MinMax4NE_Experiment.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "b677a86b-802b-49c1-a8bd-73ae655eb687",
   "metadata": {},
   "source": [
    "## Example 1: Nonconvex Minimax Problems without Constraints.\n",
    "We consider the following possibly nonconvex-nonconcave minimax problem:\n",
    "\\begin{equation}\n",
    "\\min_{u \\in \\mathbb{R}^{p_1}}\\max_{v\\in\\mathbb{R}^{p_2}}\\Big\\{ \\mathcal{L}(u,v) = f(u) + \\frac{1}{n}\\sum_{i=1}^n\\big[u^{\\top}A_iu + u^{\\top}L_iv - v^{\\top}B_iv + b_i^{\\top}u - c_i^{\\top}v\\big] - g(v) \\Big\\},\n",
    "\\end{equation}\n",
    "where $A_i \\in\\mathbb{R}^{p_1\\times p_1}$ and $B_i \\in \\mathbb{R}^{p_2\\times p_2}$ are symmetric matrices, $b_i \\in\\mathbb{R}^{p_1}$, $c_i \\in\\mathbb{R}^{p_2}$, and $L_i \\in \\mathbb{R}^{p_1\\times p_2}$ is a given matrix.\n",
    "Here, we add two terms $f(u)$ and $g(v)$ to possibly handle constraints or regularizers associated with $u$ and $v$, respectively.\n",
    "\n",
    "In this experiment, we consider the case $f = 0$ and $g=0$, without constraints. The optimality condition of the minimax problem reduces to $G(x) = 0$ as a linear equation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "32cb27db-072b-4b4c-a74f-c3190fb29a8a",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from numpy import linalg as la\n",
    "import matplotlib.pyplot as plt\n",
    "import random\n",
    "import scipy as sci\n",
    "import solvers.ne_solvers as solver\n",
    "import solvers.ame_solvers as acc_solver"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e13801fe-7f6b-48d6-8aba-bc627bec58db",
   "metadata": {},
   "source": [
    "## Linear Operator $G(x)$\n",
    "Define the perator $G(x) := [Au + Lv + b, -L^{\\top}u + Bv + c]$ for our experiments, where $A := \\frac{1}{n}\\sum_{i=1}^nA_i$, $B := \\frac{1}{n}\\sum_{i=1}^nB_i$, $L := \\frac{1}{n}\\sum_{i=1}^nL_i$, $b := \\frac{1}{n}\\sum_{i=1}^nb_i$, and $c := \\frac{1}{n}\\sum_{i=1}^nc_i$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "2730b330-8391-410d-a607-ccb231edd706",
   "metadata": {},
   "outputs": [],
   "source": [
    "def G_op_eval(data, x, mb_id, mb_size):\n",
    "    \"\"\"\n",
    "    -- Define the operator G for our methods\n",
    "    -- Args: data, x=x(k), mb_id=the indices of minibatch-size, mb_size=the size of mini-batch.\n",
    "    -- Returns: G(x).\n",
    "    \"\"\"\n",
    "    K = data.get(\"K\")\n",
    "    b = data.get(\"b\")\n",
    "    return np.mean([K[i].dot(x) + b[i] for i in mb_id], axis=0)\n",
    "\n",
    "def Gb_op_eval(data, x, mb_id, mb_size):\n",
    "    \"\"\"\n",
    "    -- Define the operator G and also output the output components.\n",
    "    \"\"\"\n",
    "    K   = data.get(\"K\")\n",
    "    b   = data.get(\"b\")\n",
    "    y_b = [K[i].dot(x) + b[i] for i in mb_id]\n",
    "    return np.mean(y_b, axis=0), y_b"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5f8c15bc-ec12-49d1-a35a-5247033128ec",
   "metadata": {},
   "source": [
    "## Generate Data for Our Experiments"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "8f66ce84-58ff-4d5c-b3e9-ba2b7dd612e0",
   "metadata": {},
   "outputs": [],
   "source": [
    "def generate_data(n, p):\n",
    "    \n",
    "    # generate a symmetric matrix.\n",
    "    p1 = int(p/2)\n",
    "    p2 = p-p1\n",
    "    K, b = [], []\n",
    "    K_sum = []\n",
    "    LB_value = 0.0\n",
    "    for i in range(n):\n",
    "        # Generate KKT matrix\n",
    "        D1  = np.maximum(np.diag(np.random.randn(p1)), LB_value)\n",
    "        #D1  = D1/np.sum(D1, axis=1, keepdims=True)\n",
    "        Q1, _ = la.qr(np.random.randn(p1, p1))\n",
    "        Ai = np.matmul(np.matmul(Q1, D1), Q1.T)\n",
    "        \n",
    "        D2  = np.maximum(np.diag(np.random.randn(p2)), LB_value)\n",
    "        #D2  = D2/np.sum(D2, axis=1, keepdims=True)\n",
    "        Q2, _ = la.qr(np.random.randn(p2, p2))\n",
    "        Bi = np.matmul(np.matmul(Q2, D2), Q2.T)\n",
    "        \n",
    "        Li = np.random.randn(p1,p2)\n",
    "        #Li = Li/np.sum(Li, axis=1, keepdims=True)\n",
    "        \n",
    "        Ki = np.block([[Ai,  Li], [-Li, Bi]])\n",
    "        ##Ki = Ki/np.sum(Ki, axis=1, keepdims=True)\n",
    "        ##Ki = Ki/la.norm(Ki, ord=2)\n",
    "    \n",
    "        # generate bi.\n",
    "        bi = np.random.randn(p)\n",
    "        #bi = bi/np.sum(bi)\n",
    "        \n",
    "        K.append(Ki)\n",
    "        b.append(bi)\n",
    "        if i==0:\n",
    "            K_sum = Ki\n",
    "        else:\n",
    "            K_sum = K_sum + Ki\n",
    "            \n",
    "    K_sum = (1/n)*K_sum\n",
    "    Lips = la.norm(K_sum, ord=2)    \n",
    "    my_data = dict({\"K\": K, \"b\": b, \"n\": n, \"p\": p, \"L\": Lips})\n",
    "    print(\"The Lipschitz constant:\", Lips)\n",
    "    \n",
    "    ## general an initial point.\n",
    "    x0 = np.random.randn(p)\n",
    "    \n",
    "    # return results.\n",
    "    return my_data, x0"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "414cc4bb-3fc3-4cc9-b290-2800be77ddfe",
   "metadata": {},
   "source": [
    "## Prepare to Run All Algorithm at once"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "0ea44591-5aea-4f34-8b2d-4544f4bebe6d",
   "metadata": {},
   "outputs": [],
   "source": [
    "def OneRun(my_data, x0):\n",
    "    ## check the problem size.\n",
    "    print(\"Problem size: n = \", my_data.get('n'), \", p = \", my_data.get('p'))\n",
    "\n",
    "    ## Extract information from the data.\n",
    "    Lips = my_data.get(\"L\")\n",
    "    n    = my_data.get(\"n\")\n",
    "    p    = my_data.get(\"p\")\n",
    "    \n",
    "    ## set default parameters\n",
    "    def_gamma        = 0.75\n",
    "    def_lr           = 0.5\n",
    "    def_max_iters    = 10000\n",
    "    def_tol          = 1e-20\n",
    "    def_n_max_epochs = 100\n",
    "    def_mb_size      = int(n**0.67)\n",
    "    my_is_term       = False  ## using termination condition or not.\n",
    "    my_verbose       = 0\n",
    "    output_all       = []\n",
    "\n",
    "    ## Optimistic gradient method\n",
    "    my_max_iters = def_n_max_epochs\n",
    "    my_lr1       = 1.0/(2.0*my_data.get(\"L\"))\n",
    "    print(\"Learning rate: \", my_lr1)\n",
    "    output1      = solver.OG(my_data, G_op_eval, x0, gamma=0.5, eta=my_lr1,\\\n",
    "                             verbose=my_verbose, n_max_iters=my_max_iters,\\\n",
    "                             tol=def_tol, is_term = my_is_term)\n",
    "    print(\"Solver status (DetOG)   :\", output1[\"message\"])\n",
    "    output_all.append(output1)\n",
    "        \n",
    "    ## Accelerated optimistic gradient method\n",
    "    my_max_iters2 = def_n_max_epochs\n",
    "    my_lr2        = 1.0/(2.0*my_data.get(\"L\"))\n",
    "    my_shift2     = 10.0\n",
    "    print(\"Learning rate: \", my_lr2)\n",
    "    output2       = acc_solver.AOG(my_data, G_op_eval, x0, beta=my_lr2, r_shift=my_shift2, \\\n",
    "                                   verbose=my_verbose, n_max_iters=my_max_iters2,\\\n",
    "                                   tol=def_tol, is_term = my_is_term)\n",
    "    print(\"Solver status (DetAcOG) :\", output2[\"message\"])\n",
    "    output_all.append(output2)\n",
    "\n",
    "    ## SAGA-Root-Finding method from Davis' paper\n",
    "    my_mb_size3 = int(0.5*n**0.67) # default\n",
    "    my_lr3      = 0.25/my_data.get(\"L\")\n",
    "    print(\"Learning rate: \", my_lr3)\n",
    "    output3     = solver.SagaRF(my_data, G_op_eval, Gb_op_eval, x0, mb_size=my_mb_size3, \\\n",
    "                                eta=my_lr3, verbose=my_verbose, n_epochs=def_n_max_epochs, \\\n",
    "                                tol=def_tol, is_term = my_is_term)\n",
    "    print(\"Solver status (SagaVrRF):\", output3[\"message\"])\n",
    "    output_all.append(output3)\n",
    "\n",
    "    ## SVRG Forward-Reflected-Backward Splitting Method\n",
    "    my_prob4    = 1.0/(n**0.33)    # default value\n",
    "    my_mb_size4 = int(0.5*n**0.67) # default value\n",
    "    # Theoretical learning rate\n",
    "    my_lr4      = 5.0 * 0.99*(1 - np.sqrt(1-my_prob4))/(2.0*my_data.get(\"L\"))\n",
    "    print(\"Learning rate: \", my_lr4)\n",
    "    output4     = solver.VrFRBS(my_data, G_op_eval, x0, mb_size=my_mb_size4, prob=my_prob4, \\\n",
    "                                eta=my_lr4, verbose=my_verbose, n_epochs=def_n_max_epochs, \\\n",
    "                                tol=def_tol, is_term = my_is_term)\n",
    "    print(\"Solver status (SvrgFRBS):\", output4[\"message\"])\n",
    "    output_all.append(output4)\n",
    "    \n",
    "    ## SVRG Extragradient method\n",
    "    my_prob5    = np.minimum(1.0/(n**0.33), 0.99) # default value\n",
    "    my_mb_size5 = int(0.5*n**0.67) # default value\n",
    "    my_alpha5   = 1 - my_prob5 \n",
    "    my_lr5      = 0.99*np.sqrt(1-my_alpha5)/my_data.get(\"L\")\n",
    "    print(\"Learning rate: \", my_lr5)\n",
    "    output5     = solver.VrEG(my_data, G_op_eval, x0, mb_size=my_mb_size5, prob=my_prob5, \\\n",
    "                              eta=my_lr5, alpha=my_alpha5, verbose=my_verbose, n_epochs=def_n_max_epochs,\\\n",
    "                              tol=def_tol, is_term = my_is_term)\n",
    "    print(\"Solver status (SvrgEG):  \", output5[\"message\"])\n",
    "    output_all.append(output5)\n",
    "    \n",
    "    ## Accelerated SVRG-Forward Reflected method\n",
    "    my_mb_size6 = int(0.5*n**0.67)  # default value\n",
    "    my_prob6    = 1.0/(n**0.33)     # default value\n",
    "    my_lr6      = 0.15 / my_data.get(\"L\") \n",
    "    my_shift6   = 20.0\n",
    "    print(\"Learning rate: \", my_lr6)\n",
    "    output6     = acc_solver.SvrgAcFR(my_data, G_op_eval, x0, mb_size=my_mb_size6, \\\n",
    "                                    beta=my_lr6, verbose=my_verbose, r_shift = my_shift6, \\\n",
    "                                    n_epochs=def_n_max_epochs, tol=def_tol, is_term = my_is_term)\n",
    "    print(\"Solver status (SvrgAcFR):\", output6[\"message\"])\n",
    "    output_all.append(output6)\n",
    "    \n",
    "    ## Accelerated SAGA-FR method.\n",
    "    my_mb_size7 = int(0.5*n**0.67) # default value\n",
    "    my_lr7      = 0.25 / my_data.get(\"L\")\n",
    "    my_shift7   = 20.0\n",
    "    print(\"Learning rate: \", my_lr7)\n",
    "    output7     = acc_solver.SagaAcFR(my_data, G_op_eval, Gb_op_eval, x0, mb_size=my_mb_size7, beta=my_lr7, \\\n",
    "                                  r_shift = my_shift7, verbose=my_verbose, n_epochs=def_n_max_epochs, \\\n",
    "                                  tol=def_tol, is_term = my_is_term)\n",
    "    print(\"Solver status (SagaAcFR):\", output7[\"message\"])\n",
    "    output_all.append(output7)\n",
    "\n",
    "    # return all outputs\n",
    "    return output_all\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f671cea9-1d0c-4000-94d4-42685c449792",
   "metadata": {},
   "source": [
    "## Run Experiment 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "1cc1c888-b026-43ff-859c-789724da28c8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "************************ We are running Experiment 1 ... ************************\n",
      "=======> We are running time  0  of Experiment 1\n",
      "The Lipschitz constant: 0.538001750827387\n",
      "Problem size: n =  5000 , p =  100\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.13951780760324822\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4480896380421147\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.27679039551772544\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "=======> We are running time  1  of Experiment 1\n",
      "The Lipschitz constant: 0.5363716811060406\n",
      "Problem size: n =  5000 , p =  100\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.13951780760324822\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4480896380421147\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.27679039551772544\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "=======> We are running time  2  of Experiment 1\n",
      "The Lipschitz constant: 0.5382117981094838\n",
      "Problem size: n =  5000 , p =  100\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.13951780760324822\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4480896380421147\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.27679039551772544\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "=======> We are running time  3  of Experiment 1\n",
      "The Lipschitz constant: 0.5337525004600231\n",
      "Problem size: n =  5000 , p =  100\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.13951780760324822\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4480896380421147\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.27679039551772544\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "=======> We are running time  4  of Experiment 1\n",
      "The Lipschitz constant: 0.538336712668066\n",
      "Problem size: n =  5000 , p =  100\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.13951780760324822\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4480896380421147\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.27679039551772544\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "=======> We are running time  5  of Experiment 1\n",
      "The Lipschitz constant: 0.5400798253564998\n",
      "Problem size: n =  5000 , p =  100\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.13951780760324822\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4480896380421147\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.27679039551772544\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "=======> We are running time  6  of Experiment 1\n",
      "The Lipschitz constant: 0.5375321890283996\n",
      "Problem size: n =  5000 , p =  100\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.13951780760324822\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4480896380421147\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.27679039551772544\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "=======> We are running time  7  of Experiment 1\n",
      "The Lipschitz constant: 0.5335129114308109\n",
      "Problem size: n =  5000 , p =  100\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.13951780760324822\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4480896380421147\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.27679039551772544\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "=======> We are running time  8  of Experiment 1\n",
      "The Lipschitz constant: 0.5416199731402533\n",
      "Problem size: n =  5000 , p =  100\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.13951780760324822\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4480896380421147\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.27679039551772544\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "=======> We are running time  9  of Experiment 1\n",
      "The Lipschitz constant: 0.5325185050139014\n",
      "Problem size: n =  5000 , p =  100\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.9226346517257514\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.13951780760324822\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4480896380421147\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.27679039551772544\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4613173258628757\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Finished!\n"
     ]
    }
   ],
   "source": [
    "## Experiment 1\n",
    "print(\"************************ We are running Experiment 1 ... ************************\")\n",
    "n1, p1 = 5000, 100\n",
    "n_num_of_runs = 10\n",
    "experiment1 = []\n",
    "for n_run in range(n_num_of_runs):\n",
    "    print(\"=======> We are running time \", n_run, \" of Experiment 1\")\n",
    "    my_data1, x0 = generate_data(n1, p1)\n",
    "    output1 = OneRun(my_data, x0)\n",
    "    experiment1.append(output1)\n",
    "print(\"Finished!\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0c5d1c68-f0d2-4970-b6eb-5478cb078b8f",
   "metadata": {},
   "source": [
    "## Run Experiment 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "beecba3c-7e5a-42fa-a3e5-902ebac56e8e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "************************ We are running Experiment 2 ... ************************\n",
      "=======> We are running time  0  of Experiment 2\n",
      "The Lipschitz constant: 0.5422886230369549\n",
      "Problem size: n =  10000 , p =  200\n",
      "Learning rate:  0.9220182367092125\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.9220182367092125\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.46100911835460623\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.11056233658754482\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.39939691074110184\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.2766054710127637\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.46100911835460623\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "=======> We are running time  1  of Experiment 2\n",
      "The Lipschitz constant: 0.5391415953385038\n",
      "Problem size: n =  10000 , p =  200\n",
      "Learning rate:  0.9274001566992276\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.9274001566992276\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4637000783496138\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.11120770088266661\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4017282336285316\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.27822004700976827\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4637000783496138\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "=======> We are running time  2  of Experiment 2\n",
      "The Lipschitz constant: 0.5396391560822613\n",
      "Problem size: n =  10000 , p =  200\n",
      "Learning rate:  0.9265450706541783\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.9265450706541783\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4632725353270892\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.11110516461238472\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.401357830190494\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.2779635211962535\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4632725353270892\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "=======> We are running time  3  of Experiment 2\n",
      "The Lipschitz constant: 0.5423735790467202\n",
      "Problem size: n =  10000 , p =  200\n",
      "Learning rate:  0.9218738141315874\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.9218738141315874\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4609369070657937\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.1105450183860142\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.3993343502308558\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.2765621442394762\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4609369070657937\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "=======> We are running time  4  of Experiment 2\n",
      "The Lipschitz constant: 0.5431139801957415\n",
      "Problem size: n =  10000 , p =  200\n",
      "Learning rate:  0.9206170679307445\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.9206170679307445\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.46030853396537225\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.11039431768300138\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.39878995693122454\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.2761851203792233\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.46030853396537225\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "=======> We are running time  5  of Experiment 2\n",
      "The Lipschitz constant: 0.5392925704802092\n",
      "Problem size: n =  10000 , p =  200\n",
      "Learning rate:  0.927140530704472\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.927140530704472\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.463570265352236\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.11117656824832578\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4016157696705262\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.2781421592113416\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.463570265352236\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "=======> We are running time  6  of Experiment 2\n",
      "The Lipschitz constant: 0.5398642783841461\n",
      "Problem size: n =  10000 , p =  200\n",
      "Learning rate:  0.926158703251375\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.926158703251375\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4630793516256875\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.11105883398557664\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4011904647947275\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.27784761097541244\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4630793516256875\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "=======> We are running time  7  of Experiment 2\n",
      "The Lipschitz constant: 0.5448088265950567\n",
      "Problem size: n =  10000 , p =  200\n",
      "Learning rate:  0.9177531192453274\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.9177531192453274\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4588765596226637\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.11005089187435722\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.39754936079989517\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.2753259357735982\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4588765596226637\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "=======> We are running time  8  of Experiment 2\n",
      "The Lipschitz constant: 0.5411888353931212\n",
      "Problem size: n =  10000 , p =  200\n",
      "Learning rate:  0.9238919343870028\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.9238919343870028\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4619459671935014\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.11078701803642214\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.40020855310822373\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.27716758031610084\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4619459671935014\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "=======> We are running time  9  of Experiment 2\n",
      "The Lipschitz constant: 0.5402215626910082\n",
      "Problem size: n =  10000 , p =  200\n",
      "Learning rate:  0.9255461731467135\n",
      "Solver status (DetOG)   : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.9255461731467135\n",
      "Solver status (DetAcOG) : Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.46277308657335675\n",
      "Solver status (SagaVrRF): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.11098538342147145\n",
      "Solver status (SvrgFRBS): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.4009251309631421\n",
      "Solver status (SvrgEG):   Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.27766385194401405\n",
      "Solver status (SvrgAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Learning rate:  0.46277308657335675\n",
      "Solver status (SagaAcFR): Exceed the maximum number of epochs. Increase it to run further ...\n",
      "Finished!\n"
     ]
    }
   ],
   "source": [
    "## Experiment 2\n",
    "print(\"************************ We are running Experiment 2 ... ************************\")\n",
    "n2, p2 = 10000, 200\n",
    "n_num_of_runs = 10\n",
    "experiment2 = []\n",
    "for n_run in range(n_num_of_runs):\n",
    "    print(\"=======> We are running time \", n_run, \" of Experiment 2\")\n",
    "    my_data2, x0_2 = generate_data(n2, p2)\n",
    "    output2 = OneRun(my_data2, x0_2)\n",
    "    experiment2.append(output2)\n",
    "print(\"Finished!\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0f6fc923-0e47-4543-b7e9-f84b64b09287",
   "metadata": {},
   "source": [
    "## Plot the results of Experiment 1 and Experiment 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "9d960c25-f9d2-47b3-a122-b87b9b553524",
   "metadata": {},
   "outputs": [],
   "source": [
    "## For Exmperiment 1.\n",
    "n_algs = len(experiment1[0])\n",
    "iter_all1, op_norm_all1 = [], []\n",
    "for k in range(n_algs):\n",
    "    iter_run, op_norm_run = [], []\n",
    "    for i in range(n_num_of_runs):\n",
    "        output_all1 = experiment1[i]\n",
    "        output_tmp1 = output_all1[k]\n",
    "        iter_run.append( [h.get(\"epoch\") for h in output_tmp1.get(\"epoch_hist\")] )\n",
    "        op_norm_run.append( [h.get(\"op_norm\") for h in output_tmp1.get(\"epoch_hist\")] )\n",
    "    iter_all1.append(iter_run)\n",
    "    op_norm_all1.append(op_norm_run)\n",
    "\n",
    "## Compute the mean value of the iterate.\n",
    "avg_iter1, avg_norm_un1 = [], []\n",
    "for k in range(n_algs):\n",
    "    avg_iter1.append( np.mean(iter_all1[k], axis=0) )\n",
    "    avg_norm_un1.append( np.array(op_norm_all1[k]) )\n",
    "\n",
    "## Compute the mean value of the norm.\n",
    "avg_norm_n1 = []\n",
    "for k in range(n_algs):\n",
    "    avg_norm_n1.append( np.mean(avg_norm_un1[k], axis=0)  )\n",
    "\n",
    "## Normalize\n",
    "avg_norm_f1 = []\n",
    "for k in range(n_algs):\n",
    "    tmp = avg_norm_n1[k]\n",
    "    avg_norm_f1.append( tmp / np.maximum(1, tmp[0]) )\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "fd8a065f-c89d-4047-9573-90a761623bba",
   "metadata": {},
   "outputs": [],
   "source": [
    "## For Exmperiment 2.\n",
    "n_algs = len(experiment2[0])\n",
    "iter_all2, op_norm_all2 = [], []\n",
    "for k in range(n_algs):\n",
    "    iter_run, op_norm_run = [], []\n",
    "    for i in range(n_num_of_runs):\n",
    "        output_all2 = experiment2[i]\n",
    "        output_tmp2 = output_all2[k]\n",
    "        iter_run.append( [h.get(\"epoch\") for h in output_tmp2.get(\"epoch_hist\")] )\n",
    "        op_norm_run.append( [h.get(\"op_norm\") for h in output_tmp2.get(\"epoch_hist\")] )\n",
    "    iter_all2.append(iter_run)\n",
    "    op_norm_all2.append(op_norm_run)\n",
    "\n",
    "## Compute the mean value of the iterate.\n",
    "avg_iter2, avg_norm_un2 = [], []\n",
    "for k in range(n_algs):\n",
    "    avg_iter2.append( np.mean(iter_all2[k], axis=0) )\n",
    "    avg_norm_un2.append( np.array(op_norm_all2[k]) )\n",
    "\n",
    "## Compute the mean value of the norm.\n",
    "avg_norm_n2 = []\n",
    "for k in range(n_algs):\n",
    "    avg_norm_n2.append( np.mean(avg_norm_un2[k], axis=0)  )\n",
    "\n",
    "## Normalize\n",
    "avg_norm_f2 = []\n",
    "for k in range(n_algs):\n",
    "    tmp = avg_norm_n2[k]\n",
    "    avg_norm_f2.append( tmp / np.maximum(1, tmp[0]) )\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "5da0c308-de60-4018-9452-2ab36d61ce93",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x350 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "## Now we plot the results.\n",
    "p_e1, n_e1  = p1, n1\n",
    "p_e2, n_e2  = p2, n2\n",
    "\n",
    "fig, axs = plt.subplots(1, 2, figsize=(12, 3.5))\n",
    "plt.rc('text', usetex=True)\n",
    "alg_names = ['OG', 'AOG', 'SagaRF', 'VrFRBS', 'VrEG', 'AVFR-Svrg', 'AVFR-Saga']\n",
    "l_colors   = ['C0', 'g', 'm', 'y', 'k', 'r', 'b', 'C1', 'C2']\n",
    "l_line_styles = ['--', '-', '-.', ':', '-', '--', '-.', '-', '--']\n",
    "l_markers = ['o', '*', 'd', 'v', '^', '>', '<', 's', 'o']\n",
    "\n",
    "for k in range(n_algs):\n",
    "    axs[0].semilogy(avg_iter1[k], avg_norm_f1[k], marker= l_markers[k], linewidth=2, linestyle=l_line_styles[k], color=l_colors[k], \\\n",
    "                    markersize=4, markevery=5, label=alg_names[k]);   \n",
    "axs[0].set_title(\"Experiment 1: n = \" + str(n_e1) + \" and p = \" + str(p_e1), fontsize=11)\n",
    "axs[0].legend()\n",
    "axs[0].grid(True)\n",
    "axs[0].set_xlabel('Number of epochs')\n",
    "axs[0].set_ylabel('Relative operator norm $\\Vert Gx^k \\Vert/\\Vert Gx^0\\Vert$')\n",
    "\n",
    "for k in range(n_algs):\n",
    "    axs[1].semilogy(avg_iter2[k], avg_norm_f2[k], marker= l_markers[k], linewidth=2, linestyle=l_line_styles[k], color=l_colors[k], \\\n",
    "                    markersize=4, markevery=5, label=alg_names[k]);   \n",
    "axs[1].set_title(\"Experiment 2: n = \" + str(n_e2) + \" and p = \" + str(p_e2), fontsize=11)\n",
    "axs[1].legend()\n",
    "axs[1].grid(True)\n",
    "axs[1].set_xlabel('Number of epochs')\n",
    "axs[1].set_ylabel('Relative operator norm $\\Vert Gx^k\\Vert/\\Vert Gx^0\\Vert$')\n",
    "\n",
    "plt.show()\n",
    "fig.savefig(\"figs/minimax4ne_experiment1_and_2_7algs.pdf\", format=\"pdf\")\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "eddba287-2bf5-436a-86d5-b2ad626ae7b5",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "737eff53-3376-4351-8742-b19e0607441f",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.11.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}