ForwardBackwardSparseL1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Forward-Backward Splitting\n",
    "===================================\n",
    "\n",
    "$\\newcommand{\\dotp}[2]{\\langle #1, #2 \\rangle}$\n",
    "$\\newcommand{\\enscond}[2]{\\lbrace #1, #2 \\rbrace}$\n",
    "$\\newcommand{\\pd}[2]{ \\frac{ \\partial #1}{\\partial #2} }$\n",
    "$\\newcommand{\\umin}[1]{\\underset{#1}{\\min}\\;}$\n",
    "$\\newcommand{\\umax}[1]{\\underset{#1}{\\max}\\;}$\n",
    "$\\newcommand{\\umin}[1]{\\underset{#1}{\\min}\\;}$\n",
    "$\\newcommand{\\uargmin}[1]{\\underset{#1}{\\arg\\min}\\;}$\n",
    "$\\newcommand{\\norm}[1]{\\|#1\\|}$\n",
    "$\\newcommand{\\abs}[1]{\\left|#1\\right|}$\n",
    "$\\newcommand{\\choice}[1]{ \\left\\{  \\begin{array}{l} #1 \\end{array} \\right. }$\n",
    "$\\newcommand{\\pa}[1]{\\left(#1\\right)}$\n",
    "$\\newcommand{\\diag}[1]{{diag}\\left( #1 \\right)}$\n",
    "$\\newcommand{\\qandq}{\\quad\\text{and}\\quad}$\n",
    "$\\newcommand{\\qwhereq}{\\quad\\text{where}\\quad}$\n",
    "$\\newcommand{\\qifq}{ \\quad \\text{if} \\quad }$\n",
    "$\\newcommand{\\qarrq}{ \\quad \\Longrightarrow \\quad }$\n",
    "$\\newcommand{\\ZZ}{\\mathbb{Z}}$\n",
    "$\\newcommand{\\CC}{\\mathbb{C}}$\n",
    "$\\newcommand{\\RR}{\\mathbb{R}}$\n",
    "$\\newcommand{\\EE}{\\mathbb{E}}$\n",
    "$\\newcommand{\\Zz}{\\mathcal{Z}}$\n",
    "$\\newcommand{\\Ww}{\\mathcal{W}}$\n",
    "$\\newcommand{\\Vv}{\\mathcal{V}}$\n",
    "$\\newcommand{\\Nn}{\\mathcal{N}}$\n",
    "$\\newcommand{\\NN}{\\mathcal{N}}$\n",
    "$\\newcommand{\\Hh}{\\mathcal{H}}$\n",
    "$\\newcommand{\\Bb}{\\mathcal{B}}$\n",
    "$\\newcommand{\\Ee}{\\mathcal{E}}$\n",
    "$\\newcommand{\\Cc}{\\mathcal{C}}$\n",
    "$\\newcommand{\\Gg}{\\mathcal{G}}$\n",
    "$\\newcommand{\\Ss}{\\mathcal{S}}$\n",
    "$\\newcommand{\\Pp}{\\mathcal{P}}$\n",
    "$\\newcommand{\\Ff}{\\mathcal{F}}$\n",
    "$\\newcommand{\\Xx}{\\mathcal{X}}$\n",
    "$\\newcommand{\\Mm}{\\mathcal{M}}$\n",
    "$\\newcommand{\\Ii}{\\mathcal{I}}$\n",
    "$\\newcommand{\\Dd}{\\mathcal{D}}$\n",
    "$\\newcommand{\\Ll}{\\mathcal{L}}$\n",
    "$\\newcommand{\\Tt}{\\mathcal{T}}$\n",
    "$\\newcommand{\\si}{\\sigma}$\n",
    "$\\newcommand{\\al}{\\alpha}$\n",
    "$\\newcommand{\\la}{\\lambda}$\n",
    "$\\newcommand{\\ga}{\\gamma}$\n",
    "$\\newcommand{\\Ga}{\\Gamma}$\n",
    "$\\newcommand{\\La}{\\Lambda}$\n",
    "$\\newcommand{\\si}{\\sigma}$\n",
    "$\\newcommand{\\Si}{\\Sigma}$\n",
    "$\\newcommand{\\be}{\\beta}$\n",
    "$\\newcommand{\\de}{\\delta}$\n",
    "$\\newcommand{\\De}{\\Delta}$\n",
    "$\\newcommand{\\phi}{\\varphi}$\n",
    "$\\newcommand{\\th}{\\theta}$\n",
    "$\\newcommand{\\om}{\\omega}$\n",
    "$\\newcommand{\\Om}{\\Omega}$\n"
   ]
  },
  {
   "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 __future__ import division\n",
    "%pylab inline\n",
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This numerical tour presents the Forward-Backward (FB) algorithm to\n",
    "minimize the sum of a smooth and a simple function. It shows an\n",
    "application to sparse deconvolution."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Forward-Backward Algorithm\n",
    "--------------------------\n",
    "We consider the problem of minimizing the sum of two functions\n",
    "$$ E^\\star = \\umin{x \\in \\RR^N} E(x) = f(x) + g(x). $$\n",
    "\n",
    "So, we want to find a vector $x^\\star$ solution to the problem, i.e. a minimizer of $E=f+g$.\n",
    "\n",
    "\n",
    "We assume that $f$ is a $C^1$ function with $\\beta$-Lipschitz gradient.\n",
    "\n",
    "\n",
    "\n",
    "We also assume that $g$ is \"simple\", in the sense that one can compute exactly and quickly its proximity operator, which is defined as\n",
    "$$ \\text{prox}_{\\ga g}(x) = \\uargmin{y \\in \\RR^N} \\frac{1}{2}\\norm{x-y}^2 + \\ga g(y). $$\n",
    "for any $\\ga > 0$.\n",
    "\n",
    "\n",
    "The forward-backward algorithm reads, after initializing $x^{(0)} \\in \\RR^N$,\n",
    "$$ x^{(k+1)} = \\text{prox}_{\\ga g}\\pa{ x^{(k)} - \\ga \\nabla f( x^{(k)} )  }. $$\n",
    "\n",
    "\n",
    "If $0 < \\ga < 2/\\beta$, then this scheme converges to a minimizer of\n",
    "$f+g$.\n",
    "\n",
    "Sparse Regularization of Inverse Problems\n",
    "-----------------------------------------\n",
    "We consider a linear inverse problem\n",
    "$$ y = A x^\\sharp + w  \\in \\RR^P$$\n",
    "where $x^\\sharp \\in \\RR^N$ is the (unknown) signal to recover, $w \\in\n",
    "\\RR^P$ is a noise vector, and $A \\in \\RR^{P \\times N}$ models the\n",
    "acquisition device.\n",
    "\n",
    "\n",
    "To recover an estimate of the signal $x^\\sharp$, we consider  basis\n",
    "pursuit denoising, which makes use of the $\\ell^1$ norm as sparsity\n",
    "enforcing penalty:\n",
    "$$ \\umin{x \\in \\RR^N} \\frac{1}{2} \\norm{A x-y}^2 + \\la \\norm{x}_1, $$\n",
    "where the $\\ell^1$ norm is defined as\n",
    "$$ \\norm{x}_1 = \\sum_i \\abs{x_i}. $$\n",
    "\n",
    "\n",
    "The parameter $\\la$ should be set in accordance to the noise level\n",
    "$\\norm{w}$.\n",
    "\n",
    "\n",
    "This minimization problem can be cast in the form of minimizing $f+g$\n",
    "where\n",
    "$$ f(x) = \\frac{1}{2} \\norm{Ax-y}^2\n",
    "\\qandq g(x) = \\la \\norm{x}_1. $$\n",
    "\n",
    "\n",
    "$f$ is smooth; we have\n",
    "$$ \\nabla f(x) = A^* (A x - y),  $$\n",
    "which is $\\beta$-Lipschitz continuous, with\n",
    "$$ \\beta = \\norm{ A^* A }. $$\n",
    "\n",
    "\n",
    "The $\\ell^1$-norm is \"simple\", because its proximal operator is soft\n",
    "thresholding:\n",
    "$$ \\big(\\text{prox}_{\\ga g}(x)\\big)_n = \\max\\pa{ 0, 1 - \\frac{\\la \\ga}{\\abs{x_n}} } x_n. $$\n",
    "\n",
    "Signal Deconvolution on Synthetic Sparse Data\n",
    "-----------------------------------------------------\n",
    "A simple linearized model of seismic acquisition considers a linear filtering\n",
    "operator (convolution):\n",
    "$$ A x = h \\ast x $$\n",
    "\n",
    "The filter $h$ is called the impulse response, or the poind spread function, of the acquisition process $x\\mapsto Ax$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "N = 1024"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We define the width of the filter $h$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "s = 5"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We define $h$ as the second derivative of a\n",
    "Gaussian."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "t = arange(-N/2,N/2)\n",
    "h = (1-t**2/s**2)*exp(-(t**2)/(2*s**2))\n",
    "h = h - h.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We define the operator $A$. For simplicity, here periodic boundary conditions are used, so that the convolution is efficiently implemented as a product in Fourier domain."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "h_tf = fft.fft(fft.fftshift(h))\n",
    "opA = lambda u : real(fft.ifft(fft.fft(u) * h_tf))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We display the filter $h$ and its spectrum (amplitude of its Fourier transform)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-100, 100)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAbgAAAEzCAYAAABDvRaoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nASdJGiUDTpI0Sv8fKB+DmkQ0b9MAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f018c07d650>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "figsize(7,5)\n",
    "plot(t,h)\n",
    "xlim(-100,100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f016e99c090>]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAawAAAE4CAYAAAAHP8f5AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nkqS5YGBJkuaCgSVJmgsGliRpLvx/5IZqrPd8ZfMAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f016e9eef10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot(t,fft.fftshift(abs(h_tf)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We generate a synthetic sparse signal $x^\\sharp$, with only a small number of nonzero\n",
    "coefficients."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "random.seed(80) # we set the seed of the random number generator for reproducibility purpose."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "s = round(N*.01) # number of nonzero elements of xsharp\n",
    "sel = random.permutation(N)\n",
    "sel = sel[0:s]   # indices of the nonzero elements of xsharp\n",
    "xsharp = zeros(N)\n",
    "xsharp[sel] = sign(randn(s)) * (1-0.3*rand(s))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "noiselevel = 0.2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Compute the measurements $y=A x^\\sharp + w$ where $w$ is a realization\n",
    "of white Gaussian noise."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = opA(xsharp) + noiselevel * randn(N)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f016e9a7b50>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAzYAAAFFCAYAAAA+ZngHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\neb97JyXyU2sAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f016e93acd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "figsize(14,5)\n",
    "stem(sel,xsharp[sel])\n",
    "xlim(0,N-1)\n",
    "title('signal $x^\\sharp$')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f016e658f90>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAzYAAAFECAYAAAD1OquiAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f016e951e10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "figsize(14,5)\n",
    "plot(range(N),y)\n",
    "xlim(0,N-1)\n",
    "title('signal $y$')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Deconvolution\n",
    "---------------------------\n",
    "We now implement the foward-backward algorithm to recover an estimate of the sparse signal\n",
    "\n",
    "\n",
    "We define the regularization parameter $\\la$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "Lambda = 3"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We define the proximity operator of $\\ga g$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def prox_gamma_g (x, gamma, Lambda) :\n",
    "    return x - x/maximum(abs(x)/(Lambda*gamma),1) # soft-thresholding"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We define the gradient operator of $f$. Note that $A^*=A$ because the filter $h$ is symmetric."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>\n",
    " <font color='magenta'>\n",
    "__Exercise: write the code of the function grad_f.__</font></h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "grad_f = lambda x : opA(opA(x))-opA(y)# put your code here"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We define the Lipschitz constant $\\beta$ of $\\nabla f$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "85.032184045506753"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "beta = abs(fft.fft(h)).max()**2\n",
    "beta"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We define the stepsize $\\ga$, which must be smaller than $2/\\beta$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "gamma = 1.9 / beta"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We compute the solution of $\\ell_1$ deconvolution (basis pursuit denoising).\n",
    "We keep track of the energy $E_k=f(x^{(k)})+g(x^{(k)})$."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>\n",
    " <font color='magenta'>\n",
    "__Exercise: write the code of the forward-backward iteration.__</font></h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "nbiter = 2000\n",
    "x = y\n",
    "En_array = zeros(nbiter+1)\n",
    "En_array[0] = norm(opA(x) - y)**2/2 + Lambda*norm(x, ord=1)\n",
    "for iter in range(nbiter):  # iter goes from 0 to nbiter-1\n",
    "    x = prox_gamma_g(x-gamma*grad_f(x),gamma,Lambda)\n",
    "    En_array[iter+1] = norm(opA(x) - y)**2/2 + Lambda*norm(x, ord=1)  \n",
    "x_restored = x"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We display the result."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x7f016e78d7d0>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA7EAAAG0CAYAAADkcaDDAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nxv8Px1+i7W32KaAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f016e9d02d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, (subfig1,subfig2) = subplots(1,2,figsize=(16,7)) # one figure with two horizontal subfigures\n",
    "subfig1.stem(xsharp)\n",
    "subfig2.stem(x_restored)\n",
    "subfig1.set_title('$x^\\sharp$')\n",
    "subfig2.set_title('$x_\\mathrm{restored}$')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We plot the relative error $(E_k-E^\\star)/(E_0-E^\\star)$ in log-scale with respect to $k$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f61020c5c10>]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAz4AAAE4CAYAAACaDSbsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSZLU8ww+kiRJknre/wfClW7cED54TQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f6108a05e50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot(log10((En_array[0:1800]-En_array[-1])/(En_array[0]-En_array[-1])))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Over-relaxed Forward-Backward\n",
    "-----------------------------\n",
    "It is possible to introduce a relaxation parameter $\\rho$ with $0 < \\rho < 1$. The over-relaxed foward-backward algorithm initializes $x^{(0)} \\in \\RR^N$,\n",
    "and then iterates, for $k=1,2,\\ldots$\n",
    "$$ z^{(k)} = \\text{prox}_{\\ga g}\\pa{\n",
    "              x^{(k-1)} - \\ga \\nabla f( x^{(k-1)} )  }.\n",
    "$$\n",
    "$$ x^{(k)} = z^{(k)} +\n",
    "      \\rho \\pa{ z^{(k)} - x^{(k-1)} } $$\n",
    "\n",
    "\n",
    "Let us assume $\\gamma=1/\\beta$. Convergence of the iterates $x^{(k)}$ and $z^{(k)}$ to a solution is guaranteed\n",
    "for $ 0 < \\rho < 1/2 $. The weaker property of convergence of $ E(x^{(k)}) $\n",
    "to $E^\\star$ is proved, when $ 1/2\\leq \\rho <1 $."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>\n",
    " <font color='magenta'>\n",
    "__Exercise: write the code of the over-relaxed forward-backward iteration.__</font></h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "gamma = 1/beta\n",
    "nbiter = 1700\n",
    "rho = 0.95\n",
    "x = y\n",
    "En_array_overrelaxed = zeros(nbiter+1)\n",
    "En_array_overrelaxed[0] = norm(opA(x) - y)**2/2 + Lambda*norm(x, ord=1)\n",
    "for iter in range(nbiter):  \n",
    "    z = prox_gamma_g(x-gamma*grad_f(x),gamma,Lambda)\n",
    "    x = z + rho*(z-x)\n",
    "    # put your code here\n",
    "    En_array_overrelaxed[iter+1] = norm(opA(x) - y)**2/2 + Lambda*norm(x, ord=1)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f016c1ab910>]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAz4AAAE4CAYAAACaDSbsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nkiT13kv4iZJkWBxFnAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f016e7eca90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot(log10((En_array[0:1800]-En_array[-1])/(En_array[0]-En_array[-1])),'r')\n",
    "plot(log10((En_array_overrelaxed[0:1800]-En_array[-1])/(En_array[0]-En_array[-1])))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we see, in this example, over-relaxation does not bring any speedup, because $\\gamma$ is lower than without over-relaxation. There are other setting parameters or other problems, for which over-relaxation does bring a significant speedup. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "FISTA-like Accelerated Forward-Backward Algorithm\n",
    "---------------------------\n",
    "We consider the FISTA algorithm introduced in:\n",
    "\n",
    "\n",
    "A. Beck and M. Teboulle,\n",
    "\"A Fast Iterative Shrinkage-Thresholding Algorithm for Linear Inverse Problems\",\n",
    "_SIAM Journal on Imaging Sciences_, 2009.\n",
    "\n",
    "More precisely, we consider a slightly modified version of FISTA, whose convergence is proved, see\n",
    "A. Chambolle and C. Dossal, \"On the convergence of the iterates of \"FISTA\"\", preprint, 2015."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Given an initial estimate $x^{(0)}$ of the solution and a parameter $a>2$, the algorithm sets $\\gamma=1/\\beta$, sets $z^{(0)}=x^{(0)} \\in \\RR^N$,\n",
    "and iterates, for $k=1,2,\\ldots$\n",
    "$$ x^{(k)} = \\text{prox}_{\\ga g}\\pa{\n",
    "              z^{(k-1)} - \\ga \\nabla f( z^{(k-1)} )  }.\n",
    "$$\n",
    "$$ \\alpha_k=(k-1)/(k+a) $$\n",
    "$$ z^{(k)} = x^{(k)} +\n",
    "      \\alpha_k\n",
    "  \\pa{ x^{(k)} - x^{(k-1)} } $$\n",
    "\n",
    "\n",
    "It is proved that the iterates $x^{(k)}$ converge to a solution $x^\\star$ of the problem. Moreover, \n",
    "the  optimal convergence rate for this class of problems is reached,\n",
    "namely\n",
    "$$ E_k - E^\\star = O(1/k^2), $$\n",
    "whereas the convergence rate for the normal forward-backward is only\n",
    "$O(1/k)$.\n",
    "\n",
    "Note the difference between the over-relaxed forward-backward and the accelerated forward-backward: the later is based on an inertia mechanism, of different nature than over-relaxation."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>\n",
    " <font color='magenta'>\n",
    "__Exercise: write the code of the accelerated forward-backward iteration.__</font></h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "gamma = 1/beta\n",
    "nbiter = 1700\n",
    "a = 10\n",
    "x = y\n",
    "zk=y\n",
    "xkm1=x\n",
    "En_array_fista = zeros(nbiter+1)\n",
    "En_array_fista[0] = norm(opA(x) - y)**2/2 + Lambda*norm(x, ord=1)\n",
    "for iter in range(nbiter): \n",
    "    # put your code here\n",
    "    xkm1=x\n",
    "    x = prox_gamma_g(zk-gamma*grad_f(zk),gamma,Lambda)\n",
    "    ak = (iter-1)/(iter+a)\n",
    "    zk = x + ak*(x-xkm1)\n",
    "    En_array_fista[iter+1] = norm(opA(x) - y)**2/2 + Lambda*norm(x, ord=1)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/gorthaur/lib/python2.7/site-packages/IPython/kernel/__main__.py:3: RuntimeWarning: divide by zero encountered in log10\n",
      "  app.launch_new_instance()\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f016bfd6310>]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAz4AAAE4CAYAAACaDSbsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nfx96xQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f016c043f10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot(log10((En_array[0:1800]-En_array[-1])/(En_array[0]-En_array[-1])))\n",
    "plot(log10((En_array_overrelaxed[0:1800]-En_array[-1])/(En_array[0]-En_array[-1])))\n",
    "plot(log10((En_array_fista[0:1800]-En_array[-1])/(En_array[0]-En_array[-1])))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3>\n",
    " <font color='magenta'>\n",
    "__Exercise: try different values of a. You can try a=3,10,30,50,100.__</font></h3>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can note that the accelerated forward-backward is not monotonic: the cost function E is not decreasing along with the iterations and some oscillations are present."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}