py/PythonCrashCourse_Numpy.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Numpy Package"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "#import time"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Document and Auto-completion in Jupyter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[888, 888, 888],\n",
       "       [888, 888, 888],\n",
       "       [888, 888, 888]])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.full((3,3), 888)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# np.[TAB]\n",
    "np.abs()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<function RandomState.normal>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Help document\n",
    "np.random.normal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# [Sh-Tab] to show function arguments\n",
    "m = np.random.normal(loc=0, scale=2, size=(3,3,4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.62614627, 0.6367582 , 0.27929782],\n",
       "       [0.4136021 , 0.31453627, 0.37130676],\n",
       "       [0.80182135, 0.22729903, 0.72353916],\n",
       "       ...,\n",
       "       [0.8484895 , 0.74656739, 0.56561472],\n",
       "       [0.93868938, 0.76417995, 0.49233368],\n",
       "       [0.59049809, 0.83239895, 0.99649781]])"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.random.uniform(size=(1000000,3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[ 2.96024931, -0.94150061,  0.97075215, -0.0661142 ],\n",
       "        [-0.18874499, -0.92460418, -0.76195149,  1.86989038],\n",
       "        [ 0.78978536,  1.11362441, -0.83243849, -2.18049592]],\n",
       "\n",
       "       [[-0.56223205, -1.50110584,  3.53977997,  1.56313814],\n",
       "        [-0.29271921, -2.60834539, -1.0455198 , -3.94372458],\n",
       "        [ 1.22510353, -1.15656434, -0.79876942, -0.52400139]],\n",
       "\n",
       "       [[-0.56395878, -1.54818624,  0.20061252,  1.03600575],\n",
       "        [ 3.33335696, -1.50825426,  2.75188253, -0.27286031],\n",
       "        [ 1.8149083 ,  1.95658111,  2.81549151, -1.42183983]]])"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "m"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3, (3, 3, 4))"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "m.ndim, m.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Vector & Matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0, 1, 2, 3, 4, 5, 6, 0, 1, 2, 3, 4, 5, 6, 0, 1, 2, 3, 4, 5, 6]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[0, 1, 2, 3, 4, 5, 6] * 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 1, 2, 3, 4, 5, 6])"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t = np.array([0, 1, 2, 3, 4, 5, 6])\n",
    "t"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0,  2,  4,  6,  8, 10, 12])"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t * 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 1 2 3 4 5 6]\n",
      "1\n",
      "(7,)\n",
      "0 1 6\n",
      "[2 3 4] [3 4 5]\n",
      "[0 1] [3 4 5 6]\n"
     ]
    }
   ],
   "source": [
    "print(t)\n",
    "print(t.ndim) # rank/dimension\n",
    "print(t.shape) # shape\n",
    "print(t[0], t[1], t[-1])\n",
    "print(t[2:5], t[3:-1])\n",
    "print(t[:2], t[3:])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Matrix multiplication"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[2 5 2] \n",
      " [[ 1.   0.7 -0.2]\n",
      " [ 0.2  1.   0.5]\n",
      " [ 0.2 -0.5  1. ]]\n"
     ]
    }
   ],
   "source": [
    "vec_x = np.array([2, 5, 2])\n",
    "mat_A = np.array([[1, 0.7, -0.2], [0.2, 1, 0.5], [0.2, -0.5, 1]])\n",
    "print(vec_x, '\\n', mat_A)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 4 25  4] \n",
      "\n",
      "[[1.   0.49 0.04]\n",
      " [0.04 1.   0.25]\n",
      " [0.04 0.25 1.  ]] \n",
      "\n",
      "[[ 2.   3.5 -0.4]\n",
      " [ 0.4  5.   1. ]\n",
      " [ 0.4 -2.5  2. ]] \n",
      "\n",
      "[[ 2.   3.5 -0.4]\n",
      " [ 0.4  5.   1. ]\n",
      " [ 0.4 -2.5  2. ]] \n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Vector * Vector (Matrix * Matrix): Component-wise\n",
    "print(vec_x * vec_x, '\\n')\n",
    "print(mat_A * mat_A, '\\n')\n",
    "\n",
    "# Matrix x Vector ??: component-wise in each vector\n",
    "print( mat_A * vec_x, '\\n')\n",
    "print( vec_x * mat_A, '\\n') # Same"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11])"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.arange(12)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0,  1,  2,  3],\n",
       "       [ 4,  5,  6,  7],\n",
       "       [ 8,  9, 10, 11]])"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.arange(12).reshape((3,4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0  1  2  3]\n",
      " [ 4  5  6  7]\n",
      " [ 8  9 10 11]] [0 1 2 3]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[ 0,  1,  4,  9],\n",
       "       [ 0,  5, 12, 21],\n",
       "       [ 0,  9, 20, 33]])"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mat_B = np.arange(12).reshape((3,4))\n",
    "vec_y = np.arange(4)\n",
    "print(mat_B, vec_y)\n",
    "mat_B * vec_y "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.arange(12)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "42\n",
      "42\n",
      "42\n"
     ]
    }
   ],
   "source": [
    "# Dot(inner) product\n",
    "vec_y = vec_x + 1\n",
    "print( np.dot(vec_x, vec_y) )\n",
    "print( vec_x.dot(vec_y) ) # same\n",
    "print( vec_x@vec_y )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 1.1   1.5  -0.05]\n",
      " [ 0.5   0.89  0.96]\n",
      " [ 0.3  -0.86  0.71]] \n",
      "\n",
      "[ 5.1  6.4 -0.1] \n",
      "\n",
      "[3.4 5.4 4.1]\n"
     ]
    }
   ],
   "source": [
    "# Matrix multiplication ?\n",
    "print(mat_A @ mat_A, '\\n')\n",
    "# print(np.matmul(mat_A, mat_A), '\\n')  # @ is exactly same as np.matmul\n",
    "\n",
    "print(mat_A @ vec_x, '\\n')  # vec_x is treated as a column vector\n",
    "\n",
    "print(vec_x @ mat_A) # vec_x is treated as a row vector"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([2, 5, 2])"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vec_x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[2],\n",
       "       [5],\n",
       "       [2]])"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#vec_x.reshape((3,1))\n",
    "vec_x[:,None]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[[2]\n",
      "  [5]\n",
      "  [2]]]\n",
      "[[2]\n",
      " [5]\n",
      " [2]]\n",
      "[[2]\n",
      " [5]\n",
      " [2]]\n"
     ]
    }
   ],
   "source": [
    "# Convert vector to a matrix with single column\n",
    "vec_x_col = vec_x[None, :, None]\n",
    "print(vec_x_col)\n",
    "print( np.reshape(vec_x,(3,1)) )\n",
    "print( vec_x.reshape((3,1)) )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0 1 2 3 4]\n"
     ]
    }
   ],
   "source": [
    "vec_x = np.arange(5)\n",
    "print(vec_x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 1, 2, 3, 4])"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vec_x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0],\n",
       "       [1],\n",
       "       [2],\n",
       "       [3],\n",
       "       [4]])"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vec_x[:,None]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0,  0,  0,  0,  0],\n",
       "       [ 0,  1,  2,  3,  4],\n",
       "       [ 0,  2,  4,  6,  8],\n",
       "       [ 0,  3,  6,  9, 12],\n",
       "       [ 0,  4,  8, 12, 16]])"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "MAT = vec_x * vec_x[:,None]\n",
    "MAT"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0],\n",
       "       [1],\n",
       "       [2],\n",
       "       [3],\n",
       "       [4]])"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "MAT[1,:,None]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 1.   0.7 -0.2]\n",
      " [ 0.2  1.   0.5]\n",
      " [ 0.2 -0.5  1. ]] [2 5 2]\n"
     ]
    }
   ],
   "source": [
    "print(mat_A, vec_x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 2. ,  1.4, -0.4],\n",
       "       [ 1. ,  5. ,  2.5],\n",
       "       [ 0.4, -1. ,  2. ]])"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mat_A * vec_x[:, None]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## N-d Array Initialization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1. 1. 1. 1. 1. 1. 1. 1. 1. 1.] [[0. 0. 0.]\n",
      " [0. 0. 0.]\n",
      " [0. 0. 0.]\n",
      " [0. 0. 0.]\n",
      " [0. 0. 0.]]\n"
     ]
    }
   ],
   "source": [
    "vec_y = np.ones(10)\n",
    "vec_z = np.zeros((3,3), dtype=float)\n",
    "print(vec_y, vec_z)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 1. ,  0.7, -0.2],\n",
       "       [ 0.2,  1. ,  0.5],\n",
       "       [ 0.2, -0.5,  1. ]])"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mat_A"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1. 1. 1.]\n",
      " [1. 1. 1.]\n",
      " [1. 1. 1.]]\n"
     ]
    }
   ],
   "source": [
    "#mat_B = np.zeros((3,3))\n",
    "mat_B = np.zeros_like(mat_A)\n",
    "mat_B = np.ones_like(mat_A)\n",
    "print(mat_B)\n",
    "\n",
    "x_norm = np.random.normal(10,2,size=(10000,5))\n",
    "x_unif = np.random.uniform(size=(10000,5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.70156283, 0.56775488, 0.0666352 , 0.03694798, 0.37853412],\n",
       "       [0.1124156 , 0.06954032, 0.4774412 , 0.62776954, 0.22402903],\n",
       "       [0.5589337 , 0.22771066, 0.99585508, 0.36031552, 0.62483539],\n",
       "       ...,\n",
       "       [0.71380004, 0.52613986, 0.55704213, 0.15593052, 0.26376138],\n",
       "       [0.3610137 , 0.64398466, 0.19240332, 0.01738119, 0.94389583],\n",
       "       [0.89186146, 0.66886961, 0.83174258, 0.94356677, 0.33247091]])"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_unif"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[14.6163539 ,  6.03994971,  9.37496929, 11.46478543, 11.21584236],\n",
       "       [12.65332235,  8.6515051 , 10.37758043, 10.14274288,  6.62902868],\n",
       "       [10.28128975, 13.81702823,  8.90160183,  7.61369815, 10.69750229],\n",
       "       ...,\n",
       "       [ 8.4998295 , 11.10313026, 10.97030237, 13.16339777,  8.0392472 ],\n",
       "       [ 9.87901298, 10.13459584, 11.04148707,  8.57544703, 13.25993821],\n",
       "       [10.54603188, 10.85936352, 11.97892541,  7.59488134, 10.25469146]])"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x_norm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Operations on Matrix (N-d array)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9.993267895639137"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(x_norm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[10.00776035  9.96292112  9.99514243  9.99407847 10.00643711]\n",
      "[1.99246068 1.98708009 1.99839084 2.01379241 1.96705789]\n"
     ]
    }
   ],
   "source": [
    "# Axis 0: Column, Axis 1: Row\n",
    "print( np.mean(x_norm, axis=0) )\n",
    "print( np.std(x_norm, axis=0) )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "9.993267895639137\n",
      "[10.00776035  9.96292112  9.99514243  9.99407847 10.00643711]\n",
      "[10.54238014  9.69083589 10.26222405 ... 10.35518142 10.57809623\n",
      " 10.24677872]\n",
      "9.993267895639137\n"
     ]
    }
   ],
   "source": [
    "print( np.mean(x_norm) )\n",
    "print( np.mean(x_norm,axis=0) )\n",
    "print( np.mean(x_norm,axis=1) )\n",
    "print( np.mean(x_norm,axis=(0,1)) )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 1, -1, -1, ..., -1, -1,  1],\n",
       "       [ 1,  1, -1, ...,  1,  1, -1],\n",
       "       [ 1, -1,  1, ..., -1,  1, -1],\n",
       "       ...,\n",
       "       [-1,  1,  1, ..., -1, -1, -1],\n",
       "       [-1,  1, -1, ...,  1, -1,  1],\n",
       "       [-1,  1, -1, ...,  1, -1, -1]])"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Random Walk Simulation\n",
    "x_rw = np.random.binomial(1,0.5,size=(1000,10))*2 - 1\n",
    "x_rw"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[  1  -1  -1 ...  -1  -1   1]\n",
      " [  2   0  -2 ...   0   0   0]\n",
      " [  3  -1  -1 ...  -1   1  -1]\n",
      " ...\n",
      " [ 66   0 -12 ...   8  10  42]\n",
      " [ 65   1 -13 ...   9   9  43]\n",
      " [ 64   2 -14 ...  10   8  42]]\n"
     ]
    }
   ],
   "source": [
    "x_cum = np.cumsum(x_rw, axis=0)\n",
    "print(x_cum)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x2098a3499d0>,\n",
       " <matplotlib.lines.Line2D at 0x2098a349a00>,\n",
       " <matplotlib.lines.Line2D at 0x2098a349a30>,\n",
       " <matplotlib.lines.Line2D at 0x2098a349b80>,\n",
       " <matplotlib.lines.Line2D at 0x2098a349ca0>,\n",
       " <matplotlib.lines.Line2D at 0x2098a349dc0>,\n",
       " <matplotlib.lines.Line2D at 0x2098a349ee0>,\n",
       " <matplotlib.lines.Line2D at 0x2098a35d040>,\n",
       " <matplotlib.lines.Line2D at 0x2098a35d160>,\n",
       " <matplotlib.lines.Line2D at 0x2098a35d280>]"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "plt.plot(x_cum)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "mat_B = np.arange(12).reshape((3,4))\n",
    "vec_x = np.arange(4)\n",
    "vec_y = np.arange(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([0, 1, 2, 3]),\n",
       " array([0, 1, 2]),\n",
       " array([[ 0,  1,  2,  3],\n",
       "        [ 4,  5,  6,  7],\n",
       "        [ 8,  9, 10, 11]]))"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vec_x, vec_y, mat_B"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0  1  2  3]\n",
      " [ 4  5  6  7]\n",
      " [ 8  9 10 11]\n",
      " [ 0  1  2  3]\n",
      " [ 4  5  6  7]\n",
      " [ 8  9 10 11]]\n",
      "[[ 0  1  2  3]\n",
      " [ 4  5  6  7]\n",
      " [ 8  9 10 11]\n",
      " [ 0  1  2  3]]\n"
     ]
    }
   ],
   "source": [
    "print(np.vstack([mat_B, mat_B]))\n",
    "print(np.vstack([mat_B, vec_x]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0  1  2  3  0]\n",
      " [ 4  5  6  7  1]\n",
      " [ 8  9 10 11  2]]\n"
     ]
    }
   ],
   "source": [
    "#print(np.hstack([mat_B, mat_B]))\n",
    "#print(np.hstack([mat_B, vec_y]))\n",
    "print(np.hstack([mat_B, vec_y[:,None]]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# Vectorization is fast\n",
    "* Using `for` or `while` loop is very slow\n",
    "* Always find proper vectorization operation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 2.99745439, -0.13782727,  0.87820482,  0.80318842, -1.02527728],\n",
       "       [-0.19455223,  1.16750393,  0.76165588,  0.81497754, -0.27265223],\n",
       "       [ 0.16339742, -0.17187894, -0.29636528, -1.39157575,  0.08720416],\n",
       "       ...,\n",
       "       [ 1.41298468, -0.7536215 ,  2.5007458 ,  0.34204669, -0.70076537],\n",
       "       [ 0.2486229 , -0.60635544,  2.20467683,  1.45205303,  0.08571512],\n",
       "       [ 0.46537275, -0.06263523, -1.59448421,  0.88472256,  0.00873597]])"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# We're going to simulate the mean of the maximum of 5 normal distribution.\n",
    "nn = 100000\n",
    "x_norm = np.random.normal(size=(nn,5))\n",
    "x_norm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Using `for` loop"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mean = 1.1592777341201512\n",
      "Wall time: 93.7 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "x_max = np.ones(x_norm.shape[0])\n",
    "for k in range(nn):\n",
    "    x_max[k] = max(x_norm[k,:])\n",
    "val = x_max.mean()\n",
    "print(f'mean = {val}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Vectorized Version"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mean = 1.1592777341201512\n",
      "Wall time: 5.32 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "val = np.amax(x_norm, axis=1).mean()\n",
    "print(f'mean = {val}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Array view and Copy\n",
    "* Array view shares the same memory space\n",
    "* Any change in the view will make the same change in the original array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0  1  2  3]\n",
      " [ 4  5  6  7]\n",
      " [ 8  9 10 11]] [4 5 6 7]\n"
     ]
    }
   ],
   "source": [
    "mat_B = np.arange(12).reshape((3,4))\n",
    "#print(mat_B)\n",
    "vec_x = mat_B[1,]\n",
    "#vec_x = mat_B[1,].copy()\n",
    "print(mat_B, vec_x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [],
   "source": [
    "vec_x[2] = -99"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0  1  2  3]\n",
      " [ 4  5  6  7]\n",
      " [ 8  9 10 11]] [  4   5 -99   7]\n"
     ]
    }
   ],
   "source": [
    "print(mat_B, vec_x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[  0   1   2   3]\n",
      " [  4 -99   6   7]\n",
      " [  8   9  10  11]] \n",
      "\n",
      " [[ 0  1  2  3]\n",
      " [ 4  5  6  7]\n",
      " [ 8  9 10 11]]\n"
     ]
    }
   ],
   "source": [
    "# mat_C is not a view\n",
    "mat_B = np.arange(12).reshape((3,4))\n",
    "mat_C = mat_B.copy()\n",
    "mat_B[1,1] = -99\n",
    "print(mat_B, \"\\n\\n\", mat_C)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.uniform"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}