4_matrix_math/4_matrix_math.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3 style='color:blue'>Calculate profit/loss from revenue and expenses</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "revenue = np.array([[180,200,220],[24,36,40],[12,18,20]])\n",
    "expenses = np.array([[80,90,100],[10,16,20],[8,10,10]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[100, 110, 120],\n",
       "       [ 14,  20,  20],\n",
       "       [  4,   8,  10]])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "profit = revenue - expenses\n",
    "profit"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3 style='color:blue'>Calculate total sales from units and price per unit using matrix multiplication</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "price_per_unit = np.array([1000,400,1200])\n",
    "units = np.array([[30,40,50],[5,10,15],[2,5,7]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[30000, 16000, 60000],\n",
       "       [ 5000,  4000, 18000],\n",
       "       [ 2000,  2000,  8400]])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "price_per_unit*units"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In above case numpy is using broadcasting so it expands price_per_unit array from 1 row, 3 columns to 3 row and 3 columns. Correct way to do matrix multiplication is to use dot product as shown below"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([34400, 50000, 64400])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.dot(price_per_unit,units)"
   ]
  }
 ],
 "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}