FW_NQP_2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "a8b0fa2a",
   "metadata": {},
   "outputs": [],
   "source": [
    "import cvxpy as cp\n",
    "import numpy as np\n",
    "import pdb\n",
    "from tqdm import tqdm\n",
    "\n",
    "class FW:\n",
    "    def __init__(self, constraint_var, constraints, H, A, h, u_bar):\n",
    "      \n",
    "        self.constraint_var = constraint_var\n",
    "        self.constraints = constraints\n",
    "        self.H, self.A, self.h, self.u_bar = H, A, h, u_bar\n",
    "        self.n = H.shape[1]\n",
    "        self.m = A.shape[0]\n",
    "\n",
    "    def project(self, constraints, grad):\n",
    "\n",
    "        x = self.constraint_var\n",
    "        objective = cp.Maximize(cp.scalar_product(x, grad)) \n",
    "        prob = cp.Problem(objective, constraints)\n",
    "        prob.solve() \n",
    "        return x.value\n",
    "\n",
    "\n",
    "    def stochastic_continuous_greedy_step(self, x, grad, epoch, alpha):\n",
    "       \n",
    "        s = np.zeros_like(grad)\n",
    "        \n",
    "        for i in range(s.shape[0]):\n",
    "            s[i] = self.project( self.constraints, grad[i])\n",
    "        \n",
    "        \n",
    "        return (1- alpha)*x + alpha*s \n",
    "\n",
    "    \n",
    "    def compute_value_grad(self, x, noise_scale=10):\n",
    "        noise = np.random.normal(scale=noise_scale, size=x.shape)\n",
    "        \n",
    "        value = 1/2*x @ self.H @ x.T + self.h.T @ x.T\n",
    "        gradient = self.H@x.T + self.h\n",
    "        \n",
    "        return value[0][0], gradient + noise\n",
    "    \n",
    "    \n",
    "    def stochastic_continuous_greedy(self, epoch, weight_shape):\n",
    "\n",
    "        x = np.zeros(shape=weight_shape) + 1e-5\n",
    "        values = []\n",
    "        for e in tqdm(range(epoch)):\n",
    "            alpha = (0.1 / (e + 1))\n",
    "            #print(alpha)\n",
    "            value, grad = self.compute_value_grad(x)\n",
    "            x = self.stochastic_continuous_greedy_step(x, grad, epoch, alpha)\n",
    "            values.append(value)\n",
    "\n",
    "        return values "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "443214fc",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 21.80it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.08it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.53it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.76it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.85it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.79it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.81it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.81it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.77it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.48it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.25it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.68it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.80it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.75it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.71it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.62it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.68it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.69it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.82it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.25it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:15<00:00, 19.22it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.52it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.09it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.57it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.85it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.74it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.82it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.14it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 21.94it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.15it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.13it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.09it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.66it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.14it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.09it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.27it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.11it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.15it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.26it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.48it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.80it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.50it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.70it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.71it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.75it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 21.94it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:15<00:00, 18.92it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.16it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.30it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.74it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.69it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.76it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.91it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.60it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.53it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 23.00it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:12<00:00, 23.15it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:12<00:00, 23.08it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:12<00:00, 23.40it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.65it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.83it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.78it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.75it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:12<00:00, 23.19it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:12<00:00, 23.17it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:12<00:00, 23.31it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.95it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.25it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.59it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.44it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 21.78it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.41it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.33it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.49it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.50it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 21.99it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.10it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.36it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.16it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 21.99it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.34it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.14it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 21.87it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 21.81it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.38it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 21.99it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.36it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 21.99it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.34it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 21.83it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.29it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.36it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.58it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:14<00:00, 21.42it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 21.98it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 21.86it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 21.83it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.49it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.46it/s]\n",
      "100%|█████████████████████████████████████████| 300/300 [00:13<00:00, 22.42it/s]\n"
     ]
    }
   ],
   "source": [
    "Val = []\n",
    "run = 100\n",
    "for i in range(run):\n",
    "    n = 10\n",
    "    m = 50\n",
    "    b = 10\n",
    "    u_bar = np.ones((1,n))\n",
    "    H = np.random.uniform(-100, 0, (n, n))\n",
    "    A = np.random.uniform(0, 1, (m, n))\n",
    "    h = -1 * H.T @ u_bar.T\n",
    "    train_iter = 300\n",
    "    x = cp.Variable(shape=(1,n))\n",
    "    constraints = [0 <= x, x <= u_bar, A @ x.T <= b]\n",
    "    fw = FW(x, constraints, H, A, h, u_bar)\n",
    "    values = fw.stochastic_continuous_greedy(train_iter, (1,n))\n",
    "    Val.append(values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "d9b538b8",
   "metadata": {},
   "outputs": [
    {
     "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": [
    "import matplotlib.pyplot as plt\n",
    "plt.figure()\n",
    "#plt.axis([0, 50, -2500, 2500])\n",
    "plt.plot(values)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "2951db7f",
   "metadata": {},
   "outputs": [],
   "source": [
    "data = np.array(Val)\n",
    "average = np.average(data, axis=0)\n",
    "minimum = np.min(np.array(Val), axis=0) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "2586b2dd",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "plt.figure(figsize=(10,8))\n",
    "plt.plot(minimum, color='orange', label='minimum')\n",
    "plt.plot(average, color='blue', label= 'average')\n",
    "plt.legend(loc=\"upper left\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "37ce2d0d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "P-correlation 0.9999987990147075\n"
     ]
    }
   ],
   "source": [
    "from scipy import stats\n",
    "r, f = stats.pearsonr(average, minimum)\n",
    "print(f\"P-correlation {r}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "86639abd",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1f3d9594",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3ff26a8f",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}