.dim == 3:\n",
    "            z = self.z\n",
    "        layer_z = mDict[\"layer_z\"]\n",
    "        return (\n",
    "            (self.f0(mDict, z) - mDict[\"val_background\"])*\n",
    "            self._atanfctDeriv(z - layer_z, self.slope)\n",
    "        )\n",
    "    \n",
    "    def _deriv_func_coeffs(self, mDict):\n",
    "        if self.mesh.dim == 2:\n",
    "            z = self.y\n",
    "        elif self.mesh.dim == 3:\n",
    "            z = self.z\n",
    "        return(\n",
    "            np.vstack([\n",
    "                self._atanLayer(mDict),\n",
    "                z * self._atanLayer(mDict)\n",
    "            ])\n",
    "        )\n",
    "\n",
    "    def deriv(self, m):\n",
    "        if self.mesh.dim == 2:\n",
    "            z = self.y\n",
    "        elif self.mesh.dim == 3:\n",
    "            z = self.z\n",
    "        mDict = self.mDict(m)\n",
    "        return sp.csr_matrix(\n",
    "            np.vstack(\n",
    "                [\n",
    "                    self._deriv_val_background(mDict),\n",
    "                    self._deriv_layer_z(mDict),\n",
    "                    self._deriv_func_coeffs(mDict)\n",
    "                ]\n",
    "            ).T\n",
    "        )\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Testing ParametricLayer(21700,4)\n",
      "==================== checkDerivative ====================\n",
      "iter    h         |ft-f0|   |ft-f0-h*J0*dx|  Order\n",
      "---------------------------------------------------------\n",
      " 0   1.00e-01    1.217e+01     1.683e-03      nan\n",
      " 1   1.00e-02    1.217e+00     1.795e-05      1.972\n",
      " 2   1.00e-03    1.217e-01     1.813e-07      1.996\n",
      " 3   1.00e-04    1.217e-02     1.814e-09      2.000\n",
      "========================= PASS! =========================\n",
      "Once upon a time, a happy little test passed.\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(-11.75, 11.75)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x144 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# test\n",
    "parametric_layer = ParametricLayer(mesh,slopeFact=1e3)\n",
    "m_background = -5\n",
    "z_layer = -0.5\n",
    "m_coeffs = np.r_[-1,0]\n",
    "m =  np.r_[m_background, z_layer, m_coeffs]\n",
    "\n",
    "parametric_layer.test(m)\n",
    "\n",
    "fig, ax = plt.subplots(1, 1, figsize=(10, 2))\n",
    "cbar = plt.colorbar(mesh.plotImage(parametric_layer * m, ax=ax)[0], ax=ax)\n",
    "cbar.set_label(r'log($\\sigma$)', rotation=270,labelpad=20,fontsize=14)\n",
    "ax.set_ylim([-5,0.5])\n",
    "ax.set_xlim([min(survey.electrode_locations.T[0]),max(survey.electrode_locations.T[0])])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Inversion"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  6.63e+05  0.00e+00  6.63e+05    1.73e+04      0              \n",
      "   1  1.00e+00  6.52e+05  2.89e-01  6.52e+05    3.70e+05      3              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 1.0528e+04 <= tolF*(1+|f0|) = 6.6257e+04\n",
      "0 : |xc-x_last| = 4.8217e+02 <= tolX*(1+|x0|) = 3.9558e+01\n",
      "0 : |proj(x-g)-x|    = 3.7045e+05 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 3.7045e+05 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  6.80e+05  2.88e-01  6.80e+05    6.92e+03      0              \n",
      "   1  1.00e+00  2.83e+05  1.03e+01  2.83e+05    1.13e+04      1              \n",
      "------------------------- STOP! -------------------------\n",
      "0 : |fc-fOld| = 3.9790e+05 <= tolF*(1+|f0|) = 6.8050e+04\n",
      "0 : |xc-x_last| = 2.4539e+02 <= tolX*(1+|x0|) = 3.2219e+01\n",
      "0 : |proj(x-g)-x|    = 1.1341e+04 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 1.1341e+04 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "1\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  2.44e+05  1.03e+01  2.44e+05    4.92e+05      0              \n",
      "   1  1.00e+00  1.02e+05  1.03e+01  1.02e+05    4.93e+04      0              \n",
      "------------------------- STOP! -------------------------\n",
      "0 : |fc-fOld| = 1.4139e+05 <= tolF*(1+|f0|) = 2.4356e+04\n",
      "1 : |xc-x_last| = 5.2274e-01 <= tolX*(1+|x0|) = 7.4180e+01\n",
      "0 : |proj(x-g)-x|    = 4.9292e+04 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 4.9292e+04 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  1.04e+05  1.03e+01  1.04e+05    1.03e+04      0              \n",
      "   1  1.00e+00  1.99e+04  2.20e+01  1.99e+04    8.76e+03      0              \n",
      "------------------------- STOP! -------------------------\n",
      "0 : |fc-fOld| = 8.4479e+04 <= tolF*(1+|f0|) = 1.0436e+04\n",
      "0 : |xc-x_last| = 8.5867e+01 <= tolX*(1+|x0|) = 5.6076e+01\n",
      "0 : |proj(x-g)-x|    = 8.7579e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 8.7579e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "2\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  2.53e+04  2.20e+01  2.54e+04    1.48e+05      0              \n",
      "   1  1.00e+00  2.00e+04  2.20e+01  2.01e+04    9.94e+03      0              \n",
      "------------------------- STOP! -------------------------\n",
      "0 : |fc-fOld| = 5.3040e+03 <= tolF*(1+|f0|) = 2.5369e+03\n",
      "1 : |xc-x_last| = 7.1996e-02 <= tolX*(1+|x0|) = 8.0309e+01\n",
      "0 : |proj(x-g)-x|    = 9.9390e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 9.9390e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  2.44e+04  2.20e+01  2.44e+04    7.94e+03      0              \n",
      "   1  1.00e+00  1.31e+04  2.57e+01  1.31e+04    3.64e+03      0              \n",
      "------------------------- STOP! -------------------------\n",
      "0 : |fc-fOld| = 1.1299e+04 <= tolF*(1+|f0|) = 2.4412e+03\n",
      "1 : |xc-x_last| = 1.7614e+01 <= tolX*(1+|x0|) = 6.3966e+01\n",
      "0 : |proj(x-g)-x|    = 3.6361e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 3.6361e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "3\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  1.23e+04  2.57e+01  1.23e+04    5.74e+04      0              \n",
      "   1  1.00e+00  1.15e+04  2.57e+01  1.15e+04    4.75e+03      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 8.3679e+02 <= tolF*(1+|f0|) = 1.2350e+03\n",
      "1 : |xc-x_last| = 2.9113e-02 <= tolX*(1+|x0|) = 8.0548e+01\n",
      "0 : |proj(x-g)-x|    = 4.7510e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 4.7510e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  1.61e+04  2.57e+01  1.61e+04    3.33e+03      0              \n",
      "   1  1.00e+00  1.09e+04  2.89e+01  1.10e+04    3.55e+03      0              \n",
      "------------------------- STOP! -------------------------\n",
      "0 : |fc-fOld| = 5.1469e+03 <= tolF*(1+|f0|) = 1.6120e+03\n",
      "1 : |xc-x_last| = 1.7194e+01 <= tolX*(1+|x0|) = 6.4266e+01\n",
      "0 : |proj(x-g)-x|    = 3.5480e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 3.5480e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "4\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  9.48e+03  2.89e+01  9.51e+03    5.13e+04      0              \n",
      "   1  1.00e+00  8.82e+03  2.89e+01  8.85e+03    4.16e+03      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 6.5730e+02 <= tolF*(1+|f0|) = 9.5079e+02\n",
      "1 : |xc-x_last| = 2.5416e-02 <= tolX*(1+|x0|) = 8.0489e+01\n",
      "0 : |proj(x-g)-x|    = 4.1615e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 4.1615e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  1.37e+04  2.89e+01  1.37e+04    3.58e+03      0              \n",
      "   1  1.00e+00  1.01e+04  3.31e+01  1.01e+04    3.29e+03      0              \n",
      "------------------------- STOP! -------------------------\n",
      "0 : |fc-fOld| = 3.6098e+03 <= tolF*(1+|f0|) = 1.3737e+03\n",
      "1 : |xc-x_last| = 1.3322e+01 <= tolX*(1+|x0|) = 6.4193e+01\n",
      "0 : |proj(x-g)-x|    = 3.2920e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 3.2920e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "5\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  8.09e+03  3.31e+01  8.12e+03    4.13e+04      0              \n",
      "   1  1.00e+00  7.70e+03  3.31e+01  7.73e+03    2.91e+04      2              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 3.9199e+02 <= tolF*(1+|f0|) = 8.1239e+02\n",
      "0 : |xc-x_last| = 2.5848e+02 <= tolX*(1+|x0|) = 8.0709e+01\n",
      "0 : |proj(x-g)-x|    = 2.9078e+04 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 2.9078e+04 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  7.81e+03  3.31e+01  7.84e+03    3.79e+03      0              \n",
      "   1  1.00e+00  5.68e+03  3.42e+01  5.71e+03    2.16e+03      0              \n",
      "------------------------- STOP! -------------------------\n",
      "0 : |fc-fOld| = 2.1296e+03 <= tolF*(1+|f0|) = 7.8405e+02\n",
      "1 : |xc-x_last| = 6.4408e+00 <= tolX*(1+|x0|) = 6.4466e+01\n",
      "0 : |proj(x-g)-x|    = 2.1633e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 2.1633e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "6\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  5.08e+03  3.42e+01  5.12e+03    1.75e+04      0              \n",
      "   1  1.00e+00  4.77e+03  3.42e+01  4.80e+03    2.16e+03      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 3.1258e+02 <= tolF*(1+|f0|) = 5.1169e+02\n",
      "0 : |xc-x_last| = 1.2208e+02 <= tolX*(1+|x0|) = 6.7909e+01\n",
      "0 : |proj(x-g)-x|    = 2.1559e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 2.1559e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  5.04e+03  3.42e+01  5.08e+03    2.09e+03      0              \n",
      "   1  1.00e+00  3.89e+03  3.76e+01  3.93e+03    1.79e+03      0              \n",
      "------------------------- STOP! -------------------------\n",
      "0 : |fc-fOld| = 1.1485e+03 <= tolF*(1+|f0|) = 5.0798e+02\n",
      "1 : |xc-x_last| = 4.8798e+00 <= tolX*(1+|x0|) = 6.3666e+01\n",
      "0 : |proj(x-g)-x|    = 1.7852e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 1.7852e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "7\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  3.70e+03  3.76e+01  3.73e+03    4.01e+03      0              \n",
      "   1  1.00e+00  3.69e+03  3.76e+01  3.72e+03    3.14e+03      2              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 9.6034e+00 <= tolF*(1+|f0|) = 3.7337e+02\n",
      "1 : |xc-x_last| = 6.1457e+01 <= tolX*(1+|x0|) = 6.4749e+01\n",
      "0 : |proj(x-g)-x|    = 3.1419e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 3.1419e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  3.71e+03  3.76e+01  3.75e+03    1.80e+03      0              \n",
      "   1  1.00e+00  2.38e+03  4.16e+01  2.42e+03    1.99e+03      0              \n",
      "------------------------- STOP! -------------------------\n",
      "0 : |fc-fOld| = 1.3267e+03 <= tolF*(1+|f0|) = 3.7477e+02\n",
      "1 : |xc-x_last| = 7.3855e+00 <= tolX*(1+|x0|) = 6.3900e+01\n",
      "0 : |proj(x-g)-x|    = 1.9855e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 1.9855e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "8\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  2.31e+03  4.16e+01  2.35e+03    4.58e+03      0              \n",
      "   1  1.00e+00  2.30e+03  4.16e+01  2.35e+03    3.71e+03      2              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 3.6106e+00 <= tolF*(1+|f0|) = 2.3509e+02\n",
      "1 : |xc-x_last| = 2.7761e+01 <= tolX*(1+|x0|) = 6.4503e+01\n",
      "0 : |proj(x-g)-x|    = 3.7141e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 3.7141e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  2.32e+03  4.16e+01  2.36e+03    1.98e+03      0              \n",
      "   1  1.00e+00  1.73e+03  4.24e+01  1.77e+03    8.90e+02      0              \n",
      "------------------------- STOP! -------------------------\n",
      "0 : |fc-fOld| = 5.8893e+02 <= tolF*(1+|f0|) = 2.3590e+02\n",
      "1 : |xc-x_last| = 2.9650e+00 <= tolX*(1+|x0|) = 6.4360e+01\n",
      "0 : |proj(x-g)-x|    = 8.8954e+02 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 8.8954e+02 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "9\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  1.72e+03  4.24e+01  1.76e+03    1.48e+03      0              \n",
      "   1  1.00e+00  1.71e+03  4.24e+01  1.75e+03    2.13e+03      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 1.1497e+01 <= tolF*(1+|f0|) = 1.7611e+02\n",
      "0 : |xc-x_last| = 1.0827e+02 <= tolX*(1+|x0|) = 6.4479e+01\n",
      "0 : |proj(x-g)-x|    = 2.1276e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 2.1276e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  2.12e+03  4.24e+01  2.17e+03    1.15e+03      0              \n",
      "   1  1.00e+00  1.66e+03  4.50e+01  1.70e+03    1.23e+03      0              \n",
      "------------------------- STOP! -------------------------\n",
      "0 : |fc-fOld| = 4.6225e+02 <= tolF*(1+|f0|) = 2.1669e+02\n",
      "1 : |xc-x_last| = 5.1303e+00 <= tolX*(1+|x0|) = 6.4376e+01\n",
      "0 : |proj(x-g)-x|    = 1.2303e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 1.2303e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "10\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  1.49e+03  4.50e+01  1.54e+03    1.09e+04      0              \n",
      "   1  1.00e+00  1.46e+03  4.50e+01  1.51e+03    1.17e+03      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 3.1877e+01 <= tolF*(1+|f0|) = 1.5382e+02\n",
      "1 : |xc-x_last| = 4.6989e+01 <= tolX*(1+|x0|) = 6.5855e+01\n",
      "0 : |proj(x-g)-x|    = 1.1702e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 1.1702e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  1.61e+03  4.50e+01  1.65e+03    1.23e+03      0              \n",
      "   1  1.00e+00  1.35e+03  4.59e+01  1.40e+03    6.12e+02      0              \n",
      "------------------------- STOP! -------------------------\n",
      "0 : |fc-fOld| = 2.5366e+02 <= tolF*(1+|f0|) = 1.6522e+02\n",
      "1 : |xc-x_last| = 2.7372e+00 <= tolX*(1+|x0|) = 6.4650e+01\n",
      "0 : |proj(x-g)-x|    = 6.1157e+02 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 6.1157e+02 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "11\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  1.28e+03  4.59e+01  1.32e+03    5.60e+03      0              \n",
      "   1  1.00e+00  1.26e+03  4.59e+01  1.31e+03    6.70e+02      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 1.6703e+01 <= tolF*(1+|f0|) = 1.3235e+02\n",
      "1 : |xc-x_last| = 6.3411e+01 <= tolX*(1+|x0|) = 6.5271e+01\n",
      "0 : |proj(x-g)-x|    = 6.6984e+02 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 6.6984e+02 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  1.26e+03  4.59e+01  1.31e+03    6.13e+02      0              \n",
      "   1  1.00e+00  1.07e+03  4.72e+01  1.12e+03    8.79e+02      0              \n",
      "------------------------- STOP! -------------------------\n",
      "0 : |fc-fOld| = 1.9060e+02 <= tolF*(1+|f0|) = 1.3112e+02\n",
      "1 : |xc-x_last| = 6.0357e+00 <= tolX*(1+|x0|) = 6.4759e+01\n",
      "0 : |proj(x-g)-x|    = 8.7928e+02 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 8.7928e+02 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "12\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  1.05e+03  4.72e+01  1.10e+03    7.40e+03      0              \n",
      "   1  1.00e+00  1.05e+03  4.72e+01  1.10e+03    6.88e+03      3              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 3.3698e+00 <= tolF*(1+|f0|) = 1.1031e+02\n",
      "1 : |xc-x_last| = 1.7844e+00 <= tolX*(1+|x0|) = 6.5194e+01\n",
      "0 : |proj(x-g)-x|    = 6.8796e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 6.8796e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  1.06e+03  4.72e+01  1.11e+03    8.72e+02      0              \n",
      "   1  1.00e+00  9.33e+02  4.76e+01  9.81e+02    4.51e+02      0              \n",
      "------------------------- STOP! -------------------------\n",
      "0 : |fc-fOld| = 1.2983e+02 <= tolF*(1+|f0|) = 1.1115e+02\n",
      "1 : |xc-x_last| = 2.2218e+00 <= tolX*(1+|x0|) = 6.5185e+01\n",
      "0 : |proj(x-g)-x|    = 4.5121e+02 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 4.5121e+02 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "13\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  9.25e+02  4.76e+01  9.72e+02    2.75e+03      0              \n",
      "   1  1.00e+00  9.12e+02  4.76e+01  9.59e+02    4.51e+02      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 1.3368e+01 <= tolF*(1+|f0|) = 9.7346e+01\n",
      "1 : |xc-x_last| = 1.0737e+00 <= tolX*(1+|x0|) = 6.5353e+01\n",
      "0 : |proj(x-g)-x|    = 4.5146e+02 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 4.5146e+02 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  9.14e+02  4.76e+01  9.62e+02    4.33e+02      0              \n",
      "   1  1.00e+00  8.02e+02  4.86e+01  8.51e+02    7.40e+02      0              \n",
      "------------------------- STOP! -------------------------\n",
      "0 : |fc-fOld| = 1.1148e+02 <= tolF*(1+|f0|) = 9.6302e+01\n",
      "1 : |xc-x_last| = 6.2709e+00 <= tolX*(1+|x0|) = 6.5258e+01\n",
      "0 : |proj(x-g)-x|    = 7.4010e+02 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 7.4010e+02 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "14\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  7.95e+02  4.86e+01  8.44e+02    1.04e+03      0              \n",
      "   1  1.00e+00  7.93e+02  4.86e+01  8.42e+02    5.29e+03      1              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 2.2810e+00 <= tolF*(1+|f0|) = 8.4499e+01\n",
      "1 : |xc-x_last| = 2.8275e+00 <= tolX*(1+|x0|) = 6.5776e+01\n",
      "0 : |proj(x-g)-x|    = 5.2947e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 5.2947e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  7.89e+02  4.86e+01  8.38e+02    7.15e+02      0              \n",
      "   1  1.00e+00  7.05e+02  4.88e+01  7.54e+02    3.29e+02      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 8.3769e+01 <= tolF*(1+|f0|) = 8.3866e+01\n",
      "1 : |xc-x_last| = 1.2136e+00 <= tolX*(1+|x0|) = 6.5733e+01\n",
      "0 : |proj(x-g)-x|    = 3.2916e+02 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 3.2916e+02 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "15\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  7.52e+02  4.88e+01  8.00e+02    7.64e+03      0              \n",
      "   1  1.00e+00  7.37e+02  4.88e+01  7.86e+02    6.63e+02      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 1.4817e+01 <= tolF*(1+|f0|) = 8.0141e+01\n",
      "1 : |xc-x_last| = 3.8915e-03 <= tolX*(1+|x0|) = 6.5732e+01\n",
      "0 : |proj(x-g)-x|    = 6.6306e+02 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 6.6306e+02 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  7.05e+02  4.88e+01  7.54e+02    3.46e+02      0              \n",
      "   1  1.00e+00  6.28e+02  4.93e+01  6.77e+02    5.22e+02      0              \n",
      "------------------------- STOP! -------------------------\n",
      "0 : |fc-fOld| = 7.6401e+01 <= tolF*(1+|f0|) = 7.5478e+01\n",
      "1 : |xc-x_last| = 4.3611e+00 <= tolX*(1+|x0|) = 6.5725e+01\n",
      "0 : |proj(x-g)-x|    = 5.2211e+02 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 5.2211e+02 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "16\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  6.66e+02  4.93e+01  7.15e+02    8.16e+03      0              \n",
      "   1  1.00e+00  6.49e+02  4.93e+01  6.98e+02    6.97e+02      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 1.6950e+01 <= tolF*(1+|f0|) = 7.1635e+01\n",
      "1 : |xc-x_last| = 4.1734e-03 <= tolX*(1+|x0|) = 6.6016e+01\n",
      "0 : |proj(x-g)-x|    = 6.9743e+02 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 6.9743e+02 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  6.27e+02  4.93e+01  6.76e+02    5.16e+02      0              \n",
      "   1  1.00e+00  5.66e+02  4.94e+01  6.15e+02    3.04e+02      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 6.0807e+01 <= tolF*(1+|f0|) = 6.7721e+01\n",
      "1 : |xc-x_last| = 1.5381e+00 <= tolX*(1+|x0|) = 6.6009e+01\n",
      "0 : |proj(x-g)-x|    = 3.0442e+02 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 3.0442e+02 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "17\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  6.16e+02  4.94e+01  6.65e+02    1.81e+03      0              \n",
      "   1  1.00e+00  5.74e+02  4.94e+01  6.23e+02    2.69e+03      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 4.1906e+01 <= tolF*(1+|f0|) = 6.6632e+01\n",
      "1 : |xc-x_last| = 1.1061e+00 <= tolX*(1+|x0|) = 6.6065e+01\n",
      "0 : |proj(x-g)-x|    = 2.6863e+03 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 2.6863e+03 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  5.68e+02  4.94e+01  6.17e+02    3.25e+02      0              \n",
      "   1  1.00e+00  5.31e+02  4.98e+01  5.81e+02    3.88e+02      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 3.6799e+01 <= tolF*(1+|f0|) = 6.1837e+01\n",
      "1 : |xc-x_last| = 2.8154e+00 <= tolX*(1+|x0|) = 6.6056e+01\n",
      "0 : |proj(x-g)-x|    = 3.8762e+02 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 3.8762e+02 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "18\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  5.24e+02  4.98e+01  5.74e+02    2.33e+03      0              \n",
      "   1  1.00e+00  5.23e+02  4.98e+01  5.73e+02    3.50e+02      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 1.3959e+00 <= tolF*(1+|f0|) = 5.7525e+01\n",
      "1 : |xc-x_last| = 4.4324e-02 <= tolX*(1+|x0|) = 6.6304e+01\n",
      "0 : |proj(x-g)-x|    = 3.4953e+02 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 3.4953e+02 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  5.28e+02  4.98e+01  5.78e+02    3.92e+02      0              \n",
      "   1  1.00e+00  4.96e+02  5.00e+01  5.46e+02    2.65e+02      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 3.2298e+01 <= tolF*(1+|f0|) = 5.7882e+01\n",
      "1 : |xc-x_last| = 9.9028e-01 <= tolX*(1+|x0|) = 6.6291e+01\n",
      "0 : |proj(x-g)-x|    = 2.6486e+02 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 2.6486e+02 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "19\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  4.99e+02  5.00e+01  5.49e+02    2.24e+03      0              \n",
      "   1  1.00e+00  4.94e+02  5.00e+01  5.44e+02    9.10e+02      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 5.3220e+00 <= tolF*(1+|f0|) = 5.5044e+01\n",
      "1 : |xc-x_last| = 1.0788e+00 <= tolX*(1+|x0|) = 6.6302e+01\n",
      "0 : |proj(x-g)-x|    = 9.1001e+02 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 9.1001e+02 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n",
      "SimPEG.InvProblem will set Regularization.mref to m0.\n",
      "\n",
      "        SimPEG.InvProblem is setting bfgsH0 to the inverse of the eval2Deriv.\n",
      "        ***Done using same Solver and solverOpts as the problem***\n",
      "model has any nan: 0\n",
      "============================ Inexact Gauss Newton ============================\n",
      "  #     beta     phi_d     phi_m       f      |proj(x-g)-x|  LS    Comment   \n",
      "-----------------------------------------------------------------------------\n",
      "x0 has any nan: 0\n",
      "   0  1.00e+00  4.96e+02  5.00e+01  5.46e+02    2.71e+02      0              \n",
      "   1  1.00e+00  4.68e+02  5.03e+01  5.19e+02    3.54e+02      0              \n",
      "------------------------- STOP! -------------------------\n",
      "1 : |fc-fOld| = 2.7610e+01 <= tolF*(1+|f0|) = 5.4711e+01\n",
      "1 : |xc-x_last| = 2.6190e+00 <= tolX*(1+|x0|) = 6.6287e+01\n",
      "0 : |proj(x-g)-x|    = 3.5404e+02 <= tolG          = 1.0000e-01\n",
      "0 : |proj(x-g)-x|    = 3.5404e+02 <= 1e3*eps       = 1.0000e-02\n",
      "1 : maxIter   =       1    <= iter          =      1\n",
      "------------------------- DONE! -------------------------\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean percent data error =  1.1345126625157504\n",
      "\r",
      " Time: 81 mins\n"
     ]
    }
   ],
   "source": [
    "\n",
    "np.random.seed(0)\n",
    "\n",
    "rho_min = min(data.dobs/IO.G)\n",
    "rho_mean = mean(data.dobs/IO.G)\n",
    "rho_max = max(data.dobs/IO.G)\n",
    "\n",
    "m_i_hybrid = []\n",
    "m_i_parametric = []\n",
    "m_i_smooth = []\n",
    "phid = []\n",
    "\n",
    "\n",
    "max_iter = 20\n",
    "\n",
    "start_time = time.time()\n",
    "\n",
    "for i in range(0,max_iter):\n",
    "    \n",
    "    print(i)\n",
    "    ### do parametric step\n",
    "    \n",
    "    # mapping\n",
    "    \n",
    "    # parametric component\n",
    "    parametric_layer = ParametricLayer(mesh,slopeFact=1e3)\n",
    "    add_z_layer = maps.InjectActiveCells(mesh, np.r_[True, False, True, False], [z_layer,0], nC=4)\n",
    "    exp_map_parametric = maps.ExpMap(mesh)\n",
    "    parametric_map = (exp_map_parametric * parametric_layer * add_z_layer)\n",
    "\n",
    "    # smooth component\n",
    "    active_inds = mesh.gridCC[:, -1] < z_layer\n",
    "    n_active = active_inds.sum()\n",
    "    active_cells = maps.InjectActiveCells(mesh, active_inds , 0)\n",
    "    exp_map_smooth = maps.ExpMap(nP=int(n_active))\n",
    "    smooth_map = active_cells * exp_map_smooth \n",
    "\n",
    "    # combine parameteric + smooth\n",
    "    wires = maps.Wires((\"parametric\", parametric_map.nP), (\"smooth\", int(n_active)))\n",
    "    sum_map = maps.SumMap([parametric_map * wires.parametric, smooth_map * wires.smooth])\n",
    "\n",
    "    # simulation \n",
    "    sim = DC.Simulation2DCellCentered(\n",
    "        mesh, survey=survey, sigmaMap=sum_map, solver=Solver, bc_type=\"Neumann\")\n",
    "    \n",
    "    # reference model\n",
    "    if i==0:\n",
    "        m0_parametric = np.r_[np.log(1e-4), np.log(1./rho_min)]\n",
    "        m0_smooth = np.ones(n_active)*np.log(1./rho_mean)\n",
    "    else:\n",
    "        m0_parametric = np.r_[mopt_p[0:2]]\n",
    "        m0_smooth = mopt_s \n",
    "        \n",
    "    m0 = np.hstack([m0_parametric, m0_smooth])\n",
    "    \n",
    "    # data misfit\n",
    "    dmisfit = data_misfit.L2DataMisfit(simulation=sim, data=data)\n",
    "    \n",
    "    # regularization mapping \n",
    "    regmap = wires.smooth\n",
    "    \n",
    "    # inversion components\n",
    "    reg = regularization.Tikhonov(mesh, mapping=regmap, indActive=active_inds)\n",
    "    opt = optimization.InexactGaussNewton(maxIter=1)\n",
    "    invProb = inverse_problem.BaseInvProblem(dmisfit, reg, opt) \n",
    "    invProb.beta = 1.\n",
    "    target = directives.TargetMisfit()\n",
    "    inv = inversion.BaseInversion(invProb, directiveList=[target])\n",
    "\n",
    "    # run\n",
    "    mopt_p = inv.run(m0) \n",
    "    m_i_parametric.append(mopt_p)\n",
    "    m_i_hybrid.append(sum_map*mopt_p)\n",
    "    phid.append(invProb.phi_d)\n",
    "\n",
    "\n",
    "\n",
    "    ### do smooth step\n",
    "    \n",
    "    # mapping\n",
    "    active_inds = mesh.gridCC[:, -1] < z_layer\n",
    "    n_active = active_inds.sum()\n",
    "    active_cells = maps.InjectActiveCells(mesh, active_inds , exp(mopt_p[1]))\n",
    "    exp_map_smooth = maps.ExpMap(nP=int(n_active))\n",
    "    smooth_map = active_cells * exp_map_smooth \n",
    "\n",
    "    # simulation \n",
    "    sim_smooth = DC.Simulation2DCellCentered(\n",
    "        mesh, survey=survey, sigmaMap=smooth_map, solver=Solver, bc_type=\"Neumann\"\n",
    "    )\n",
    "\n",
    "    # reference model \n",
    "    m0_smooth = (mopt_p[2:])\n",
    "    \n",
    "    # data misfit\n",
    "    dmisfit = data_misfit.L2DataMisfit(simulation=sim_smooth, data=data)\n",
    "\n",
    "    # inversion components\n",
    "    regmap = maps.IdentityMap(nP=active_cells.nP)\n",
    "    reg = regularization.Tikhonov(mesh,indActive=active_inds, mapping=regmap)\n",
    "    opt = optimization.InexactGaussNewton(maxIter=1)\n",
    "    invProb = inverse_problem.BaseInvProblem(dmisfit, reg, opt) \n",
    "    invProb.beta = 1.\n",
    "    target = directives.TargetMisfit()\n",
    "    inv = inversion.BaseInversion(invProb, directiveList=[target])\n",
    "\n",
    "    # run\n",
    "    mopt_s = inv.run(m0_smooth) \n",
    "    m_i_smooth.append(mopt_s)\n",
    "    m_i_hybrid.append(smooth_map * mopt_s)\n",
    "    phid.append(invProb.phi_d)\n",
    "    \n",
    "mopt_hybrid = smooth_map * mopt_s\n",
    "dpred = sim_smooth.dpred(mopt_s)\n",
    "mpe = (mean(abs(data.dobs-dpred)/abs(data.dobs))*100)\n",
    "\n",
    "print('Mean percent data error = ',mpe)\n",
    "print ('\\r Time: {:.0f} mins'.format((time.time() - start_time)/60.))\n",
    "  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plot results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'X (m)')"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x648 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fs = 16\n",
    "fig, ax = plt.subplots(3,1, figsize=[13,9])\n",
    "\n",
    "dat = mesh.plotImage(log(sig),ax=ax[0],clim=[min(log(sig)),max(log(sig))])\n",
    "ax[0].set_title(r'm$_{true}$')\n",
    "\n",
    "dat = mesh.plotImage(log(mapping*mopt_smooth),ax=ax[1],clim=[min(log(sig)),max(log(sig))])\n",
    "ax[1].set_title(r'm$_{est}$, standard inversion')\n",
    "\n",
    "dat = mesh.plotImage(log(mopt_hybrid),ax=ax[2],clim=[min(log(sig)),max(log(sig))])\n",
    "ax[2].set_title(r'm$_{est}$\\, hybrid inversion')\n",
    "\n",
    "for a in ax:\n",
    "    cbar = colorbar(dat[0],ax=a)\n",
    "    cbar.set_label(r'log($\\sigma$)', rotation=270,labelpad=20,fontsize=fs)\n",
    "    a.set_aspect(1)\n",
    "    a.set_ylim([-5,0.5])\n",
    "    a.set_xlim([min(survey.electrode_locations.T[0]),max(survey.electrode_locations.T[0])])\n",
    "    a.plot(survey.electrode_locations.T[0],survey.electrode_locations.T[1],'wo',markersize=4)\n",
    "    a.set_xlabel('')\n",
    "    a.set_ylabel('Z (m)',fontsize=fs)   \n",
    "    a.axvline(x=0,color='w',linestyle='dashed',linewidth=2)\n",
    "    a.axvline(x=7.5,color='w',linestyle='dashed',linewidth=2)\n",
    "ax[2].set_xlabel('X (m)', fontsize=fs)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Compare standard and hybrid inversions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x316.8 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# convergence plots\n",
    "fig, ax = plt.subplots(1,1, figsize=(4, 4.4))\n",
    "\n",
    "iters_standard = np.loadtxt('InversionModel-'+dtn_standard+'.txt') # file created with inversion\n",
    "iters_hybrid = phid.copy()\n",
    "ax.plot(linspace(1,20,20),log10(np.hstack([iters_standard.T[2]])),'bo-',label='Standard inversion')\n",
    "ax.plot(linspace(1,20,40),log10(iters_hybrid),'ro-',label='Hybrid inversion')\n",
    "ax.set_xticks(np.arange(0, 21, step=2));\n",
    "ax.set_xlabel('Iteration')\n",
    "ax.set_ylabel(r'log$_{10}$($\\phi_d$)')\n",
    "ax.legend(loc='upper right',bbox_to_anchor=(0.99, 0.99),fontsize=11);\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 518.4x316.8 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot iterations\n",
    "\n",
    "# standard:\n",
    "\n",
    "n_iter = len(iters_standard.T[0])\n",
    "\n",
    "# load model at each iteration\n",
    "m_i = []\n",
    "for i in range(1, n_iter+1):\n",
    "    if i  <10:\n",
    "        m_i.append(np.load('00%d-InversionModel-'%i+dtn_standard+'.npy'))\n",
    "    else:\n",
    "        m_i.append(np.load('0%d-InversionModel-'%i+dtn_standard+'.npy'))\n",
    "m_i = np.array(m_i)\n",
    "\n",
    "# plot\n",
    "colors = plt.cm.jet(np.linspace(0,1,n_iter))\n",
    "xsl = int(mesh.nCx/2)\n",
    "\n",
    "fig, ax = plt.subplots(1,2, figsize=[7.2,4.4])\n",
    "\n",
    "import matplotlib.gridspec as gridspec\n",
    "gs1 = gridspec.GridSpec(1,2)\n",
    "gs1.update(wspace=0.05, hspace=0.05) # set the spacing between axes. \n",
    "ax[0] = plt.subplot(gs1[0])\n",
    "ax[1] = plt.subplot(gs1[1])\n",
    "\n",
    "for i in range(0,n_iter):\n",
    "    ax[0].plot((np.reshape(log(mapping * m_i[i]),[mesh.nCx,mesh.nCy],order='F'))[xsl],\n",
    "             (mesh.vectorCCy),'-',color=colors[i],label='iter %.0f'%(i+1),linewidth=1)\n",
    "\n",
    "# hybrid:\n",
    "\n",
    "n_iter = len(m_i_hybrid)\n",
    "\n",
    "# load model at each iteration\n",
    "\n",
    "# plot\n",
    "colors = plt.cm.jet(np.linspace(0,1,n_iter))\n",
    "xsl = int(mesh.nCx/2)\n",
    "\n",
    "for i in range(0,n_iter):\n",
    "\n",
    "    ax[1].plot((np.reshape(log(m_i_hybrid[i]),[mesh.nCx,mesh.nCy],order='F'))[xsl],\n",
    "               (mesh.vectorCCy),'-',color=colors[i],label='iter %.0f'%(i+1),linewidth=1)\n",
    "\n",
    "for a in ax:\n",
    "    a.plot(np.reshape(log(sig),[mesh.nCx,mesh.nCy],order='F')[xsl],\n",
    "             (mesh.vectorCCy),'k-',label=r'm$_{true}$',linewidth=2.5)\n",
    "    a.set_ylim([-5,0])\n",
    "    a.set_xlabel(r'log($\\sigma$)')\n",
    "    if a==ax[0]:\n",
    "        a.set_ylabel(\"Z (m)\",labelpad=-1)\n",
    "    a.set_xlim([-8,0])\n",
    "    a.set_xticks(np.arange(-7, 1, step=1));\n",
    "ax[0].set_title('Standard')\n",
    "ax[1].set_title('Hybrid')\n",
    "ax[1].tick_params(axis='y',left=False)\n",
    "ax[0].legend(loc='lower left',bbox_to_anchor=(0.01, 0.01),\n",
    "             labelspacing=0.04,fontsize=8,framealpha=1)\n",
    "\n",
    "frame1 = plt.gca()\n",
    "frame1.axes.yaxis.set_ticklabels([]);\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot slices\n",
    "\n",
    "fig, ax = plt.subplots(1,2, figsize=[13,6])\n",
    "\n",
    "# at x=0m\n",
    "xsl = int(mesh.nCx/2)\n",
    "ax[0].plot(np.reshape(log(sig),[mesh.nCx,mesh.nCy],order='F')[xsl],\n",
    "           (mesh.vectorCCy),'k-',label=r'm$_{true}$',linewidth=2.5)\n",
    "\n",
    "ax[0].plot(np.reshape(log(exp(mopt_smooth)),[mesh.nCx,mesh.nCy],order='F')[xsl],\n",
    "           (mesh.vectorCCy),'b',label=r'm$_{est}$, standard inversion',linewidth=1.5)\n",
    "\n",
    "ax[0].plot(np.reshape(log(mopt_hybrid),[mesh.nCx,mesh.nCy],order='F')[xsl],\n",
    "           (mesh.vectorCCy),'r-',label=r'm$_{est}$, hybrid inversion',linewidth=1.5)\n",
    "\n",
    "ax[0].set_title('Vertical slice through X=0 m',y=1.02);    \n",
    "\n",
    "# at x=7.5m \n",
    "xsl = int(mesh.nCx/2)+60\n",
    "ax[1].plot(np.reshape(log(1./rho),[mesh.nCx,mesh.nCy],order='F')[xsl],\n",
    "           (mesh.vectorCCy),'k-',label=r'm$_{true}$',linewidth=2.5)\n",
    "\n",
    "ax[1].plot(np.reshape(log(exp(mopt_smooth)),[mesh.nCx,mesh.nCy],order='F')[xsl],\n",
    "           (mesh.vectorCCy),'b',label=r'm$_{est}$, standard inversion',linewidth=1.5)\n",
    "\n",
    "ax[1].plot(np.reshape(log(mopt_hybrid),[mesh.nCx,mesh.nCy],order='F')[xsl],\n",
    "           (mesh.vectorCCy),'r-',label=r'm$_{est}$, hybrid inversion',linewidth=1.5)\n",
    "\n",
    "ax[1].set_title('Vertical slice through X=75 m',y=1.02);\n",
    "\n",
    "for a in ax:\n",
    "    a.set_ylim([-5,0])\n",
    "    a.set_xlim([-7,0])\n",
    "    a.set_xlabel(r'log($\\sigma$)',fontsize=fs)\n",
    "ax[0].set_ylabel('Z (m)',fontsize=fs)\n",
    "plt.legend();\n"
   ]
  }
 ],
 "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": "3.7.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}