WEBVTT 1 00:00:00.120 --> 00:00:03.560 Welcome to the deep dive. You send us practical Julia 2 00:00:03.600 --> 00:00:07.559 dot pdf, and we're going to explore why this language, 3 00:00:07.639 --> 00:00:10.839 Julia is making ways, especially in science and tech. 4 00:00:11.039 --> 00:00:13.919 Exactly. Our mission really is to pull out the key 5 00:00:13.960 --> 00:00:16.760 ideas about what makes Julia special and why it could 6 00:00:16.800 --> 00:00:19.120 be a well, a really useful tool for you. 7 00:00:19.559 --> 00:00:22.640 Think of this as the quick guide to Julia's main straints. 8 00:00:23.160 --> 00:00:25.559 We want to hit the highlights what really sets it 9 00:00:25.600 --> 00:00:28.679 apart without getting totally bogged down in every little command. 10 00:00:28.800 --> 00:00:30.039 Yeah, keep it focused. 11 00:00:30.160 --> 00:00:32.479 It's got quite an origin story actually came out of 12 00:00:32.600 --> 00:00:37.280 MIT back in twenty twelve. The creators were they jokingly 13 00:00:37.320 --> 00:00:39.880 call themselves greedy, greedy in what way they wandered it 14 00:00:39.920 --> 00:00:41.799 all in one language. They're the ease of use you 15 00:00:41.799 --> 00:00:43.240 get from something like Python. 16 00:00:43.000 --> 00:00:45.520 Uh huh, the interactive feel, right yeah. 17 00:00:45.280 --> 00:00:47.960 But also the raw speed, the high performance you usually 18 00:00:48.000 --> 00:00:50.240 need languages like C or Fortran four. 19 00:00:50.719 --> 00:00:53.240 Okay, so the best of both worlds. And yeah, if 20 00:00:53.280 --> 00:00:58.280 you've used interactive environments Pythons, repl matt Lab, Octave, some 21 00:00:58.320 --> 00:01:00.560 parts of Julia will feel pretty familiar, like. 22 00:01:00.520 --> 00:01:02.479 Typing a command and getting an answer straight. 23 00:01:02.280 --> 00:01:06.159 Away exactly that and defining variables on the fly, not 24 00:01:06.200 --> 00:01:10.159 always needing to specify their type beforehand. That feels comfortable, 25 00:01:10.280 --> 00:01:10.959 right it does. 26 00:01:11.359 --> 00:01:13.879 But okay, let's pivot slightly because this is where Julia 27 00:01:13.959 --> 00:01:16.079 gets well, really unique, isn't it? 28 00:01:16.359 --> 00:01:20.840 Definitely? What makes it stand out, especially for scientists and engineers, 29 00:01:21.079 --> 00:01:23.239 is this thing called multiple dispatch. 30 00:01:23.480 --> 00:01:26.799 Multiple dispatch sounds technical, it is a bit, but the 31 00:01:26.920 --> 00:01:27.959 concept is key. 32 00:01:28.400 --> 00:01:32.159 It's powered by Julia's type system, which is very sophisticated, 33 00:01:32.599 --> 00:01:35.840 and it's this feature, combined with the interactivity and the 34 00:01:35.879 --> 00:01:37.840 speed that's driving its adoption. 35 00:01:38.280 --> 00:01:40.560 So what does multiple dispatch do differently? 36 00:01:41.000 --> 00:01:43.640 Okay, so imagine you want to add two things in 37 00:01:43.680 --> 00:01:46.040 many languages, the decision of how to add them depends 38 00:01:46.120 --> 00:01:48.239 only on the type of the first thing. Right, With 39 00:01:48.359 --> 00:01:51.000 multiple dispatches, Julia looks at the types of all the 40 00:01:51.040 --> 00:01:54.159 things involved. So adding an integer and a floating point 41 00:01:54.239 --> 00:01:56.920 number might use a different internal method than adding two 42 00:01:57.000 --> 00:01:59.400 integers or adding an integer in a fraction. 43 00:01:59.640 --> 00:02:02.680 Ah. So it picks the most specific recipe based on 44 00:02:02.760 --> 00:02:04.519 everything you give it precisely. 45 00:02:04.920 --> 00:02:08.719 That makes it incredibly flexible and crucially very fast when 46 00:02:08.759 --> 00:02:10.560 you're dealing with all the different kinds of data you 47 00:02:10.599 --> 00:02:11.919 find in scientific computing. 48 00:02:12.159 --> 00:02:15.879 Interesting. Okay, so the material you sent practical Julia. How 49 00:02:15.879 --> 00:02:17.759 does it structure this give us the roadmap? 50 00:02:17.879 --> 00:02:22.639 Sure, it starts naturally with the basics, setting things up, 51 00:02:22.680 --> 00:02:24.919 the syntax, the fundamental building. 52 00:02:24.599 --> 00:02:26.400 Blocks, the usual starting point. Yep. 53 00:02:26.680 --> 00:02:31.280 Then it moves into organizing code, modules, packages, then visualization 54 00:02:31.439 --> 00:02:32.240 which is huge. 55 00:02:32.360 --> 00:02:34.199 Plotting yeah, always important, Then. 56 00:02:34.120 --> 00:02:38.560 Different ways to store data collections. It even gets into metaprogramming. 57 00:02:38.000 --> 00:02:40.039 Whoa code that writes code. 58 00:02:39.879 --> 00:02:44.639 Essentially, Yeah, pretty advanced stuff. Then things like diagrams, animations, 59 00:02:44.879 --> 00:02:49.120 the type sister itself in detail, handling physical units. Wow, okay, 60 00:02:49.360 --> 00:02:54.800 And finally it shows Julia in action across different fields physics, biology, stats, math, 61 00:02:55.039 --> 00:02:58.280 signal processing, image analysis, even parallel computing. 62 00:02:58.400 --> 00:02:59.599 So it covers a lot of ground. 63 00:02:59.680 --> 00:03:03.879 It really does, speaks volumes about Julia's versatility. Actually, following 64 00:03:03.879 --> 00:03:06.599 this structure should give you a really solid grasp from 65 00:03:06.680 --> 00:03:08.759 the core ideas out to the more specialized stuff. 66 00:03:08.840 --> 00:03:11.120 All right, let's dive into that core. Then, getting started, 67 00:03:11.280 --> 00:03:14.319 the book talks about the r EPL that interactive command line. 68 00:03:14.479 --> 00:03:16.400 Uh huh, the read evil print loop. 69 00:03:16.599 --> 00:03:19.159 Right, and it has different modes, like a normal mode 70 00:03:19.199 --> 00:03:19.879 than a shell mode. 71 00:03:20.000 --> 00:03:23.159 Yeah, you hit a semicolon, and you're basically at your 72 00:03:23.199 --> 00:03:26.319 system's command line right inside Julia. Super handy, and. 73 00:03:26.280 --> 00:03:28.240 A package mode too with a bracket. 74 00:03:28.280 --> 00:03:31.319 That's right, And that integrated package system is honestly a 75 00:03:31.319 --> 00:03:34.280 bit of a game changer. Oh so well, finding, installing 76 00:03:34.360 --> 00:03:38.280 and managing external libraries, you know, extra tools people have 77 00:03:38.319 --> 00:03:41.159 written can be a real nightmare in some ecosystem. 78 00:03:41.240 --> 00:03:43.560 Oh yeah, dependency hell exactly. 79 00:03:43.639 --> 00:03:47.360 That Julia's built in package manager, accessed with that Hall 80 00:03:47.479 --> 00:03:50.520 is designed to make that whole process much much smoother. 81 00:03:50.719 --> 00:03:53.360 It's like having the app store right there in the language. 82 00:03:53.400 --> 00:03:56.400 Okay, that sounds genuinely useful. So after the environment, it's 83 00:03:56.439 --> 00:03:59.919 the basic data types, right, the absolute fundamentals yep. 84 00:04:00.080 --> 00:04:07.240 Numbers, integers, floating point, decimals, even fractions, rationals, and things 85 00:04:07.280 --> 00:04:10.840 like infinity and logic, true and false, true and false. 86 00:04:11.360 --> 00:04:13.879 And the book points out something called short circuiting for 87 00:04:14.080 --> 00:04:15.159 nd A and D. 88 00:04:15.360 --> 00:04:18.199 And ore, meaning it stops evaluating early. 89 00:04:18.079 --> 00:04:20.639 Exactly if you have condition one and condition two and 90 00:04:20.680 --> 00:04:23.560 condition one is false, Julia doesn't even bother checking condition 91 00:04:23.600 --> 00:04:26.000 two because the whole thing has to be false. Saves time, 92 00:04:26.480 --> 00:04:27.920 prevents potential errors. 93 00:04:28.079 --> 00:04:31.639 Smart, okay, and it differentiates between zero mb it does 94 00:04:31.839 --> 00:04:34.439 checks if the values are equal, So five equals five 95 00:04:34.480 --> 00:04:37.439 point zero would be true because five equals five point 96 00:04:37.560 --> 00:04:39.800 zero numerically right, But. 97 00:04:40.199 --> 00:04:42.519 Checks for identity. Are they the exact same thing and 98 00:04:42.560 --> 00:04:45.040 memory the same type? So five equals five point zero 99 00:04:45.040 --> 00:04:47.680 is false. One's an integer, one's afloat. 100 00:04:47.480 --> 00:04:50.720 Got it value versus identity, Like checking if two coins 101 00:04:50.759 --> 00:04:53.000 have the same value versus checking if they're the exact 102 00:04:53.040 --> 00:04:53.439 same coin. 103 00:04:53.519 --> 00:04:54.399 That's a perfect analogy. 104 00:04:54.480 --> 00:04:59.160 Yeah, okay, Next up collections putting data together a raise. 105 00:04:59.319 --> 00:05:03.439 First, arrays are fundamental. They can be simplests vectors or 106 00:05:03.480 --> 00:05:06.360 tables like matrices or even higher dimensional things. 107 00:05:06.439 --> 00:05:08.319 And you access elements with square. 108 00:05:08.040 --> 00:05:11.560 Bracket yep, square brackets. And it's important to note you 109 00:05:11.560 --> 00:05:14.399 can get a single element with one index, or a 110 00:05:14.439 --> 00:05:17.560 whole slice like a row or column using a colon red. 111 00:05:17.879 --> 00:05:21.000 And you need the right number of indices for your array, right, 112 00:05:21.040 --> 00:05:23.480 like two indices for a matrix spot on. 113 00:05:24.079 --> 00:05:27.360 Using the wrong number will give you an error. Understanding 114 00:05:27.399 --> 00:05:30.399 how to index and slice arrays efficiently is crucial for 115 00:05:30.519 --> 00:05:31.279 numerical work. 116 00:05:31.600 --> 00:05:35.720 Then strings text data defined with double quotes double quotes. 117 00:05:35.800 --> 00:05:38.639 Yeah, the booknotes they're kind of like lists of characters, 118 00:05:38.639 --> 00:05:41.040 but with a key difference. They're immutable. 119 00:05:41.199 --> 00:05:43.800 Immutable meaning you can't change them once they're created. 120 00:05:43.920 --> 00:05:46.079 Exactly. You can't just swap out one character in the 121 00:05:46.120 --> 00:05:48.399 middle of an existing string, You'd have to create a new. 122 00:05:48.240 --> 00:05:51.680 One, okay. And joining strings uses the asterisker. 123 00:05:51.439 --> 00:05:54.519 Yep concatenates to them. And you can check for characters 124 00:05:54.600 --> 00:05:58.000 using in or for substrings using the a cursin function. 125 00:05:58.240 --> 00:05:59.439 And there's a warning about. 126 00:05:59.240 --> 00:06:02.920 Unicode, because Julia handles Unicode characters, which can take up 127 00:06:02.920 --> 00:06:06.319 different numbers of bytes. Trying to access a specific bite 128 00:06:06.399 --> 00:06:08.360 might land you in the middle of a character, causing 129 00:06:08.439 --> 00:06:10.800 a string index error. Just something to be aware of. 130 00:06:11.079 --> 00:06:15.560 And for multiline text, triple quotes like this very handy. Okay, 131 00:06:16.040 --> 00:06:18.639 what about ranges you mentioned eight point two point one. 132 00:06:18.600 --> 00:06:21.920 Right, ranges define sequences? Eight point two point one means 133 00:06:21.920 --> 00:06:24.199 start at eight step by Naze's two, so go down, 134 00:06:24.480 --> 00:06:26.600 stop when you get to one, or pass it. So 135 00:06:26.720 --> 00:06:30.399 eight six four to two super useful in loops. 136 00:06:30.560 --> 00:06:32.879 Speaking of loops, the four loop you can use in 137 00:06:33.319 --> 00:06:36.040 or or even that fancy symbol uh huh, I'll work. 138 00:06:35.879 --> 00:06:38.600 For iterating, and you can nest them easily even with 139 00:06:38.639 --> 00:06:43.000 a compact comma syntax. For simple cases, loops, work of arranges, arrays, 140 00:06:43.319 --> 00:06:43.959 all sorts. 141 00:06:43.720 --> 00:06:45.959 Of a collections, gotcha tooples are mentioned too. 142 00:06:46.040 --> 00:06:50.600 Parentheses immutable exactly like lists, you can't change defined with 143 00:06:50.920 --> 00:06:53.759 and the in operator works for checking membership and collections 144 00:06:54.040 --> 00:06:56.959 and checks for non membership. Ranges are memory efficient too. 145 00:06:56.959 --> 00:06:59.120 They don't store every number, just the start, step and 146 00:06:59.279 --> 00:06:59.639 end right. 147 00:06:59.639 --> 00:07:03.319 That makes sense, okay. Let's talk functions, reusable blocks of code. 148 00:07:03.120 --> 00:07:06.600 Defined with the function keyword pretty standard. The key concept 149 00:07:06.639 --> 00:07:08.199 the book emphasizes here is. 150 00:07:08.279 --> 00:07:10.920 Scope scope, meaning where variables live. 151 00:07:11.199 --> 00:07:15.040 Pretty much, variables inside a function are typically local to 152 00:07:15.079 --> 00:07:17.800 that function. They don't usually mess with variables of the 153 00:07:17.839 --> 00:07:20.759 same name outside the funder totally well. The book mentions 154 00:07:20.759 --> 00:07:24.279 a nuance with loops, specifically, if you're running code from 155 00:07:24.319 --> 00:07:27.079 a file and you use a variable inside a loop 156 00:07:27.079 --> 00:07:30.000 that has the same name as a global variable outside, 157 00:07:30.240 --> 00:07:34.240 Julia might warn you about potential shadowing or unintended modification. 158 00:07:34.759 --> 00:07:36.800 It's a helpful safety check, okay. 159 00:07:37.000 --> 00:07:39.759 And functions can change their inputs, there's a naming convention. 160 00:07:39.920 --> 00:07:40.160 Yes. 161 00:07:40.560 --> 00:07:43.720 Functions that modify their arguments often end with an exclamation 162 00:07:43.759 --> 00:07:46.600 mark like push to add to an array or pop 163 00:07:46.720 --> 00:07:47.279 to remove. 164 00:07:47.439 --> 00:07:50.199 So the inna day is a warning sign. This might 165 00:07:50.319 --> 00:07:51.800 change your data exactly. 166 00:07:52.079 --> 00:07:55.519 It's a really useful convention. Contrast that with mutable types 167 00:07:55.560 --> 00:07:58.480 like strings, where operations always create new strings makes sense. 168 00:07:58.519 --> 00:08:00.959 Anonymous functions are covered too little one liners yep. 169 00:08:01.120 --> 00:08:05.600 Using the syntax stores great for quick operations and function composition. 170 00:08:05.680 --> 00:08:08.120 Combining functions plus do blocks? 171 00:08:08.240 --> 00:08:09.279 What do blocks for? 172 00:08:09.560 --> 00:08:12.639 They're clean? Way to pass a multiline anonymous function to 173 00:08:12.720 --> 00:08:15.759 another function makes the code much more readable when a 174 00:08:15.759 --> 00:08:18.920 function takes another function as an argument, which happens a 175 00:08:18.920 --> 00:08:19.759 lot in Julia. 176 00:08:19.800 --> 00:08:25.519 Interesting. Okay. Scaling up modules and packages for bigger. 177 00:08:25.240 --> 00:08:29.279 Projects, right, You use using or import to bring code 178 00:08:29.279 --> 00:08:31.959 from other modules into your current scope. The book gives 179 00:08:31.959 --> 00:08:35.639 examples like linear algebra or random from the standard library. 180 00:08:35.960 --> 00:08:38.600 Is there a difference between using and import? 181 00:08:39.120 --> 00:08:43.120 There is a subtle one. Using module brings all exported 182 00:08:43.200 --> 00:08:46.320 names into your name space, while import module only brings 183 00:08:46.440 --> 00:08:49.000 the module name itself, so you have to write module dot, 184 00:08:49.039 --> 00:08:51.679 function name dot, import module dot function one, function two 185 00:08:51.720 --> 00:08:54.639 is also possible. Using is more common for interactive use. 186 00:08:54.840 --> 00:08:56.960 Import can be clearer in large projects. 187 00:08:57.159 --> 00:08:59.440 Got it and this ties back to the package manager 188 00:08:59.440 --> 00:08:59.960 in the RPA. 189 00:09:00.240 --> 00:09:02.799 Absolutely, that's where you add these external packages, and the 190 00:09:02.799 --> 00:09:05.600 book highlights the project dot tom l and manifest dot 191 00:09:05.639 --> 00:09:06.240 tom l file. 192 00:09:06.399 --> 00:09:09.200 Those are the files that tract dependencies exactly. 193 00:09:09.320 --> 00:09:12.840 Project dot com lists your direct dependencies, and manifest dot 194 00:09:12.840 --> 00:09:17.240 com records the exact versions of all dependencies, including dependencies 195 00:09:17.279 --> 00:09:21.320 of dependencies used in your project. This is crucial for reproducibility. 196 00:09:21.879 --> 00:09:25.399 Anyone else or you later on can recreate the exact 197 00:09:25.480 --> 00:09:27.480 environment no more dependency. 198 00:09:27.519 --> 00:09:32.240 Hell, that sounds incredibly valuable for collaboration and just sanity. 199 00:09:32.679 --> 00:09:34.000 How do you find these packages? 200 00:09:34.320 --> 00:09:37.279 GitHub is the main place. The book suggests searching GitHub 201 00:09:37.279 --> 00:09:40.000 and adding language Julia to your search query to narrow 202 00:09:40.039 --> 00:09:40.360 it down. 203 00:09:40.399 --> 00:09:41.960 Good tip and documentation. 204 00:09:42.159 --> 00:09:45.360 There's a central place pkg d D three dot julia 205 00:09:45.440 --> 00:09:49.200 lang dot org, which often hosts documentation for many registered packages. 206 00:09:49.279 --> 00:09:53.399 Excellent okay, visualization, you said it was huge. The plos package. 207 00:09:53.080 --> 00:09:55.679 Plots dot jl is the go to metapackage doesn't do 208 00:09:55.720 --> 00:09:58.240 the drawing itself. It uses different back ends. Gr is 209 00:09:58.279 --> 00:10:01.200 the default. It's fast and usually just work, but you 210 00:10:01.240 --> 00:10:03.440 can install and use others if needed. 211 00:10:03.200 --> 00:10:05.240 And you just use plot often. Yeah. 212 00:10:05.480 --> 00:10:07.679 Plot xy is the basic idea, and if you want 213 00:10:07.679 --> 00:10:09.720 to add to an existing plot, you use plot with 214 00:10:09.759 --> 00:10:10.639 the exclamation mark. 215 00:10:10.679 --> 00:10:14.320 Again ah, Modifying the plot object makes sense, and customization 216 00:10:14.679 --> 00:10:16.399 labels legends all there. 217 00:10:16.480 --> 00:10:19.559 You can set labels, control the legend position legends on 218 00:10:19.679 --> 00:10:23.320 top left, give it a title, legend title, change line styles, LS, 219 00:10:23.720 --> 00:10:27.159 line widths, LW, markers, colors, lots of options. 220 00:10:27.200 --> 00:10:29.080 What about math formulas in plots? 221 00:10:29.159 --> 00:10:31.480 That's where latex strings comes in. You just put an 222 00:10:31.559 --> 00:10:34.759 L before a string like L alpha plus beta and 223 00:10:34.919 --> 00:10:37.799 plots with the right back end will render it as 224 00:10:37.919 --> 00:10:41.320 proper math using latex really nice for technical plots. 225 00:10:41.440 --> 00:10:44.799 Very cool and more complex plots. Contours heat maps YEP. 226 00:10:44.639 --> 00:10:48.320 Contour heat map surface plots are mentioned and arranging multiple 227 00:10:48.320 --> 00:10:49.000 plots together. 228 00:10:49.120 --> 00:10:49.840 How does that work? 229 00:10:50.039 --> 00:10:52.279 You can use the at layout macro to define a 230 00:10:52.320 --> 00:10:55.120 grid structure and then pass multiple plot objects to the 231 00:10:55.159 --> 00:10:59.039 plot function. Makes creating figures with subplots pretty straightforward. 232 00:10:59.080 --> 00:11:04.279 Okay, Moving beyond rays and strings. Other collection types dictionaries. 233 00:11:03.679 --> 00:11:06.919 YEP, dicked for key value pairs like a Python dictionary 234 00:11:07.000 --> 00:11:09.519 or a hash map. You look up values using keys 235 00:11:09.720 --> 00:11:12.559 and the sets set for unordered collections of unique elements 236 00:11:12.879 --> 00:11:16.360 useful for membership, testing and operations like intersect to find 237 00:11:16.440 --> 00:11:19.120 common elements or union to combine elements. 238 00:11:19.120 --> 00:11:20.120 Okay, and structs. 239 00:11:20.399 --> 00:11:23.480 Structs let you define your own custom composite data types. 240 00:11:23.919 --> 00:11:27.039 You bundle related data fields together. Think of defining a 241 00:11:27.080 --> 00:11:29.679 point type with x and y fields. 242 00:11:29.320 --> 00:11:31.799 And named tuples like tuples, but. 243 00:11:31.840 --> 00:11:34.639 Each element has a name, so instead of accessing my 244 00:11:34.759 --> 00:11:38.600 tupule one, you might access my nametuple dot x makes 245 00:11:38.639 --> 00:11:39.600 code more readable. 246 00:11:39.799 --> 00:11:43.360 So lots of tools for organizing data. Let's circle back 247 00:11:43.360 --> 00:11:45.480 to a raise for a bit more advanced concepts. 248 00:11:45.480 --> 00:11:49.559 Definitely, you can create arrays using functions like comprehensions, which 249 00:11:49.559 --> 00:11:52.039 are a very concise way to build arrays based on 250 00:11:52.080 --> 00:11:55.159 loops and conditions. Think I two for i in one 251 00:11:55.200 --> 00:11:55.960 point five. 252 00:11:55.799 --> 00:11:57.720 Ah list comprehensions from Python. 253 00:11:57.840 --> 00:12:01.919 But for arrays exactly, there's also repeat and fill fill 254 00:12:02.000 --> 00:12:04.799 value DIMS creates an array where every element is the 255 00:12:04.840 --> 00:12:07.840 exact same object in memory. If that object is mutable, 256 00:12:07.960 --> 00:12:09.320 changing one changes them all. 257 00:12:09.440 --> 00:12:11.759 Ooh tricky. What about repeat. 258 00:12:11.440 --> 00:12:14.279 BP actually copies the elements, so it behaves more like 259 00:12:14.320 --> 00:12:16.679 you might initially expect good distinction to know. 260 00:12:16.759 --> 00:12:19.279 And helper functions like zeros and ones yep. 261 00:12:19.200 --> 00:12:21.600 Quickly create a rays filled with zeros or ones super 262 00:12:21.600 --> 00:12:24.480 common and reshape lets you change the dimensions of an 263 00:12:24.559 --> 00:12:26.399 array without changing the data itself. 264 00:12:26.559 --> 00:12:30.320 Useful. What about numerical operations combining arrays? 265 00:12:30.600 --> 00:12:36.200 You can concatenate arrays vertically using semicolones AB or horizontally 266 00:12:36.279 --> 00:12:37.639 using spaces AB. 267 00:12:38.000 --> 00:12:40.080 Logical indexing using true false. 268 00:12:40.240 --> 00:12:42.960 Yeah, this is powerful. If you have an array A 269 00:12:43.200 --> 00:12:45.759 and a boolian array B of the same size, AB 270 00:12:45.919 --> 00:12:48.399 gives you only the elements of A where the corresponding 271 00:12:48.440 --> 00:12:49.480 element in B is true. 272 00:12:49.799 --> 00:12:54.960 Great for filtering and matrix stuff transpose multiplications. 273 00:12:54.360 --> 00:12:57.840 The single quote gives you the joint complex conjugate transpose, 274 00:12:57.879 --> 00:13:01.799 which is just the transpose for real matrices. Standard matrix multiplication. 275 00:13:01.360 --> 00:13:05.440 Is just a okay. And iterating enumerate and zip. 276 00:13:05.480 --> 00:13:10.039 Enumerate collection gives you pairs of index value as you loop. 277 00:13:10.720 --> 00:13:13.679 ZIP call one, call two gives you pairs of value 278 00:13:13.679 --> 00:13:15.320 one value two from each collection. 279 00:13:15.759 --> 00:13:18.039 Very useful, and pairs for dictionary. 280 00:13:17.759 --> 00:13:21.399 Peirs iterates through key value pairs. The book also warns 281 00:13:21.399 --> 00:13:25.000 about mixing up the iteration number from enumerate with say 282 00:13:25.320 --> 00:13:28.559 the character index in a string, especially with unicode. They 283 00:13:28.600 --> 00:13:29.519 aren't always the same. 284 00:13:29.639 --> 00:13:31.960 Good point. Okay, this feels like we're getting into some 285 00:13:32.000 --> 00:13:35.519 really powerful features now, functions, metaprogramming errors. 286 00:13:35.679 --> 00:13:39.399 Yeah, this is where Julia really starts to flex its muscles. Vargs. 287 00:13:39.440 --> 00:13:43.759 For instance, using its in a function definition lets it 288 00:13:43.799 --> 00:13:48.039 accept any number of arguments like printal, NBC exactly. Keyword 289 00:13:48.080 --> 00:13:51.200 arguments are also common, letting you pass optional arguments by 290 00:13:51.320 --> 00:13:55.600 name like plot x, y, line style dot. There's even 291 00:13:55.600 --> 00:13:58.639 a concise syntax if the keyword is just a variable name. 292 00:13:58.519 --> 00:14:01.240 And splatting illy guy in a call. Right. 293 00:14:01.519 --> 00:14:04.039 If you have arguments in a tuple or dictionary, you 294 00:14:04.080 --> 00:14:08.399 can splat them into a function call. My funk ARGs 295 00:14:08.440 --> 00:14:11.879 passes elements of the tupple ARBs as individual arguments. For 296 00:14:12.000 --> 00:14:15.200 dictionaries with symbol keys, you use myfunk quarks. 297 00:14:15.399 --> 00:14:18.240 Handy string interpolation with docker yep. 298 00:14:18.600 --> 00:14:22.000 Variable name is expression embeds values directly in strings, and 299 00:14:22.159 --> 00:14:25.039 raw strings are useful if you have lots of backslashes 300 00:14:25.120 --> 00:14:26.200 or dollar signs you don't want. 301 00:14:26.240 --> 00:14:28.759 Interpreted operators are functions. You can define your. 302 00:14:28.639 --> 00:14:32.360 Own yes plus thalk. They're all just functions, and you 303 00:14:32.360 --> 00:14:35.519 can define your own custom infix operators, though you need 304 00:14:35.519 --> 00:14:38.320 to follow rules about allowed characters and precedents, which are 305 00:14:38.360 --> 00:14:39.320 in the Julida docs. 306 00:14:39.440 --> 00:14:42.159 Higher order functions, map, filter. 307 00:14:42.000 --> 00:14:46.279 Reduce standard functional programming tools. Map applies a function to 308 00:14:46.279 --> 00:14:49.799 each element, Filter keeps elements based on a condition. Reduce 309 00:14:49.840 --> 00:14:52.919 combines elements using a function. The book notes to be 310 00:14:52.960 --> 00:14:56.039 careful with reduce and non associated operators like division. 311 00:14:56.320 --> 00:14:59.440 The order matters and do blocks pop up again here? 312 00:14:59.639 --> 00:15:03.440 Yep? Because map, filter, etc. Often take functions as arguments. 313 00:15:03.720 --> 00:15:07.159 Do blocks make the syntax cleaner for multiline function arguments? 314 00:15:07.279 --> 00:15:08.039 Okay? 315 00:15:07.919 --> 00:15:12.519 Symbol type symbols some name are like efficient unique strings, 316 00:15:12.559 --> 00:15:15.320 often used internally or for things like dictionary keys or 317 00:15:15.399 --> 00:15:18.720 keyword arguments like we saw in plotting attributes line style. 318 00:15:18.559 --> 00:15:24.120 Dot dash right and then metaprogramming code about code. 319 00:15:24.279 --> 00:15:27.679 This is advanced but powerful. You can capture Julia code 320 00:15:27.720 --> 00:15:31.480 as data using quote, dot end or race. This creates 321 00:15:31.519 --> 00:15:34.759 an EXPRUNO object representing the code structure. 322 00:15:34.360 --> 00:15:35.879 And you can manipulate that structure. 323 00:15:35.919 --> 00:15:38.200 You can, and then you can evaluate it using Evil, 324 00:15:38.240 --> 00:15:41.120 though the book rightly warns to use Evil carefully. You 325 00:15:41.120 --> 00:15:44.320 can even write functions that programmatically create variables. 326 00:15:44.519 --> 00:15:46.960 And macros build on this at macro. 327 00:15:46.759 --> 00:15:49.840 Name macros are functions that run at parse time. They 328 00:15:49.879 --> 00:15:53.559 take code expressions as input and return new code expressions, 329 00:15:53.600 --> 00:15:57.200 which then replace the macro call. It's like programmable copy paste, 330 00:15:57.279 --> 00:15:58.240 but much smarter. 331 00:15:58.519 --> 00:16:00.480 What are they used for? Things? 332 00:16:00.879 --> 00:16:05.840 Simplifying syntax, generating repetitive code, optimizing performance. The book mentions 333 00:16:05.879 --> 00:16:08.440 at time and at a lapse for basic timing at 334 00:16:08.480 --> 00:16:12.399 b time from benchmark tools for proper benchmarking at code 335 00:16:12.399 --> 00:16:16.320 worn type to check type stability crucial for performance, and 336 00:16:16.360 --> 00:16:21.440 at FastMATH to allow aggressive, potentially non standard floating point optimizations. 337 00:16:21.679 --> 00:16:24.240 That at code one type sounds really useful for squeezing 338 00:16:24.240 --> 00:16:24.919 out speed. 339 00:16:24.720 --> 00:16:28.799 It absolutely is. And macros need sometimes to control variable scoping, 340 00:16:28.919 --> 00:16:32.039 ensuring variables from the macro's definition don't clash with variables 341 00:16:32.039 --> 00:16:34.360 where the macro is used. It's called hygiene. 342 00:16:34.559 --> 00:16:37.360 Okay, deep stuff. Finally, error handling essential. 343 00:16:37.759 --> 00:16:40.639 You get stack traces when errors occur, showing the sequence 344 00:16:40.639 --> 00:16:44.080 of function calls. You can handle expected errors gracefully using 345 00:16:44.120 --> 00:16:44.759 triy dot. 346 00:16:44.639 --> 00:16:47.000 Catch blocks, and create your own errors. 347 00:16:46.919 --> 00:16:50.279 YEP using throw, often with specific error types. It domain 348 00:16:50.360 --> 00:16:52.879 er if an import value is invalid. Plus warn and 349 00:16:53.159 --> 00:16:57.320 error for logging messages without necessarily stopping execution for warnings. 350 00:16:57.360 --> 00:17:04.240 Right comprehensive error handling. What about graphic beyond plotting diagrams animations. 351 00:17:03.679 --> 00:17:07.799 The book introduces Luxor dot JL for creating static vector graphics. 352 00:17:07.880 --> 00:17:11.519 It's imperative you issue commands like set color, draw circle, draw. 353 00:17:11.400 --> 00:17:14.359 Text, like programming a drawing step by step exactly. 354 00:17:14.480 --> 00:17:17.680 The example is a solar system diagram pulling data from NASA, 355 00:17:17.960 --> 00:17:20.640 drawing planets, setting colors, dash styles, et cetera. 356 00:17:20.799 --> 00:17:22.079 Cool and animations. 357 00:17:22.160 --> 00:17:24.839 That's where Javis dot jl comes in built on Luxor. 358 00:17:25.119 --> 00:17:30.119 You define objects and then act on them over time, rotate, translate, scale, 359 00:17:30.359 --> 00:17:33.519 dot opacity, then you render the result as a video 360 00:17:33.720 --> 00:17:34.119 or gift. 361 00:17:34.359 --> 00:17:35.359 Examples given and. 362 00:17:35.319 --> 00:17:38.480 Epicycles animation is mentioned, and a simulation of a vibrating 363 00:17:38.480 --> 00:17:41.559 circle visualized with a heat map and polar coordinates. Very 364 00:17:41.559 --> 00:17:45.640 neat ways to visualize dynamic processes and interactive plots. Briefly 365 00:17:45.680 --> 00:17:49.119 touches on using Pluto notebooks. With the at bind macro, 366 00:17:49.480 --> 00:17:52.960 you can link HTML sliders or buttons to Julia variables, 367 00:17:53.119 --> 00:17:55.759 making visualizations interactive within the notebook. 368 00:17:55.880 --> 00:17:58.119 Okay, now the type system we mentioned. 369 00:17:57.839 --> 00:18:01.160 It's key, absolutely fundamental. Every value in Julia has a 370 00:18:01.200 --> 00:18:04.440 type tells you what it is. We have basic numeric types, 371 00:18:04.440 --> 00:18:08.160 but also parametric types like rational and sixty four a 372 00:18:08.279 --> 00:18:10.559 rational number made from sixty four bit integers. 373 00:18:10.880 --> 00:18:13.039 And you can check types with isa. 374 00:18:12.920 --> 00:18:16.240 Ice value type checks if the value is of that 375 00:18:16.359 --> 00:18:19.839 type or a subtype. This leads to the type hierarchy, 376 00:18:19.920 --> 00:18:22.440 a tree of types sort of Every type has a 377 00:18:22.480 --> 00:18:24.440 supertype going all the way up to the root type, 378 00:18:24.440 --> 00:18:27.200 and meet types lower down or subtypes. The book even 379 00:18:27.240 --> 00:18:31.160 shows visualizing this hierarchy using graph packages. Concrete types are 380 00:18:31.200 --> 00:18:32.680 the ones that can actually have instances. 381 00:18:32.920 --> 00:18:34.400 Can you declare variable types? 382 00:18:34.480 --> 00:18:38.200 You can globally globalgf dot float sixty four egles one 383 00:18:38.279 --> 00:18:42.400 point zero, or locally x dot nt equals five. This 384 00:18:42.640 --> 00:18:44.960 x is an assution. Assigning the wrong type will cause 385 00:18:45.000 --> 00:18:48.480 an error. Helps catch bugs and sometimes aids performance. 386 00:18:48.000 --> 00:18:50.960 And defining your own types abstract versus concrete. 387 00:18:51.160 --> 00:18:54.160 You use abstract type my abstract type end to define 388 00:18:54.160 --> 00:18:57.640 concepts and struct my concrete type end for actual data 389 00:18:57.640 --> 00:19:01.640 structures with fields, you can specify subtyping relationships using a. 390 00:19:02.079 --> 00:19:03.720 So struct my concrete type my. 391 00:19:03.720 --> 00:19:07.519 Abstract precisely, and structs can be parametric too, like structs 392 00:19:07.559 --> 00:19:10.480 selips t ND where t could be the number type 393 00:19:10.559 --> 00:19:13.440 used for coordinates. You can add constraints using war like 394 00:19:13.480 --> 00:19:15.079 selips T, whar T real. 395 00:19:15.039 --> 00:19:17.440 Very flexible m plot recipes. 396 00:19:17.240 --> 00:19:20.000 Ah at recipe and at user plot. These let you 397 00:19:20.039 --> 00:19:22.319 teach the plots package how to plot your custom types 398 00:19:22.359 --> 00:19:25.279 automatically super powerful for integration. 399 00:19:25.039 --> 00:19:27.960 Makes your own types first class citizens for plotting nice. 400 00:19:28.000 --> 00:19:29.200 What about physical units? 401 00:19:29.279 --> 00:19:31.759