WEBVTT 1 00:00:00.080 --> 00:00:02.240 Welcome to the deep dive, and today we're going to 2 00:00:02.279 --> 00:00:05.360 be taking a look at the world of racket programming. 3 00:00:05.519 --> 00:00:07.440 You know, I've got this whole stack of materials from 4 00:00:07.519 --> 00:00:10.679 racket programming the fun way, and it looks like you're 5 00:00:10.720 --> 00:00:13.199 ready to kind of move past just the basic stuff. 6 00:00:13.000 --> 00:00:16.719 Right, Yeah, definitely, we'll be going, I guess, beyond just 7 00:00:16.760 --> 00:00:20.719 the syntax, like really looking at the design choices that 8 00:00:20.800 --> 00:00:22.960 make racket kind of stand out. So we're talking about 9 00:00:22.960 --> 00:00:25.719 things like immutable data, or maybe the power of let 10 00:00:25.719 --> 00:00:29.800 for variable scope, even how racket handles recursion in an 11 00:00:29.839 --> 00:00:30.559 efficient way. 12 00:00:31.039 --> 00:00:35.799 Okay, cool, So let's start with data. The book jumps 13 00:00:35.799 --> 00:00:38.920 into you know, lists, infectors, and even these things called 14 00:00:38.960 --> 00:00:43.920 con cells. What's different about about how racket deals with data. 15 00:00:44.479 --> 00:00:47.079 Well, a lot of rackets data structures really reflect this 16 00:00:47.079 --> 00:00:51.280 this underlying philosophy of like clarity and correctness. So for example, 17 00:00:51.280 --> 00:00:54.520 you'll run into both immutable and mutable data. You know, 18 00:00:54.520 --> 00:00:56.240 so immutable data you can change it after you create it, 19 00:00:56.280 --> 00:00:58.640 but immutable data, it's kind of like a constant, which 20 00:00:58.679 --> 00:01:01.759 is really helpful for preventing you know, those unexpected side 21 00:01:01.759 --> 00:01:02.640 effects in your code. 22 00:01:02.759 --> 00:01:04.319 I see, Yeah, that makes sense. So it helps keep 23 00:01:04.319 --> 00:01:08.159 things predictable. So how do these cons cells fit in 24 00:01:08.239 --> 00:01:08.640 all of this? 25 00:01:08.879 --> 00:01:11.280 Oh, con cells those are like the foundation of lists 26 00:01:11.280 --> 00:01:15.000 and racket. Think of them like linked train cars. I 27 00:01:15.000 --> 00:01:17.239 guess each car is holding some data and then it 28 00:01:17.280 --> 00:01:19.719 also has like a link to the next car, and 29 00:01:20.040 --> 00:01:22.359 you know that this structure makes it really efficient to 30 00:01:22.400 --> 00:01:24.719 work with lists. Actually, it's it's pretty similar to how 31 00:01:24.760 --> 00:01:28.439 lists work in languages like Lisp, which Racket bars some 32 00:01:28.519 --> 00:01:29.879 ideas from So it's like a. 33 00:01:29.879 --> 00:01:32.599 Data chain you can add or remove links. Speaking of 34 00:01:32.640 --> 00:01:36.439 manipulating data, the source mentions define and set. What's the difference. 35 00:01:36.920 --> 00:01:39.680 So define you can think of that as like, you know, 36 00:01:39.840 --> 00:01:42.120 establishing a new piece of data in your program. It's 37 00:01:42.120 --> 00:01:45.719 like creating a new variable. I guess set though, that's 38 00:01:45.760 --> 00:01:48.040 that's used to actually modify data that's already there, the 39 00:01:48.120 --> 00:01:52.840 mutable kind. And keeping them separate really helps programmers, you know, 40 00:01:52.879 --> 00:01:55.079 think about which parts their code can change, which parts 41 00:01:55.120 --> 00:01:59.239 are going to stay the same. It's all about control really, okay. 42 00:01:59.000 --> 00:02:02.280 So define kind of sets up the board and then set. Yeah, 43 00:02:02.439 --> 00:02:04.959 that's you move the pieces around where you need to 44 00:02:05.439 --> 00:02:09.680 the book also talks about different types of forms in racket, 45 00:02:09.960 --> 00:02:10.960 what are those all about? 46 00:02:11.479 --> 00:02:15.560 So at their core, forms are basically the instructions you 47 00:02:15.560 --> 00:02:17.800 know that you give to Racket. But here's the thing. 48 00:02:18.879 --> 00:02:22.360 Some forms they'll evaluate every single one of their arguments, 49 00:02:22.400 --> 00:02:26.400 while others might only evaluate some or even none. And 50 00:02:27.240 --> 00:02:30.080 this concept, this is actually really key to understanding how 51 00:02:30.159 --> 00:02:32.879 Racket actually runs your code. It can be super powerful 52 00:02:32.879 --> 00:02:34.639 for controlling the flow of your program. 53 00:02:34.879 --> 00:02:37.919 So some forms are very strict about following the recipe 54 00:02:38.039 --> 00:02:41.639 using all the ingredients, and others are a bit more flexible. 55 00:02:41.680 --> 00:02:43.759 I guess what are some of the common types of 56 00:02:43.800 --> 00:02:45.400 forms you might see in Racket. 57 00:02:45.400 --> 00:02:48.039 Oh, you'll you'll use if a lot that's for conditional logic, 58 00:02:48.080 --> 00:02:50.680 and then there's gone that's for more complex branching. And 59 00:02:50.719 --> 00:02:53.360 of course you have your looping tools like four and four. 60 00:02:54.080 --> 00:02:57.479 But the real trick is understanding how each form evaluates, 61 00:02:57.520 --> 00:02:59.439 because that's, you know, that's how you make sure your 62 00:02:59.520 --> 00:03:01.319 Racket program behave the way you expect. 63 00:03:01.400 --> 00:03:04.919 All right, that makes sense. Now let's move on to well, 64 00:03:05.280 --> 00:03:09.400 something that every programmer deals with, numbers. I'm curious how 65 00:03:09.439 --> 00:03:12.360 does Racket handle different kinds of numbers. 66 00:03:12.520 --> 00:03:15.479 Racket's numerical system is, well, it's actually quite sophisticated. It 67 00:03:15.520 --> 00:03:18.520 goes beyond just you know, simple integers and decimals. It 68 00:03:18.599 --> 00:03:21.840 supports things like rationals and even complex numbers. But the 69 00:03:21.879 --> 00:03:25.439 interesting part is that racket distinguishes between what we call 70 00:03:25.599 --> 00:03:29.599 exact numbers like rationals and approximations like those decimals. 71 00:03:29.680 --> 00:03:32.240 So like if I had thirteen as irrational, that would 72 00:03:32.240 --> 00:03:34.599 be exact, but if I had point three to three three, well, 73 00:03:34.599 --> 00:03:35.759 that would just be in approximation. 74 00:03:35.919 --> 00:03:36.120 Right. 75 00:03:37.360 --> 00:03:38.840 Why is that different so important? 76 00:03:39.800 --> 00:03:43.840 Well, because by carefully handling these different number types, you 77 00:03:43.840 --> 00:03:46.319 can actually achieve a lot more precision. And that means 78 00:03:46.319 --> 00:03:50.120 you can avoid those pesky rounding errors that you know, 79 00:03:50.199 --> 00:03:52.960 can just creep up in complex computations. I mean think 80 00:03:53.000 --> 00:03:56.479 about it in financial applications, for instance, using exact rationals 81 00:03:56.479 --> 00:04:00.360 for your calculations, that can be absolutely crucial to UH 82 00:04:00.400 --> 00:04:02.000 to make sure you get accurate results. 83 00:04:02.159 --> 00:04:04.520 Yeah, I see your point there. That's that's really important. 84 00:04:04.840 --> 00:04:09.240 And speaking of complex computations, the outline mentioned something called 85 00:04:09.520 --> 00:04:12.919 UH tail call optimization. Can you explain what that is 86 00:04:12.960 --> 00:04:15.159 and why it's why it's such a big deal in racket. 87 00:04:15.520 --> 00:04:19.959 So tail call optimization, it's it's essentially racket's way of 88 00:04:20.000 --> 00:04:24.560 making recursion, which is a pretty powerful technique where you know, 89 00:04:24.680 --> 00:04:28.720 a function calls itself more efficient. Basically, when a recursive 90 00:04:28.759 --> 00:04:31.560 call is the very last thing a function does, Racket 91 00:04:31.600 --> 00:04:34.360 doesn't have to remember all the information about, you know, 92 00:04:34.480 --> 00:04:38.720 each recursive step along the way. This prevents stack overflow errors, 93 00:04:38.759 --> 00:04:40.959 which are well, they can be a real pain when 94 00:04:40.959 --> 00:04:42.680 you're dealing with a lot of recursive calls. 95 00:04:42.800 --> 00:04:44.879 So it's like being able to climb down a ladder 96 00:04:44.879 --> 00:04:47.079 without worrying about, you know, running out of rungs on 97 00:04:47.079 --> 00:04:47.720 the way back up. 98 00:04:47.920 --> 00:04:51.199 Yeah, exactly, a good analogy, and that's why that's why 99 00:04:51.240 --> 00:04:53.199 recursion is so powerful in Racket. You can use it 100 00:04:53.240 --> 00:04:55.680 without having to, you know, worry too much about those 101 00:04:55.680 --> 00:04:56.560 efficiency issues. 102 00:04:56.920 --> 00:04:58.879 So Racket's got our back when it comes to making 103 00:04:58.920 --> 00:05:04.079 recursion work works smoothly. The book also mentions let let 104 00:05:04.480 --> 00:05:07.800 and then this idea of of lexical environments. Those sound 105 00:05:07.839 --> 00:05:09.759 a bit more abstract though they. 106 00:05:09.879 --> 00:05:12.000 Are a bit abstract, but they're actually really important for 107 00:05:12.120 --> 00:05:17.759 understanding how Racket handles variables. A lexical environment, it's basically 108 00:05:18.360 --> 00:05:20.800 a region, a zone in your code, I guess where 109 00:05:20.839 --> 00:05:23.600 specific variables are defined. It's kind of like a self 110 00:05:23.639 --> 00:05:26.040 contained dictionary of names and values, you know. 111 00:05:26.279 --> 00:05:28.319 So it's like each section of code has its own 112 00:05:28.920 --> 00:05:31.839 its own little glossary of terms. How do how do 113 00:05:32.000 --> 00:05:33.399 let and let fit into this? 114 00:05:33.680 --> 00:05:35.600 Well, let and let There are the tools you use 115 00:05:35.680 --> 00:05:38.480 to actually create those glossaries, those lexical environments. You use 116 00:05:38.519 --> 00:05:41.920 them to to introduce new variables, give them specific values. 117 00:05:42.120 --> 00:05:45.800 It helps keep your code organized, and it prevents variables 118 00:05:45.879 --> 00:05:48.480 from you know, accidentally messing with each other in different 119 00:05:48.519 --> 00:05:49.399 parts of your program. 120 00:05:50.000 --> 00:05:53.519 So it's like you're organizing your variables like index cards 121 00:05:53.560 --> 00:05:56.920 or something, each card defining you know, specific term within 122 00:05:56.959 --> 00:06:02.680 its own context. Now, the outline also mentions recursion. Now, 123 00:06:02.680 --> 00:06:05.800 I know for some programmers that word it brings back 124 00:06:05.839 --> 00:06:08.560 some bad memories. Is it as scary as it sounds? 125 00:06:08.759 --> 00:06:11.879 Well, recursion it can seem pretty daunting at first, I'll 126 00:06:11.879 --> 00:06:16.399 admit that, but it's really a fundamental concept in computer science. 127 00:06:16.439 --> 00:06:19.519 It's a super powerful tool in racket. It's all about, 128 00:06:19.519 --> 00:06:21.600 you know, solving a problem by breaking it down into 129 00:06:21.600 --> 00:06:24.319 smaller but kind of self similar subproblems. 130 00:06:24.399 --> 00:06:26.800 Yeah, I remember struggling with recursion when I was first 131 00:06:26.879 --> 00:06:29.360 learning to code. Can you give me like a concrete 132 00:06:29.360 --> 00:06:31.439 example of how it works in racket. 133 00:06:31.519 --> 00:06:34.560 Sure. Imagine imagine you want to calculate the factorial of 134 00:06:34.600 --> 00:06:38.199 a number that's that's the product of all the positive 135 00:06:38.199 --> 00:06:41.879 integers less than or equal to that number. So in racket, 136 00:06:41.920 --> 00:06:44.319 you could write a recursive function that first checks if 137 00:06:44.319 --> 00:06:46.439 the input number is one. If it is, then the 138 00:06:46.439 --> 00:06:49.480 function just returns one. That's the base case. If it's 139 00:06:49.519 --> 00:06:53.040 not one, then then the function multiplies the number by 140 00:06:53.079 --> 00:06:56.160 the factorial of the number minus one. That's the recursive step. 141 00:06:56.360 --> 00:06:59.040 So it's like this chain reaction where each step triggers 142 00:06:59.120 --> 00:07:01.240 the next one into at that base case, and then 143 00:07:01.240 --> 00:07:04.120 the results kind of like cascade back up exactly. 144 00:07:04.160 --> 00:07:06.160 You got it. And remember, thanks to that tail call 145 00:07:06.240 --> 00:07:10.360 optimization we talked about earlier, Racket can handle these, you know, 146 00:07:10.480 --> 00:07:13.800 even these really deeply nested recursive calls without without breaking 147 00:07:13.839 --> 00:07:14.360 a sweat. 148 00:07:14.399 --> 00:07:18.519 So racket makes recursion both powerful and safe to use. 149 00:07:19.120 --> 00:07:23.240 I like that. The outline also mentions closures and memorization. 150 00:07:23.680 --> 00:07:26.720 Now those sound a bit like magic spells to me. 151 00:07:27.160 --> 00:07:31.800 They kind of do, but they're really powerful techniques that 152 00:07:31.800 --> 00:07:35.240 that make racket even more capable. A closure, it's basically 153 00:07:35.319 --> 00:07:39.360 a function that that can remember the environment. It was 154 00:07:39.399 --> 00:07:41.879 created in even after that environment's gone, it's like it 155 00:07:41.920 --> 00:07:44.360 has this little bit of memory attached to it. 156 00:07:44.560 --> 00:07:47.360 Can you give me a real world scenario where that 157 00:07:47.360 --> 00:07:48.199 that would be useful. 158 00:07:48.319 --> 00:07:52.240 Let's say you have a function that calculates compound interest. 159 00:07:52.759 --> 00:07:56.079 You could use a closure to UH to capture the 160 00:07:56.079 --> 00:07:58.279 initial balance and the interest rate. That way, you could 161 00:07:58.319 --> 00:08:00.560 call that function over and over again, you know, to 162 00:08:00.560 --> 00:08:03.399 calculate the balance at different points in time without having 163 00:08:03.439 --> 00:08:05.560 to re enter those initial values each time. 164 00:08:05.959 --> 00:08:07.959 So it's like having a personal banker that keeps track 165 00:08:08.000 --> 00:08:13.160 of all your investment details. Yeah, handy, What about what 166 00:08:13.240 --> 00:08:14.959 about memmalization? How does that work? 167 00:08:15.040 --> 00:08:19.680 Memorization? It's all about efficiency. It's basically a technique where 168 00:08:19.720 --> 00:08:22.560 you store the results of these expensive computations so that 169 00:08:22.600 --> 00:08:25.720 you don't have to repeat them every single time. And 170 00:08:25.759 --> 00:08:28.160 it's particularly useful for functions that have a lot of 171 00:08:28.160 --> 00:08:31.439 repeated calculations, like like the Fibonacci sequence. 172 00:08:31.800 --> 00:08:34.320 Oh right, like having a cheat sheet for those those 173 00:08:34.360 --> 00:08:37.600 tricky math problems. Does does racket have like built in 174 00:08:37.679 --> 00:08:39.759 support for memorization or do you have to kind of 175 00:08:39.840 --> 00:08:40.440 roll your own? 176 00:08:40.519 --> 00:08:42.480 Oh? No, racket makes it makes it really easy. You 177 00:08:42.480 --> 00:08:45.480 can use the cash function, and that'll that'll memoize the 178 00:08:45.559 --> 00:08:47.879 results of any function you give it. It can really 179 00:08:47.879 --> 00:08:50.200 speed up your programs, especially if they involve a lot 180 00:08:50.240 --> 00:08:52.240 of those you know, repetitive calculations. 181 00:08:52.480 --> 00:08:55.320 I'm starting to see how these these seemingly advanced concepts 182 00:08:55.320 --> 00:08:59.200 they're actually really practical and easy to use in racket 183 00:08:59.759 --> 00:09:02.159 and being a practical Our online mentions a really cool 184 00:09:02.200 --> 00:09:08.240 application of racket analyzing stock data. How does racket handle 185 00:09:08.600 --> 00:09:10.600 you know, real world data like that. 186 00:09:10.960 --> 00:09:13.480 Well, Racket's real strength is that it can combine you know, 187 00:09:13.519 --> 00:09:17.639 this really powerful computation with with practical data manipulation. It's 188 00:09:17.759 --> 00:09:20.919 it's got some robust io capabilities for reading data from 189 00:09:20.960 --> 00:09:24.600 all sorts of places like files, databases, even web APIs. 190 00:09:24.759 --> 00:09:27.440 So it's like a data import texport wizard. Yeah, but 191 00:09:27.879 --> 00:09:30.120 what can you actually do with this this data once 192 00:09:30.159 --> 00:09:31.399 you've got it into Racket. 193 00:09:31.480 --> 00:09:34.759 Well, imagine you've got a CSV file, you know, with 194 00:09:34.919 --> 00:09:38.960 historical stock prices. You could use Racket to read in 195 00:09:39.080 --> 00:09:42.000 that data. You could filter it based on you know, 196 00:09:42.080 --> 00:09:46.759 specific criteria like maybe date ranges or stock symbols, and 197 00:09:46.799 --> 00:09:50.480 you could even do calculations things like moving averages or 198 00:09:50.559 --> 00:09:53.879 standard deviation to look at like trends and volatility. 199 00:09:54.000 --> 00:09:56.240 It sounds like you could build a pretty sophisticated financial 200 00:09:56.240 --> 00:09:57.759 analysis tool in Racket. Oh. 201 00:09:57.840 --> 00:10:01.120 Absolutely. You could even combine that analysis with with rackets 202 00:10:01.120 --> 00:10:04.799 plotting libraries to actually visualize the data, you know, create 203 00:10:04.879 --> 00:10:08.039 charts and graphs that really help you understand those those 204 00:10:08.080 --> 00:10:11.879 market trends and maybe even make some smarter investment decisions. 205 00:10:12.039 --> 00:10:14.519 Now we're talking it's like having a personal data analyst 206 00:10:14.639 --> 00:10:16.919 right there in your computer. But before we get too 207 00:10:16.960 --> 00:10:20.480 lost in the world of finance, let's shift gears a 208 00:10:20.519 --> 00:10:23.279 bit and talk about search algorithms. I know that Racket 209 00:10:23.279 --> 00:10:25.960 has a pretty impressive set of tools for, you know, 210 00:10:26.000 --> 00:10:28.679 for exploring different search strategies. 211 00:10:29.000 --> 00:10:31.679 That's right, Racket's a great platform for for implementing and 212 00:10:31.720 --> 00:10:35.360 experimenting with those you know, those classic search algorithms. We're 213 00:10:35.399 --> 00:10:38.840 talking about things like depth first search DFS, breadth first 214 00:10:38.840 --> 00:10:44.120 search BFS, and even more specialized algorithms like Dichstras algorithm, 215 00:10:44.159 --> 00:10:47.399 which is used for finding the shortest paths and weighted graphs. 216 00:10:47.519 --> 00:10:49.000 Yeah, I could see how that would be useful for 217 00:10:49.039 --> 00:10:53.840 things like pathfinding in games, or maybe optimizing roads and 218 00:10:53.879 --> 00:10:57.039 logistics applications. Lots of possibilities there exactly. 219 00:10:57.159 --> 00:11:00.000 And and you know, beyond those those classic algorithms, Racket 220 00:11:00.120 --> 00:11:04.039 lets you implement those more advanced search techniques like a search, 221 00:11:04.159 --> 00:11:06.919 which uses heuristics to you know, to guide the search 222 00:11:06.960 --> 00:11:09.360 process and find solutions even faster. 223 00:11:09.600 --> 00:11:12.720 It's amazing how those algorithms can be applied to so 224 00:11:12.759 --> 00:11:15.759 many different problems. I'm starting to see why why Racket is, 225 00:11:15.799 --> 00:11:18.679 you know, is often praised for its its versatility, its 226 00:11:18.720 --> 00:11:20.200 power as a programming language. 227 00:11:20.240 --> 00:11:23.440 It's true, and and we've only scratched the surface here. 228 00:11:23.519 --> 00:11:26.200 We still have, you know, a whole world to explore 229 00:11:26.360 --> 00:11:29.519 with racklog and logic programming. That's where you get to, uh, 230 00:11:29.759 --> 00:11:31.799 you know, describe what you want to achieve, rather than 231 00:11:31.840 --> 00:11:33.799 just telling the computer how to do it step by step. 232 00:11:33.919 --> 00:11:38.080 It's it's a really fascinating area that really showcases, you know, 233 00:11:38.200 --> 00:11:41.519 just how flexible and expressive Racket can be. 234 00:11:43.360 --> 00:11:47.720 That sounds sounds incredibly intriguing or looking forward to diving 235 00:11:47.720 --> 00:11:49.840 into that. But I think we've we've covered a lot 236 00:11:49.879 --> 00:11:51.879 of ground here in this first part of our our 237 00:11:51.960 --> 00:11:52.840 Racket deep dive. 238 00:11:52.919 --> 00:11:56.240 We certainly have we've journeyed from those foundational concepts of 239 00:11:56.360 --> 00:12:00.679 you know, data and functions all the way to practical 240 00:12:00.679 --> 00:12:05.639 applications like financial analysis and search algorithms. It's really just 241 00:12:05.679 --> 00:12:07.679 a taste of what makes racket such a such a 242 00:12:07.720 --> 00:12:09.039 unique and powerful language. 243 00:12:09.159 --> 00:12:12.120 I'm feeling pretty energized by this exploration. Can't wait to 244 00:12:12.159 --> 00:12:14.480 see what other wonders away us In the next part 245 00:12:14.519 --> 00:12:15.159 of our deep dive. 246 00:12:15.639 --> 00:12:19.039 There's plenty more to uncover. We'll be delving into those 247 00:12:19.200 --> 00:12:22.320 intricacies of logic programming, and then we'll touch on the 248 00:12:22.320 --> 00:12:27.039 elegance of abstract computing machines. Until then, keep those mental 249 00:12:27.039 --> 00:12:27.759 gears turning. 250 00:12:28.279 --> 00:12:31.360 Welcome back to the deep dive. We've been exploring this 251 00:12:31.440 --> 00:12:34.919 world of racket programming, and you've given me some really 252 00:12:35.039 --> 00:12:39.480 fascinating material here. I'm particularly intrigued by this section on 253 00:12:39.639 --> 00:12:44.039 logic programming. With racklog, it almost seems like a completely 254 00:12:44.080 --> 00:12:46.120 different way of thinking about programming. 255 00:12:46.200 --> 00:12:49.159 It is it really is. With racklog, you're basically programming 256 00:12:49.200 --> 00:12:51.639 by defining facts and rules, and then you kind of 257 00:12:51.720 --> 00:12:55.480 let the system infer conclusions from those. It's more of 258 00:12:55.519 --> 00:12:58.000 a declarative style where you focus on the what, not 259 00:12:58.159 --> 00:13:00.399 the how. If that makes sense, it makes it effect 260 00:13:00.399 --> 00:13:05.440 for things like knowledge representation or rule based systems even 261 00:13:05.639 --> 00:13:08.000 solving those those tricky logical puzzles. 262 00:13:08.159 --> 00:13:10.240 Can you could you give me a concrete example here. 263 00:13:10.600 --> 00:13:12.639 I'm having a little trouble wrapping my head around it, 264 00:13:12.679 --> 00:13:13.200 to be honest. 265 00:13:13.519 --> 00:13:16.320 Sure. Sure. Let's say let's say you want to represent 266 00:13:16.360 --> 00:13:22.080 family relationships. So you could define facts like parent John Mary, 267 00:13:22.120 --> 00:13:25.080 meaning John is the parent of Mary, and then you 268 00:13:25.080 --> 00:13:27.679 know parent Mary Alice meaning Mary is the parent of Alice. 269 00:13:27.720 --> 00:13:31.960 And then you could define a rule like grandparent x Z, 270 00:13:32.480 --> 00:13:36.360 parent x Y parent y Z, and and that means 271 00:13:36.399 --> 00:13:39.600 that X is the grandparent of Z if X is 272 00:13:39.639 --> 00:13:41.879 the parent of why, and why is the parent of Z. 273 00:13:42.200 --> 00:13:45.879 So you're you're setting up these logical statements that racklog 274 00:13:45.919 --> 00:13:48.879 can then use to answer questions like who are the 275 00:13:48.879 --> 00:13:49.879 grandparents of Alice? 276 00:13:49.919 --> 00:13:52.879 Exactly? You just asked the question, and racklog uses those 277 00:13:52.879 --> 00:13:57.000 facts and rules to find the answer. It's pretty powerful stuff. 278 00:13:57.080 --> 00:14:00.000 That's fascinating. It's like it's like having a digital detail 279 00:14:00.679 --> 00:14:03.039 that can you know they can solve those logic puzzles 280 00:14:03.080 --> 00:14:04.559 based on the clues you give it. I can you 281 00:14:04.720 --> 00:14:07.080 definitely see how this would be useful for building, you know, 282 00:14:07.600 --> 00:14:10.039 expert systems or those knowledge based applications. 283 00:14:10.200 --> 00:14:12.440 You got it. Racklog really opens up a whole new 284 00:14:12.440 --> 00:14:15.320 world of possibilities, you know, for those problems that are 285 00:14:15.360 --> 00:14:17.639 kind of tailor made for this type of logical deduction. 286 00:14:18.120 --> 00:14:20.120 Before we move on, though, can you can you explain 287 00:14:20.200 --> 00:14:24.320 how how racklog actually does that reasoning, Like what, what 288 00:14:24.360 --> 00:14:26.639 are the key mechanisms involved there? 289 00:14:26.679 --> 00:14:30.399 Sure? So the two the two core mechanisms are unification 290 00:14:30.720 --> 00:14:36.000 and resolution. Unification that's that's about finding values that that 291 00:14:36.080 --> 00:14:38.840 make logical expressions match up. So, for example, to figure 292 00:14:38.879 --> 00:14:43.399 out if parent John X matches parent John, racklog would 293 00:14:43.440 --> 00:14:45.919 try to unify X with Mary basically make them the same. 294 00:14:46.080 --> 00:14:48.360 Oh so it's like finding the missing piece of the 295 00:14:48.399 --> 00:14:49.679 puzzle to complete the picture. 296 00:14:49.799 --> 00:14:51.080 Yeah, that's a good way to put it. And then 297 00:14:51.120 --> 00:14:54.879 you have resolution. Resolution that's about combining different rules and 298 00:14:54.960 --> 00:14:57.679 facts to figure out new facts. So if you have 299 00:14:57.759 --> 00:15:02.000 the fact parent John Mary and parent marry Alice, and 300 00:15:02.000 --> 00:15:04.080 then you have that rule we talked about grandparent x Z, 301 00:15:04.279 --> 00:15:07.639 parent x y parent wise, well resolution, lets raclog put 302 00:15:07.679 --> 00:15:10.559 those together and say, hey, grandparent John Alice must also 303 00:15:10.639 --> 00:15:11.080 be true. 304 00:15:11.519 --> 00:15:14.120 It's like connecting the docks between different pieces of information 305 00:15:14.759 --> 00:15:18.480 to get those new insights. I'm really starting to appreciate 306 00:15:18.519 --> 00:15:22.000 the power the elegance of this approach. It's it's definitely 307 00:15:22.440 --> 00:15:25.399 a different way of thinking about programming, but it seems 308 00:15:25.399 --> 00:15:28.840 incredibly powerful for certain kinds of problems. 309 00:15:28.960 --> 00:15:30.919 It is, and that's that's one of the really great 310 00:15:30.919 --> 00:15:33.720 things about racket. It gives you all these different tools, 311 00:15:33.759 --> 00:15:37.120 these different paradigms to tackle all these different kinds of challenges. 312 00:15:37.120 --> 00:15:38.919 You're not you're not stuck in one way of thinking. 313 00:15:39.559 --> 00:15:43.399 Speaking of different tools, you mentioned abstract computing machines earlier. 314 00:15:43.639 --> 00:15:47.480 That sounds pretty pretty high level. What what exactly are 315 00:15:47.480 --> 00:15:50.919 they and why are they important in this world of racket. 316 00:15:51.120 --> 00:15:55.039 So abstract computing machines, they're basically these these theoretical models 317 00:15:55.080 --> 00:15:59.519 of computation. They help us understand what what computers can 318 00:15:59.600 --> 00:16:02.879 and can do at a really fundamental level. Two big 319 00:16:02.879 --> 00:16:08.159 examples are finite state machines or FSMs, and then churning machines. 320 00:16:08.320 --> 00:16:11.399 Okay, let's start with those finite state machines. What are 321 00:16:11.399 --> 00:16:12.039 those all about? 322 00:16:12.279 --> 00:16:15.519 Imagine a system that can be in like one of 323 00:16:15.639 --> 00:16:19.200 a specific number of states, right, and it receives input 324 00:16:19.399 --> 00:16:22.799 and then changes between these states based on certain rules. 325 00:16:23.200 --> 00:16:25.919 A good example is a vending machine. You could model 326 00:16:25.919 --> 00:16:27.039 that as an FSM. 327 00:16:26.759 --> 00:16:30.759 So you'd have states like idle, waiting for coins, dispensing 328 00:16:30.799 --> 00:16:34.000 product and things like that, and the input like inserting 329 00:16:34.000 --> 00:16:36.799 coins or selecting a product that would trigger those those 330 00:16:36.840 --> 00:16:38.320 state transitions exactly. 331 00:16:38.519 --> 00:16:43.080 FSMs are used all over the place, from parsing text 332 00:16:43.200 --> 00:16:46.720 to even controlling hardware. Racket makes it really easy to 333 00:16:46.759 --> 00:16:49.679 define and work with FSM, so it's a great tool 334 00:16:49.720 --> 00:16:52.039 for kind of playing around with them and seeing what 335 00:16:52.080 --> 00:16:52.519 they can do. 336 00:16:52.919 --> 00:16:55.840 It's cool. And what about those Churing machines. They sound 337 00:16:55.879 --> 00:16:59.440 a bit more intimidating, I have to say. 338 00:16:59.399 --> 00:17:03.120 Turing machine there. They're a much more powerful model of computation. 339 00:17:03.360 --> 00:17:06.839 They've got this, uh, this infinite tape for storage, and 340 00:17:06.880 --> 00:17:09.559 then this head that can read and write symbols on 341 00:17:09.599 --> 00:17:12.200 that tape. They're incredibly powerful, at least in theory. In theory, 342 00:17:12.240 --> 00:17:14.920 they can compute anything that's that's actually computable. 343 00:17:15.000 --> 00:17:17.519 So like the ultimate theoretical computer you could say that. 344 00:17:17.559 --> 00:17:20.200 I mean, we we don't actually build physical tearing machines, 345 00:17:20.240 --> 00:17:23.359 but they're they're this framework for understanding, you know, the 346 00:17:23.400 --> 00:17:26.359 limits of what what computers can do, what problems they 347 00:17:26.359 --> 00:17:29.559 can solve rack. Racket has libraries that let you, uh 348 00:17:29.640 --> 00:17:33.440 let you simulate tearing machines, which is a really fascinating 349 00:17:33.480 --> 00:17:36.559 way to to kind of explore those limits of computation. 350 00:17:37.160 --> 00:17:39.880 It sounds like a real mind bending exercise, but but 351 00:17:39.960 --> 00:17:42.839 I can see how it would be incredibly rewarding for 352 00:17:42.839 --> 00:17:44.920 for someone who really wants to understand, you know, the 353 00:17:44.920 --> 00:17:47.319 foundations of computer science. 354 00:17:47.400 --> 00:17:50.640 It is. It lets you, uh lets you really grapple 355 00:17:50.680 --> 00:17:54.079 with some of those those deep questions about about what 356 00:17:54.119 --> 00:17:56.839 it actually means for something to be to be computable. 357 00:17:56.880 --> 00:17:58.839 It's it's pretty philosophical in a way. 358 00:17:59.119 --> 00:18:02.240 I'm starting to see how these these abstract models of computation, 359 00:18:02.640 --> 00:18:05.640 they can give you a much deeper understanding of what's 360 00:18:05.920 --> 00:18:09.079 what's actually happening when when we program, it's like it's 361 00:18:09.119 --> 00:18:12.400 like looking behind the curtain of computer science exactly. 362 00:18:12.799 --> 00:18:17.119 And Racket, with its focus on being expressive, on exploration, 363 00:18:17.440 --> 00:18:20.160 it really gives you a great platform for venturing into 364 00:18:20.200 --> 00:18:23.279 those those theoretical realms. It's a great tool for learning. 365 00:18:23.799 --> 00:18:26.680 Speaking of practical applications, I'm curious about that section on 366 00:18:27.200 --> 00:18:30.200 building a calculator and Racket. That seems like a great 367 00:18:30.240 --> 00:18:32.119 way to pull together a lot of the concepts we've 368 00:18:32.160 --> 00:18:33.279 we've been talking about so far. 369 00:18:33.440 --> 00:18:36.240 It is building a calculator. It lets you apply things 370 00:18:36.279 --> 00:18:42.160 like input handling, parsing, data structures, functions, even error handling. 371 00:18:42.240 --> 00:18:44.079 It's it's a really good exercise. 372 00:18:44.119 --> 00:18:47.440 So it's like our final examine in racket programming. 373 00:18:47.319 --> 00:18:50.680 You could say that. So the first step is figuring 374 00:18:50.680 --> 00:18:53.839 out how to how to read those mathematical expressions from 375 00:18:53.880 --> 00:18:55.839 the from the user. Then you're going to break them 376 00:18:55.839 --> 00:18:59.599 down into you know, individual pieces, the numbers, the operators, 377 00:18:59.640 --> 00:19:01.559 the parentheses. We call that tokenization. 378 00:19:02.039 --> 00:19:04.240 It's like taking a sentence and breaking it down into 379 00:19:04.279 --> 00:19:08.039 those individual words and punctuation marks. But then once you 380 00:19:08.119 --> 00:19:09.920 have those those pieces, how do you how do you 381 00:19:09.920 --> 00:19:11.920 actually make sense of them and do the math. 382 00:19:12.200 --> 00:19:15.000 That's where person comes in. You take that stream of 383 00:19:15.039 --> 00:19:19.359 tokens and build a tree like structure that represents the 384 00:19:19.680 --> 00:19:23.240 whole mathematical expression. The structure captures, you know, the order 385 00:19:23.279 --> 00:19:25.880 of operations and relationships between all the different parts and 386 00:19:25.960 --> 00:19:27.559 call it an abstract syntax tree. 387 00:19:27.599 --> 00:19:32.039 So you're building this this hierarchical representation of the math problem. 388 00:19:32.599 --> 00:19:35.880 Then what once you have that abstract syntax tree, you 389 00:19:35.920 --> 00:19:39.039 can uh, you can traverse it, basically walk through it 390 00:19:39.119 --> 00:19:41.799 performing the calculations in the in the right order. You 391 00:19:41.880 --> 00:19:45.119 got to handle all the different operators plus minus times divide, 392 00:19:45.519 --> 00:19:47.559 and you've got to deal with those parentheses too to 393 00:19:47.599 --> 00:19:50.599 make sure things get evaluated in the in the correct order. 394 00:19:50.640 --> 00:19:52.160 It's it's more complex than it owns. 395 00:19:52.200 --> 00:19:54.960 It sounds like like a real challenge, but a rewarding 396 00:19:54.960 --> 00:19:57.039 one too. Yeah, what are some of the key things 397 00:19:57.079 --> 00:19:58.880 you got to think about when you're when you're designing 398 00:19:58.960 --> 00:20:00.279 a calculator in in. 399 00:20:00.319 --> 00:20:02.759 Racket Error handling is a big one. You got to 400 00:20:02.799 --> 00:20:07.599 anticipate those potential errors like dividing by zero or or 401 00:20:07.599 --> 00:20:10.359