sec1_4_computation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "x = \n",
      "[1.76405235 0.40015721 0.97873798]\n",
      "y = \n",
      "[ 2.2408932   1.86755799 -0.97727788]\n",
      "\n",
      " Inner product of vectors: \n",
      "3.7438707148552592\n",
      "3.7438707148552592\n",
      "3.7438707148552592\n"
     ]
    }
   ],
   "source": [
    "# 向量内积、外积与克罗内克积\n",
    "\n",
    "np.random.seed(0)\n",
    "dim = 3\n",
    "x = np.random.randn(dim, )\n",
    "y = np.random.randn(dim, )\n",
    "print('x = ')\n",
    "print(x)\n",
    "print('y = ')\n",
    "print(y)\n",
    "\n",
    "print('\\n Inner product of vectors: ')\n",
    "print(np.inner(x, y))\n",
    "print(np.dot(x, y))\n",
    "print(x.dot(y))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "x = \n",
      "[1.76405235 0.40015721 0.97873798]\n",
      "y = \n",
      "[ 2.2408932   1.86755799 -0.97727788]\n",
      "\n",
      " Outer product of vectors:\n",
      "[[ 3.95305291  3.29447005 -1.72396934]\n",
      " [ 0.89670957  0.74731679 -0.39106479]\n",
      " [ 2.19324729  1.82784994 -0.95649898]]\n",
      "\n",
      " Kronecker  product of vectors:\n",
      "[ 3.95305291  3.29447005 -1.72396934  0.89670957  0.74731679 -0.39106479\n",
      "  2.19324729  1.82784994 -0.95649898]\n",
      "[[ 3.95305291  3.29447005 -1.72396934]\n",
      " [ 0.89670957  0.74731679 -0.39106479]\n",
      " [ 2.19324729  1.82784994 -0.95649898]]\n"
     ]
    }
   ],
   "source": [
    "# 向量的外积与克罗内克积\n",
    "\n",
    "print('x = ')\n",
    "print(x)\n",
    "print('y = ')\n",
    "print(y)\n",
    "\n",
    "print('\\n Outer product of vectors:')\n",
    "u = np.outer(x, y)\n",
    "print(u)\n",
    "\n",
    "print('\\n Kronecker  product of vectors:')\n",
    "v = np.kron(x, y)\n",
    "print(v)\n",
    "print(v.reshape(3, 3))\n",
    "\n",
    "# -----------------------------------------\n",
    "# 练习：如何操作可从u可获得v？"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "m = \n",
      "[[ 0.95008842 -0.15135721 -0.10321885]\n",
      " [ 0.4105985   0.14404357  1.45427351]\n",
      " [ 0.76103773  0.12167502  0.44386323]]\n",
      "x = \n",
      "[1.76405235 0.40015721 0.97873798]\n",
      "\n",
      " m matrix-multiply x = \n",
      "[1.51441481 2.20531004 1.82562532]\n",
      "\n",
      " m matrix-multiply m = \n",
      "[[ 0.76196752 -0.17816392 -0.36399687]\n",
      " [ 1.55600596  0.13555026  0.81259578]\n",
      " [ 1.11080937 -0.04365498  0.29540988]]\n",
      "\n",
      " trace of m = \n",
      "1.537995221431893\n"
     ]
    }
   ],
   "source": [
    "# 矩阵相关计算\n",
    "\n",
    "m = np.random.randn(dim, dim)\n",
    "print('m = ')\n",
    "print(m)\n",
    "print('x = ')\n",
    "print(x)\n",
    "print('\\n m matrix-multiply x = ')\n",
    "print(m.dot(x))\n",
    "\n",
    "print('\\n m matrix-multiply m = ')\n",
    "print(m.dot(m))\n",
    "\n",
    "print('\\n trace of m = ')\n",
    "print(m.trace())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The shape of C = \n",
      "(5, 2, 6, 4)\n"
     ]
    }
   ],
   "source": [
    "# 张量相关计算: tensordot\n",
    "\n",
    "A = np.random.randn(5, 3, 2)\n",
    "B = np.random.randn(3, 6, 4)\n",
    "C1 = np.tensordot(A, B, [[1], [0]])\n",
    "print('The shape of C = ')\n",
    "print(C1.shape)\n",
    "\n",
    "# -----------------------------------------\n",
    "# 练习：使用reshape，transpose和dot实现上述tensordot运算"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The shape of the resulting tensor = \n",
      "(5, 6, 4)\n",
      "\n",
      " The illustration of the tensor-network contraction:\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": [
    "# 张量相关计算: einsum\n",
    "\n",
    "A = np.random.randn(5, 3, 3)\n",
    "B = np.random.randn(6, 3, 3)\n",
    "C = np.random.randn(4, 3, 3)\n",
    "T = np.einsum('iab,jbc,kca->ijk', A, B, C)\n",
    "\n",
    "print('The shape of the resulting tensor = ')\n",
    "print(T.shape)\n",
    "print('\\n The illustration of the tensor-network contraction:')\n",
    "\n",
    "import PlotFun as pf\n",
    "pf.im_read_and_show('./Imgs/contract.png')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}