├── .gitignore ├── README.md ├── docs ├── Autograding.md ├── Grading Guidlines.md └── Submission Instructions.md ├── images ├── CIAction.png ├── GHActions.png ├── release.png └── runtests.png ├── notebooks ├── Intro to Julia.ipynb └── Intro-Sketch.ipynb └── tex ├── cheatsheet.pdf └── cheatsheet.tex /.gitignore: -------------------------------------------------------------------------------- 1 | .ipynb_checkpoints* 2 | *.aux 3 | *.fls 4 | *.sty 5 | *.log 6 | *.synctex.gz 7 | *.fdb_latexmk 8 | *.swp 9 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # JuliaIntro 2 | Some of the basics for getting started with Julia 3 | 4 | See `/notebooks` for some notebooks covering tips for getting started with Julia. 5 | 6 | See `/docs` for some guidlines for submitting homeworks, using the autograder, homework grading policies, etc. 7 | 8 | See `tex/cheatsheet.pdf` for the Julia cheat sheet pdf. 9 | 10 | For a video covering the basics of getting starting with Julia, see: 11 | 12 | [![IMAGE ALT TEXT HERE](https://img.youtube.com/vi/BNYzFm4-9vo/0.jpg)](https://www.youtube.com/watch?v=BNYzFm4-9vo) 13 | -------------------------------------------------------------------------------- /docs/Autograding.md: -------------------------------------------------------------------------------- 1 | # Autograding 2 | Each homework assignment will come with a set of unit tests that will be used to help us grade your homework. If your code passes all of the provided unit tests, there's a good chance you will get full credit on the assignment. We reserve the right to run your code on a different (more exhaustive) set of unit tests. If you "hack" your solutions to pass the unit tests, or modify the tests themselves (you're welcome to add more tests, just not modify the ones we provide), you will be heavily penalized. 3 | 4 | This document will show you how the autograding process works, how to run the tests locally on your computer, and how to add more tests. 5 | 6 | ## Overview 7 | We use the built-in unit testing functionality in Julia, combined with GitHub actions to run the tests. You are not required to know the details of how this works, but this section provides a high-level overview of how the testing process works. This workflow is identical to the workflow used in published Julia packages (including Julia itself). Each homework assignment is set up as a Julia package. A minimal Julia package has the following directory structure: 8 | ``` 9 | ./PackageName 10 | /src 11 | PackageName.jl 12 | /test 13 | runtests.jl 14 | Manifest.toml 15 | ``` 16 | 17 | ### Source Code 18 | The `src/PackageName.jl` will look something like this: 19 | ```julia 20 | module HW 21 | using NBInclude 22 | 23 | function studentinfo() 24 | info = Dict( 25 | "name" => "Brian Jackson", 26 | "Andrew ID" => "bjackso2" 27 | ) 28 | return info 29 | end 30 | 31 | notebook() = @nbinclude(joinpath(@__DIR__,"hw.ipynb")) 32 | 33 | end 34 | ``` 35 | 36 | You are expected to modify the `studentinfo` command with your 37 | personal information. 38 | 39 | The `notebook()` method will run the code in your notebook (with the scope of the module). Do not modify this method. 40 | 41 | All code you implement and submit for grading should be contained in the `/src` folder. 42 | 43 | ### Test files 44 | The `test/runtests.jl` file will look similar to this: 45 | 46 | ```julia 47 | using HW 48 | using Test 49 | using Pkg 50 | Pkg.status() 51 | 52 | HW.notebook() 53 | 54 | @testset "Question 1" begin 55 | @testset "Part a" begin 56 | @test 1 == 1 57 | # etc. 58 | end 59 | @testset "Part b" begin 60 | @test 2 == 2 61 | # etc. 62 | end 63 | # etc. 64 | end 65 | ``` 66 | 67 | The `HW.notebook()` line will run your Jupyter notebook, saving all methods and variables within the `HW` module. For example, if you define a variable `q1a` in your notebook, it can be accessed via `HW.q1a` in the test suite. The tests will 68 | often be grouped into different test sets, as shown. These may also contain an `include(_test.jl)` command that will include the contents of another file in the test set. 69 | 70 | ## Running Your Tests 71 | ### Autograder on GitHub 72 | Each time you push to your `main` branch, it will trigger a GitHub action that will run your code through the test set. You can check the status of your tests by selecting the "Actions" tap in GitHub (see image below). 73 | ![Action](../images/GHActions.png) 74 | 75 | If you select a "CI" workflow and then select a "Job" on the left it will show the the results of your test in the main window. You can scroll through the "Run julia-actions/julia-runtest@latest" section it will show the terminal output of the tests, and will show you any errors it encountered or any tests that failed. 76 | 77 | ![CI](../images/CIAction.png) 78 | 79 | ### Running the tests locally 80 | To run the tests on your computer you can open a REPL in the root directory of your homework repository. Enter the package manager using `]` and then activate the current environment using `activate .`. You can then run your tests using `test HW`, replacing `HW` with the name of the module in `src` (should match the name of the repository). 81 | 82 | ![tests](../images/runtests.png) 83 | 84 | ## Adding More Tests 85 | If you want to add more tests that will run automatically, feel free to include an extra `@testset` in `runtests.jl`, named "Extra Tests" or something similar. Please do NOT modify or add tests inside of the test sets or test files we provide. 86 | 87 | -------------------------------------------------------------------------------- /docs/Grading Guidlines.md: -------------------------------------------------------------------------------- 1 | # Grading Guidelines 2 | This document details the criteria we will use in grading your homework. 3 | 4 | * If your homework passes all of the provided unit tests, you will very likely receive full credit on the assignment 5 | * We do reserve the right to run your code on a different (more exhaustive) set of tests. We will not test anything that is not asked or specified in the question prompt. 6 | * If you "hack" your code to pass the tests, you will be heavily penalized. This includes techniques such as hard-coding the output of a function to match the expected value in the test. We will be looking at your code and will be able to catch this type of cheating. 7 | * If you modify the provided test suite you will be heavily penalized. 8 | * Extra credit points may be given for good coding practice, such as clean, well-commented code, extra unit tests, elegant solutions, etc. 9 | * Some points may be deducted for extremely poor coding hygiene (no comments, overly complex or convoluted solutions, poor use of spacing, etc.) 10 | -------------------------------------------------------------------------------- /docs/Submission Instructions.md: -------------------------------------------------------------------------------- 1 | # Submission Instructions 2 | Please follow these guidlines for submitting your homework. 3 | 4 | 1. Push your code to your `main` branch on GitHub 5 | 2. Tag your final submission as `v1.0`. If you need to resubmit, just up the minor revision (e.g. `v1.1`). You can also include "Submission" as the message. 6 | 7 | We will use the timestamp on your tagged submission for assessing late penalties. 8 | 9 | You can tag your code either on GitHub or via the command line. 10 | 11 | ## Tagging via GitHub (recommended) 12 | In the Code tab in GitHub, select "tags" below the header tabs or "Create a New Release" in the right panel. See the example below: 13 | 14 | ![release](../images/release.png) 15 | 16 | ## Tagging via Command Line 17 | Alternatively, you can tag via the command line: 18 | ``` 19 | git tag -a v1.0 -m "submission" 20 | git push origin v1.0 21 | ``` 22 | -------------------------------------------------------------------------------- /images/CIAction.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Optimal-Control-16-745/JuliaIntro/d031f9371a0cda8eb72909b12d0a873eef79918c/images/CIAction.png -------------------------------------------------------------------------------- /images/GHActions.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Optimal-Control-16-745/JuliaIntro/d031f9371a0cda8eb72909b12d0a873eef79918c/images/GHActions.png -------------------------------------------------------------------------------- /images/release.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Optimal-Control-16-745/JuliaIntro/d031f9371a0cda8eb72909b12d0a873eef79918c/images/release.png -------------------------------------------------------------------------------- /images/runtests.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Optimal-Control-16-745/JuliaIntro/d031f9371a0cda8eb72909b12d0a873eef79918c/images/runtests.png -------------------------------------------------------------------------------- /notebooks/Intro-Sketch.ipynb: -------------------------------------------------------------------------------- 1 | { 2 | "cells": [ 3 | { 4 | "cell_type": "markdown", 5 | "metadata": {}, 6 | "source": [ 7 | "# Intro to Julia for Roboticists\n", 8 | "This intro assumes a solid understanding of scientific computing, and proficiency in a high-level language such as Matlab or Python.\n", 9 | "The goal of the notebook is to introduce the critical concepts needed to transition quickly to Julia and starting developing code for real problems." 10 | ] 11 | }, 12 | { 13 | "attachments": { 14 | "9c060f58-5e93-415f-8624-a5cda615469d.png": { 15 | "image/png": "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" 16 | } 17 | }, 18 | "cell_type": "markdown", 19 | "metadata": {}, 20 | "source": [ 21 | "## Why Julia?\n", 22 | "**QUESTION**: What language do you use first? \n", 23 | "**QUESTION**: What do you do when performance becomes critical?\n", 24 | "\n", 25 | "Julia attempts to solve the \"two language\" problem by offering a convenient, powerful syntax while offering exceptional performance.\n", 26 | "\n", 27 | "![image.png](attachment:9c060f58-5e93-415f-8624-a5cda615469d.png)" 28 | ] 29 | }, 30 | { 31 | "cell_type": "markdown", 32 | "metadata": {}, 33 | "source": [ 34 | "## How it Compare?\n", 35 | "| | Julia | Matlab | Python | C++ |\n", 36 | "|-|-------|--------|--------|-----|\n", 37 | "|Compilation| JIT | Dynamic | Dynamic | Static |\n", 38 | "|Age| 9 years | 37 years | 30 years | 36 years |\n", 39 | "|OOP?| Multiple Dispatch | Kinda | Yes | Yes |\n", 40 | "|Indexing| 1-based | 1-based | 0-based | 0-based |" 41 | ] 42 | }, 43 | { 44 | "cell_type": "markdown", 45 | "metadata": {}, 46 | "source": [ 47 | "## Installation\n", 48 | "1. Download the current [release](https://julialang.org/downloads/)\n", 49 | "2. Extract somewhere convenient (not in Downloads).\n", 50 | "3. Add `julia-1.x.x/bin` to your system path\n", 51 | "4. All Julia packages and related content gets stored in `~/.julia`" 52 | ] 53 | }, 54 | { 55 | "cell_type": "markdown", 56 | "metadata": {}, 57 | "source": [ 58 | "# Awesome Packages\n", 59 | "Another reason Julia is great is that there are a lot of great packages for scientific computing, and they all play very \"nice\" with each other. Some packages to look at:\n", 60 | "* JuMP.jl: Optimization modeling language\n", 61 | "* DifferentialEquations.jl: State-of-the-art package for solving differential equations\n", 62 | "* ForwardDiff.jl: Fast forward-mode automatic differentiation" 63 | ] 64 | }, 65 | { 66 | "cell_type": "markdown", 67 | "metadata": {}, 68 | "source": [ 69 | "# The Basics\n", 70 | "## Numeric Types\n", 71 | "Julia has all the basics, plus some extras! \\\n", 72 | "**Examples:**\n", 73 | "* Integers\n", 74 | "* Floats (plus `Inf` and `NaN`)\n", 75 | "* Complex\n", 76 | "* Rational\n", 77 | "* Irrational\n", 78 | "\n", 79 | "**Basic Ops**\n", 80 | "* Multiplication\n", 81 | "* Exponentiation\n", 82 | "* Division" 83 | ] 84 | }, 85 | { 86 | "cell_type": "markdown", 87 | "metadata": {}, 88 | "source": [ 89 | "## Strings and Printing\n", 90 | "Julia has both `String` and `Char` types. It comes with some handy methods for working with file paths.\n", 91 | "\n", 92 | "### Examples\n", 93 | "* Construction\n", 94 | "* Indexing\n", 95 | "* File paths\n", 96 | "* Concatenation\n", 97 | "\n", 98 | "**QUESTION**: Why do you think they used multiplication instead of addition? (TIP: think mathematical properties of addition and multiplication)\n", 99 | "\n", 100 | "### Symbols\n", 101 | "Julia also has the `Symbol` type, which basically just a hashed string, useful for comparisons. Basically, use them in place of enums in C/C++.\n", 102 | "\n", 103 | "### Printing\n", 104 | "Julia has some nice options for printing." 105 | ] 106 | }, 107 | { 108 | "cell_type": "markdown", 109 | "metadata": {}, 110 | "source": [ 111 | "Symbols tend to be used when comparison is what it most important. These get hashed so comparison is very fast" 112 | ] 113 | }, 114 | { 115 | "cell_type": "markdown", 116 | "metadata": {}, 117 | "source": [ 118 | "# Arrays\n", 119 | "### Construction\n", 120 | "* Explicit construction (Vector vs Matrix)\n", 121 | "* Initializers (`zeros`, `ones`, `fill`, etc)\n", 122 | "* Comprehension and `map`\n", 123 | "\n", 124 | "### Operations\n", 125 | "* Basic arithmetic (and broadcasting)\n", 126 | "* `push!` vs `append!`\n", 127 | "* Other useful vector operations (e.g. `maximum`, `sort`, `reverse`, `findmax`)\n", 128 | "\n", 129 | "## Assignment\n", 130 | "* Copy vs alias\n", 131 | "* Julia is pass-by-reference for *mutable* types.\n", 132 | "* Testing equality\n", 133 | "\n", 134 | "## Indexing and Slicing\n", 135 | "Julia supports pretty every type of indexing operation:\n", 136 | "* ranges\n", 137 | "* integers\n", 138 | "* booleans\n", 139 | "* linear vs cartesian\n", 140 | "* views" 141 | ] 142 | }, 143 | { 144 | "cell_type": "markdown", 145 | "metadata": {}, 146 | "source": [ 147 | "## Linear Algebra\n", 148 | "For more advanced linear algebra routines we need to load the LinearAlgebra package from the standard Julia library." 149 | ] 150 | }, 151 | { 152 | "cell_type": "code", 153 | "execution_count": 22, 154 | "metadata": {}, 155 | "outputs": [], 156 | "source": [ 157 | "using LinearAlgebra" 158 | ] 159 | }, 160 | { 161 | "cell_type": "markdown", 162 | "metadata": {}, 163 | "source": [ 164 | "### More Types\n", 165 | "With LinearAlgebra, we have a bunch of new matrix types and methods we can work with:\n", 166 | "* Matrix types: `Diagonal`, `Symmetric`, `LowerTriangular` etc.\n", 167 | "* Factorizations: `qr`, `svd`, `eigen`, `cholesky`\n", 168 | "\n", 169 | "### Basic Linear Algebra\n", 170 | "* Matrix multiplication\n", 171 | "* Outer, inner, cross products\n", 172 | "* Methods like `norm` or `eigvals`\n", 173 | "\n", 174 | "### Solving Linear Systems\n", 175 | "* Backslash (inverse, least-squares, least-norm)\n", 176 | "* Using factorization types" 177 | ] 178 | }, 179 | { 180 | "cell_type": "markdown", 181 | "metadata": {}, 182 | "source": [ 183 | "# Functions\n", 184 | "Since Julia is JIT compiled, the first function call will always be much slower.\n", 185 | "\n", 186 | "**Example**\n", 187 | "Sum function in Julia vs Python.\n", 188 | "\n", 189 | "### More Advanced\n", 190 | "* Optional and keyword arguments\n", 191 | "* Varags\n", 192 | "\n", 193 | "### Multiple Dispatch\n", 194 | "Similar to operator overloading, except that it works for all functions and happens at *runtime*.\n", 195 | "\n", 196 | "**Example**\n", 197 | "Simple print function" 198 | ] 199 | } 200 | ], 201 | "metadata": { 202 | "kernelspec": { 203 | "display_name": "Julia 1.5.3", 204 | "language": "julia", 205 | "name": "julia-1.5" 206 | }, 207 | "language_info": { 208 | "file_extension": ".jl", 209 | "mimetype": "application/julia", 210 | "name": "julia", 211 | "version": "1.5.3" 212 | } 213 | }, 214 | "nbformat": 4, 215 | "nbformat_minor": 4 216 | } 217 | -------------------------------------------------------------------------------- /tex/cheatsheet.pdf: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Optimal-Control-16-745/JuliaIntro/d031f9371a0cda8eb72909b12d0a873eef79918c/tex/cheatsheet.pdf -------------------------------------------------------------------------------- /tex/cheatsheet.tex: -------------------------------------------------------------------------------- 1 | \documentclass{article} 2 | \usepackage[utf8]{inputenc} 3 | \usepackage[margin=0.5in]{geometry} 4 | \usepackage{lipsum} 5 | 6 | \usepackage{jlcode} 7 | \usepackage{multicol} 8 | \usepackage{listings} 9 | 10 | \title{Julia Cheatsheet} 11 | \author{Brian Jackson} 12 | 13 | \begin{document} 14 | \maketitle 15 | \begin{multicols*}{3} 16 | \section{Notation} 17 | \texttt{a,b,c} are scalars, 18 | \texttt{x,y,z} are vectors, and 19 | \texttt{A,B,C} are matrices. 20 | \texttt{S} is a square matrix, 21 | \texttt{s} is a string. 22 | \texttt{elw} means element-wise. 23 | 24 | \section{Assignment} 25 | \begin{verbatim} 26 | a = 1 # scalar assignment 27 | A = B # alias assignment 28 | A .= B # element-wise copy 29 | A = copy(B) # copy 30 | A = deepcopy(B) # deep copy 31 | a = 1 + 2im # imaginary 32 | b = 1 // 2 # rational 33 | \end{verbatim} 34 | 35 | \section{Scalar Arithmetic} 36 | \begin{verbatim} 37 | a+2 # addition 38 | a-1 # subtraction 39 | 2*a # multiplication 40 | 2a # multiplication 41 | a/3 # float division 42 | a÷3 # int division (\div) 43 | a^2 # exponential 44 | a%3 # modulus 45 | \end{verbatim} 46 | 47 | \section{Arrays} 48 | \subsection{Initialization} 49 | \vspace*{-2mm} 50 | \begin{verbatim} 51 | [1,2,3] # vector 52 | [1 2 3] # row vector 53 | [1; 2; 3] # col vector 54 | [1 2; 3 4] # 2x2 matrix 55 | zeros(3) # all 0s (vec) 56 | zeros(3,2) # all 0s (mat) 57 | ones(3,2) # all 1s 58 | ones(Int,3,2) # Integer 1s 59 | rand(3,2) # uniform from 0-1 60 | randn(3,2) # Std. Gaussian 61 | fill(10,3,2) # all 10s 62 | 0:10 # integer range 0-10 63 | 0:2:10 # range 0,2,4,...,10 64 | range(0,10,step=2) # same 65 | 0:0.1:10 # step of 0.1 66 | range(0,10,length=101) # same 67 | \end{verbatim} 68 | 69 | \subsection{Arithmetic} 70 | \vspace*{-2mm} 71 | \begin{verbatim} 72 | 2 .+ x # scalar add 73 | 2 .- x # scalar sub 74 | x + y # elw add 75 | x - y # elw sub 76 | x .* y # elw mult 77 | x ./ y # elw div 78 | x' * y # dot product 79 | x'y # dot product 80 | x * y' # outer product 81 | x * y # undefined 82 | A * B # matrix mult 83 | A .* B # elw mult 84 | A .^ 2 # elw square 85 | S^2 # matrix mult 86 | \end{verbatim} 87 | 88 | \subsection{Indexing} 89 | \vspace*{-2mm} 90 | Assume \texttt{size(A) == (3,2)}. 91 | \vspace*{-2mm} 92 | \begin{verbatim} 93 | x[1] # linear index 94 | A[4] # linear index 95 | A[1,2] # row,col (same) 96 | x[2:end] # 2nd to last 97 | x[1:end-2] # 1st to 3rd last 98 | A[:,1] # 1st column 99 | A[1:2,:] # first 2 rows 100 | A[1] = 2 # assign element 101 | A[:,1] .= 2 # assign range 102 | A[:,1] = x # assign range 103 | \end{verbatim} 104 | 105 | \section{Other Types} 106 | \subsection{Strings} 107 | \vspace*{-2mm} 108 | \begin{verbatim} 109 | 'c' # char 110 | "my string" # string 111 | :abc # symbol (fast) 112 | "my num: $a" # interpolation 113 | string("a","b") # concat 114 | "a" * "b1" # concat 115 | s[1] # get char 116 | s[1:2] # sub-string 117 | \end{verbatim} 118 | 119 | \subsection{Dictionaries} 120 | \vspace{-2mm} 121 | \begin{verbatim} 122 | d1 = Dict(:a=>1, :b=>2) 123 | d2 = Dict("d"=>x, "e"=>y) 124 | d1[:a] # indexing 125 | d2["g"] = x+y # new entry 126 | pop!(d2, "g) # remove entry 127 | keys(d1) # get keys 128 | values(d1) # get values 129 | for (k,v) in pairs(d1) 130 | # key k, value v 131 | end 132 | \end{verbatim} 133 | 134 | \subsection{Lists} 135 | \vspace{-2mm} 136 | \begin{verbatim} 137 | [1,2,3] # good 138 | ["a","b","c"] # good 139 | [[1,2],[2]] # good 140 | [1,"b",[1,2]] # avoid 141 | maximum(x) # maximum element 142 | minimum(x) # minimum element 143 | argmax(x) # index of max 144 | findmax(x) # (val,idx) of max 145 | push!(x,1) # add to end 146 | insert!(x,1,5) # add to start 147 | append!(x,y) # concat 148 | [x; y] # vert cat 149 | [x y] # horz cat 150 | vcat(x,y) # vert cat 151 | hcat(x,y) # horz cat 152 | a in x # exists in? 153 | sort(x) # sort 154 | sort!(x) # sort in-place 155 | sortperm(x) # sort indices 156 | \end{verbatim} 157 | 158 | 159 | \subsection{Other} 160 | \vspace{-2mm} 161 | \begin{verbatim} 162 | (1,2,3) # tuple 163 | Set((1,2,3)) # set 164 | \end{verbatim} 165 | 166 | \section{Control Flow} 167 | \subsection{Logic} 168 | \vspace*{-2mm} 169 | \begin{verbatim} 170 | a == b # are equal 171 | A == B # all elm are equal 172 | isapprox(a, b) # \approx 173 | A === B # same memory loc 174 | a != b # not equal 175 | a < b # less than 176 | a <= b # less than or equal 177 | a && b # short-circuit and 178 | a || b # short-circuit or 179 | a < b < c # b between a,c 180 | \end{verbatim} 181 | 182 | \subsection{Conditionals} 183 | \vspace*{-2mm} 184 | \begin{verbatim} 185 | if a < b 186 | # code 187 | elseif b > a 188 | # code 189 | else 190 | # code 191 | end 192 | a < b ? 1 : 0 # inline 193 | (a < b) && 1 # short-circuit 194 | \end{verbatim} 195 | 196 | \subsection{Loops} 197 | \vspace*{-2mm} 198 | \begin{verbatim} 199 | # For loops 200 | for x = 1:10 201 | # loop body 202 | end 203 | for a in x # or \in 204 | # loop body 205 | end 206 | for i = 1:10, j = 1:10 207 | # nested loop 208 | end 209 | 210 | # While Loop 211 | while (a < b) 212 | a += 1 213 | end 214 | 215 | # List comprehension 216 | x = [sin(i) for i = 1:10] 217 | A = [i+j for i in x, j in y] 218 | \end{verbatim} 219 | 220 | \subsection{Functions} 221 | \vspace*{-2mm} 222 | \begin{verbatim} 223 | function myfun(x,y,a=1;b=2) 224 | # function body 225 | return 226 | end 227 | # valid calls 228 | myfun(1,2) 229 | myfun(1,2,3) 230 | myfun(1,2,3,b=3) 231 | myfun(1,2,b=3) 232 | 233 | # anonymous functions 234 | mysum(x,y) = x+y 235 | mysub = (x,y) -> x-y 236 | \end{verbatim} 237 | 238 | 239 | \section{Linear Algebra} 240 | \vspace*{-2mm} 241 | \begin{verbatim} 242 | using LinearAlgebra 243 | norm(x) # 2 norm 244 | norm(x,Inf) # Inf norm 245 | norm(x,p) # p-norm 246 | diag(A) # get diagonal 247 | inv(S) # inverse 248 | eigvals(S) # eigenvalues 249 | rank(S) # rank 250 | cond(S) # condition num 251 | isposdef(S) # x'S*x > 0? 252 | Diagonal(x) # diag mat 253 | Symmetric(S) # symm mat 254 | y = A\x # solve Ax = y 255 | eigen(S) # Eigen decomp 256 | qr(S) # QR fact 257 | svd(S) # SVD fact 258 | cholesky(S) # Cholesky 259 | \end{verbatim} 260 | 261 | \section{Useful Macros} 262 | \subsection{Benchmarking} 263 | \vspace*{-2mm} 264 | \begin{verbatim} 265 | @time f(x) # print time 266 | @elapsed f(x) # get time 267 | @allocated f(x) # get allocs 268 | 269 | # to run many times 270 | using BenchmarkTools 271 | @btime f(x) # print time 272 | @benchmark f(x) # get details 273 | \end{verbatim} 274 | 275 | \subsection{Other} 276 | \vspace*{-2mm} 277 | \begin{verbatim} 278 | # get which method is called 279 | @which f(x) 280 | 281 | # type stability info 282 | @code_warntype f(x) 283 | \end{verbatim} 284 | 285 | \section{Packages} 286 | \vspace*{-2mm} 287 | \begin{verbatim} 288 | # load package to use 289 | using MyPackage 290 | 291 | # Don't import methods 292 | import MyPackage 293 | 294 | # shorten name 295 | const MP = MyPackage 296 | 297 | # load specific methods 298 | using MyPackage: foo, bar 299 | 300 | # load methods to redefine 301 | import MyPackage.foo 302 | \end{verbatim} 303 | 304 | \subsection{Adding/Removing} 305 | In REPL, type \texttt{]} to open 306 | package manager. Here \texttt{Pack} 307 | can be any package name. 308 | \begin{verbatim} 309 | add Pack # add 310 | add Pack@1 # add version 311 | add Pack#master # add branch 312 | rm Pack # remove 313 | activate dir # use env at dir 314 | st # list installed packages 315 | \end{verbatim} 316 | 317 | % \section{Type System} 318 | % \subsection{Basic Ops} 319 | % \vspace*{-2mm} 320 | % \begin{verbatim} 321 | % typeof(a) # Float64 322 | % typeof(x) # Array{Float64,1} 323 | % x isa Vector{Float64} # true 324 | % Vector <: Array # true 325 | % Int <: Number # true 326 | % \end{verbatim} 327 | 328 | % \subsection{Custom Types} 329 | % \vspace*{-2mm} 330 | % \begin{verbatim} 331 | % abstract type Phasors end 332 | % struct P1 <: Phasors 333 | % a::Float64 334 | % b::Float64 335 | % isnorm::Bool 336 | % end # fields can't be changed 337 | % mutable struct P2 <: Phasors 338 | % a::Float64 339 | % b::Float64 340 | % isnorm::Bool 341 | % end # fields can be changed 342 | % function foo(x::Phasors) 343 | % # define foo on both 344 | % x.a + x.b # return this 345 | % end 346 | % \end{verbatim} 347 | 348 | 349 | \end{multicols*} 350 | 351 | \end{document} --------------------------------------------------------------------------------