WEBVTT 00:00:00.100 --> 00:00:04.860 Hello and welcome to Python Bytes, where we deliver Python news and headlines directly to your earbuds. 00:00:04.860 --> 00:00:10.680 This is episode 240, recorded July 1st, 2021, How Time Does Fly. 00:00:10.680 --> 00:00:11.800 I'm Michael Kennedy. 00:00:11.800 --> 00:00:12.940 And I'm Brian Okken. 00:00:12.940 --> 00:00:14.180 And I'm Chris Moffitt. 00:00:14.180 --> 00:00:16.420 Hey, Chris. Welcome to the show. 00:00:16.420 --> 00:00:18.080 Thank you. Great to be here. 00:00:18.080 --> 00:00:19.400 Yeah, it's great to have you here. 00:00:19.400 --> 00:00:28.140 We've had you talking about the missteps of Excel and how the Python data tools chain can make that better over on Talk Python a few times. 00:00:28.140 --> 00:00:30.040 This is your first time on Python Bytes, right? 00:00:30.040 --> 00:00:31.140 It is. Yes. 00:00:31.140 --> 00:00:32.840 Yeah. Exciting to have you here. 00:00:32.840 --> 00:00:33.720 Definitely. Definitely. 00:00:33.720 --> 00:00:35.840 But maybe you want to go ahead and kick us off. 00:00:35.840 --> 00:00:38.420 I want to talk about subclassing today. 00:00:38.420 --> 00:00:41.660 But Hinnick wrote an article called Subclassing in Python. 00:00:41.660 --> 00:00:46.640 And, you know, dealing with classes is just everywhere in Python. 00:00:46.640 --> 00:00:55.100 Even if you're not using classes, Python itself has all sorts of classes and objects that you're using all the time, whether you know it or not. 00:00:55.100 --> 00:01:05.020 But when you start getting into larger design, there is a question around, you know, composition versus inheritance and stuff. 00:01:05.020 --> 00:01:13.560 So I really like this article that Hinnick put together because I think people should think about the ramifications more. 00:01:14.000 --> 00:01:19.840 So the general gist is he prefers composition over inheritance and I do, too. 00:01:19.840 --> 00:01:34.340 And but then goes through if you have to do inheritance when sometimes you do in Python, for instance, the greatest the greatest example I know of is when you're having exception hierarchies. 00:01:34.340 --> 00:01:39.560 And it's really it's really easy to build up exception hierarchies in Python. 00:01:39.560 --> 00:01:45.100 And it's there's like nothing there is except for like the class definitions and their inheritance. 00:01:45.100 --> 00:01:47.060 And that's the easiest class you've ever created. 00:01:47.060 --> 00:01:48.800 Class exception name path. 00:01:48.800 --> 00:01:49.660 Yeah. 00:01:49.660 --> 00:01:52.440 And but it's useful to do that. 00:01:52.480 --> 00:02:05.040 But then if you want to go further, there's other there's other design patterns and stuff, especially from the C world, C++ world where people might be thinking, well, I want to do something similar in Python and stuff. 00:02:05.040 --> 00:02:09.640 And so this is this is actually kind of a really great article to discussion about it. 00:02:09.640 --> 00:02:10.320 It's pretty long. 00:02:10.320 --> 00:02:13.600 I don't want to summarize it too much, but I'll jump into the three types. 00:02:13.600 --> 00:02:19.780 So he talks about three types of subclassing that's often happens subclassing for code sharing. 00:02:19.780 --> 00:02:26.560 And the short answer is it's just people are trying to do the drive principle and try to share code. 00:02:26.560 --> 00:02:29.800 And it's just it ends up being a bad idea, essentially. 00:02:29.800 --> 00:02:32.220 And there's a bunch of references for it. 00:02:32.220 --> 00:02:37.960 And if you don't, I think if you don't believe me or him, read, read this article and read a bunch. 00:02:37.960 --> 00:02:41.180 He's got a whole bunch of linked articles to that discuss it. 00:02:41.180 --> 00:02:43.620 But but I kind of agree. 00:02:43.620 --> 00:02:51.380 The the second type is inherited in is abstract data types or interfaces in a lot of languages are called the interfaces. 00:02:51.380 --> 00:02:54.800 And this is kind of a neat use of it. 00:02:54.800 --> 00:02:56.240 And there's a bunch of things. 00:02:56.240 --> 00:03:00.740 But it's it's and I thought, OK, yeah, you definitely will use. 00:03:01.960 --> 00:03:05.720 It's a little bit of inheritance and composition for for data type stuff. 00:03:05.720 --> 00:03:19.980 But there's in his discussion, he talks about some of the cool things that Python has that allow you to to have these sort of hierarchies without actually doing subclassing. 00:03:19.980 --> 00:03:28.300 So there's there's some cool features of Python, like the protocol syntax that came in recently in in typing dot protocol. 00:03:28.300 --> 00:03:33.280 Yeah, protocols like formal duck typing, which is an odd thing to combine. 00:03:33.280 --> 00:03:33.700 But yes. 00:03:33.700 --> 00:03:34.220 Yeah. 00:03:34.220 --> 00:03:37.920 But it's it's kind of really cool how it's put together in Python. 00:03:37.920 --> 00:03:38.800 So I like that. 00:03:38.800 --> 00:03:41.520 And then lastly is a specialization. 00:03:41.520 --> 00:03:44.180 And that's where kind of the exception hierarchies come in. 00:03:44.180 --> 00:03:49.820 But also he's got a great discussion about structuring data classes that have common elements. 00:03:49.820 --> 00:03:53.340 And and I think that's an interesting discussion, too. 00:03:53.340 --> 00:03:56.340 And I I think I already said this. 00:03:56.340 --> 00:04:05.580 The summary really it's really hard to summarize this article other than it's good to think about your design, especially if you're going to try to bring subclassing into it. 00:04:05.760 --> 00:04:06.820 So let's do that. 00:04:06.820 --> 00:04:07.160 Yeah. 00:04:07.160 --> 00:04:07.440 Awesome. 00:04:07.440 --> 00:04:11.320 I haven't had a chance to dive into this article, but I do want to read it and explore it. 00:04:11.320 --> 00:04:17.160 You know, it touches on a couple of things like it touches on namespaces and modules, which I think is pretty interesting. 00:04:17.160 --> 00:04:25.860 So many people coming from C++, C#, Java, et cetera, like all these really strongly OOP, especially C# and Java, where everything has to be a class. 00:04:25.860 --> 00:04:31.000 They you'll see people creating classes just for things like static variables and so on. 00:04:31.000 --> 00:04:31.440 Right. 00:04:31.440 --> 00:04:32.520 Or static functions. 00:04:32.520 --> 00:04:35.080 If you just have a bunch of static functions, you know what works really well for that? 00:04:35.400 --> 00:04:36.980 A module that has functions in it. 00:04:36.980 --> 00:04:37.160 Right. 00:04:37.160 --> 00:04:38.380 That's the same thing. 00:04:38.380 --> 00:04:39.360 You import module. 00:04:39.360 --> 00:04:45.480 Then you say module dot function name is the same as from module import class, class dot static function name. 00:04:45.480 --> 00:04:45.660 Right. 00:04:45.660 --> 00:04:48.620 Like it's just a layer that doesn't really need to be there. 00:04:48.620 --> 00:04:51.060 So the article touches on that, which I think is neat. 00:04:51.060 --> 00:04:52.860 Like sometimes you just don't need those. 00:04:52.860 --> 00:04:56.740 And then also the composition over inheritance. 00:04:56.740 --> 00:05:04.580 I think composition over inheritance is a really important thing to think about, because so often people say, well, you can't use OOP because it's horrible in all these ways. 00:05:04.580 --> 00:05:07.380 And you end up with like a robotic duck that can't quack. 00:05:07.380 --> 00:05:07.560 Right. 00:05:07.600 --> 00:05:09.600 Like you end up with these weird situations. 00:05:09.600 --> 00:05:14.860 If you like derive too many things and you put a weird specialty on the end, you're like a duck is an animal. 00:05:14.860 --> 00:05:16.940 And then like, but it has wings. 00:05:16.940 --> 00:05:18.080 But wait, now it's a robot. 00:05:18.080 --> 00:05:19.200 Now, why does it eat water? 00:05:19.200 --> 00:05:20.540 You know, it's like what happened to it? 00:05:20.540 --> 00:05:20.700 Right. 00:05:20.960 --> 00:05:27.940 But the composition allows for you to keep things much more tight and small in the inheritance stack, but still put them together in meaningful ways. 00:05:27.940 --> 00:05:29.920 So anyway, yeah, I want to see more about this. 00:05:29.920 --> 00:05:30.520 This looks great. 00:05:30.520 --> 00:05:35.360 And I'm coming from a standpoint of I'm a C++ person as well. 00:05:35.360 --> 00:05:37.300 And I've done both extremes. 00:05:37.300 --> 00:05:45.200 I've like gone way down the inheritance hierarchy thing and had like seven deep in hierarchy, maybe not seven, but like five deep. 00:05:45.200 --> 00:05:47.320 And it gets to be a nightmare. 00:05:47.540 --> 00:05:52.880 So then I got went to the other direction and didn't do any inheritance at all in a design. 00:05:52.880 --> 00:05:54.940 But there's there's problems there, too. 00:05:54.940 --> 00:05:59.640 So thinking about it and doing it smartly is you just need to. 00:05:59.640 --> 00:05:59.860 Yeah. 00:05:59.860 --> 00:06:01.820 So it's often like salt. 00:06:01.820 --> 00:06:03.920 I could see it's really good when you have some. 00:06:03.920 --> 00:06:05.540 You try to go like salt. 00:06:05.540 --> 00:06:05.840 It's great. 00:06:05.840 --> 00:06:06.700 I'm going to have that for dinner. 00:06:06.700 --> 00:06:08.100 Like, no, you shouldn't do that. 00:06:08.100 --> 00:06:17.360 I think the other thing that's really important about this is depending on how long you've been working in Python, sometimes you kind of get stuck in a rut. 00:06:17.360 --> 00:06:19.200 And you're always doing the same thing. 00:06:19.200 --> 00:06:21.920 And the language has evolved and grown over time. 00:06:21.920 --> 00:06:31.560 And so I think articles like this kind of force you to take a step back and see if you're using all the new features in a way that maybe aren't idiomatic. 00:06:31.560 --> 00:06:33.780 Also, quick comment out there from the live stream. 00:06:33.780 --> 00:06:35.680 Paul says, first time watching the live stream. 00:06:35.680 --> 00:06:36.360 Hey, Paul. 00:06:36.360 --> 00:06:38.820 Weird seeing everyone when they say the intro. 00:06:38.820 --> 00:06:40.040 Indeed, it is kind of weird. 00:06:40.200 --> 00:06:42.880 But I want to highlight this one to say hi to Paul. 00:06:42.880 --> 00:06:43.500 Thanks for being here. 00:06:43.500 --> 00:06:47.660 But also, if you're listening, you're like, hey, I'd kind of like to see what's on the screen while you're all talking about this. 00:06:47.660 --> 00:06:48.860 You'll follow us on YouTube. 00:06:48.860 --> 00:06:51.800 There's like a live stream menu right on Pythonbytes.fm. 00:06:51.800 --> 00:06:53.760 So it's easy to sign up for that. 00:06:53.760 --> 00:07:00.680 And also, Sam out there in the live stream following up on that says, currently maintaining a library with a deep templated class and hierarchies. 00:07:00.680 --> 00:07:02.580 It's very hard to keep track of it all. 00:07:03.340 --> 00:07:04.580 Yeah, I hear you that. 00:07:04.580 --> 00:07:05.280 That's for sure. 00:07:05.280 --> 00:07:06.140 All right. 00:07:06.140 --> 00:07:08.640 Let's switch over to the next one. 00:07:08.640 --> 00:07:09.740 Now, I tried. 00:07:09.740 --> 00:07:11.460 I tried to resist this, Brian. 00:07:11.460 --> 00:07:12.780 I promise I did. 00:07:12.780 --> 00:07:16.940 But I've ended up with an extra extra with seven more extras. 00:07:16.940 --> 00:07:18.040 Hear all about it. 00:07:18.040 --> 00:07:21.700 And it just had to become a main item because otherwise we'd be here for hours. 00:07:21.700 --> 00:07:23.040 It's not the idea of the show. 00:07:23.040 --> 00:07:24.840 So we've got an extra extra. 00:07:24.840 --> 00:07:25.840 Hear all about it. 00:07:25.840 --> 00:07:26.940 Nine extras. 00:07:26.940 --> 00:07:28.200 Let's pull them up. 00:07:28.200 --> 00:07:29.180 Action number one. 00:07:29.180 --> 00:07:30.880 We've talked about Pyodide. 00:07:31.000 --> 00:07:34.920 I had a whole talk Python episode on Pyodide, which is an interesting thing. 00:07:34.920 --> 00:07:39.780 It is this project by Mozilla where you take Python and you run it in the browser. 00:07:39.780 --> 00:07:46.820 And then you take many of the data science packages like NumPy and Matplotlib and stuff and compile them into the browser. 00:07:46.820 --> 00:07:51.660 And then you basically have client side Python data science, which is really interesting. 00:07:51.660 --> 00:07:56.160 This project is being spun out as its own topic, as its own project. 00:07:56.160 --> 00:07:57.580 It's no longer under Mozilla. 00:07:57.580 --> 00:07:59.660 Usually that doesn't sound good to me. 00:07:59.740 --> 00:08:01.340 It kind of sounds like it's been orphaned. 00:08:01.340 --> 00:08:04.240 I have no idea what the status of Pyodide is. 00:08:04.240 --> 00:08:05.260 People can check that out. 00:08:05.260 --> 00:08:06.960 But it's no longer under Mozilla. 00:08:06.960 --> 00:08:09.380 It's its own separate thing, as they say. 00:08:09.380 --> 00:08:11.040 So it's cruising out there. 00:08:11.040 --> 00:08:13.420 And also, it didn't get compiled to JavaScript. 00:08:13.420 --> 00:08:17.360 It got compiled to WebAssembly, which is interesting because that's faster. 00:08:17.360 --> 00:08:17.900 All right. 00:08:17.900 --> 00:08:18.540 That's number one. 00:08:18.540 --> 00:08:23.360 Number two, I just, as in a couple hours ago, released a brand new course, Python powered chat 00:08:23.360 --> 00:08:25.300 apps with Wilio and SendGrid. 00:08:25.300 --> 00:08:29.760 So the idea is if you want to have some kind of chat bot, but a lot of that conversation 00:08:29.760 --> 00:08:34.860 has to involve your database and your data and verifying things. 00:08:34.860 --> 00:08:40.380 Like the app that we built here is a tech savvy bakery where you can order cakes by sending 00:08:40.380 --> 00:08:41.380 it a WhatsApp message. 00:08:41.380 --> 00:08:43.320 And then it'll say, hey, you want a cake? 00:08:43.320 --> 00:08:44.500 Well, here's the menu. 00:08:44.500 --> 00:08:46.700 And it actually gets the menu from our Flask app. 00:08:46.940 --> 00:08:48.560 And then they pick something off the menu. 00:08:48.560 --> 00:08:51.640 And once they pick all the details, they said like this. 00:08:51.640 --> 00:08:52.400 Okay, great. 00:08:52.400 --> 00:08:55.840 We send it back to our website and figure out how much that's going to cost. 00:08:55.840 --> 00:08:56.780 They order it. 00:08:56.780 --> 00:08:57.500 It goes back. 00:08:57.500 --> 00:09:02.060 We send them a, if once they accept it, they get like a customized pretty email. 00:09:02.060 --> 00:09:03.540 It goes back to the backend. 00:09:03.540 --> 00:09:04.560 The bakers bake it. 00:09:04.560 --> 00:09:06.200 It sends them another message to let them know. 00:09:06.200 --> 00:09:11.280 So if you want to build kind of like that workflow, if Wilio and SendGrid, check it out. 00:09:11.280 --> 00:09:12.360 This course is super fun. 00:09:12.360 --> 00:09:14.520 It's six hours and it's 100% free. 00:09:14.860 --> 00:09:16.680 So people can check that out. 00:09:16.680 --> 00:09:19.400 I think if you're trying to build that kind of thing, that'll be a lot of fun. 00:09:19.400 --> 00:09:20.880 So links in the show notes there. 00:09:20.880 --> 00:09:21.880 I had something. 00:09:21.880 --> 00:09:22.440 Yes. 00:09:22.440 --> 00:09:24.920 If I can't afford free, can I get a discount code? 00:09:24.920 --> 00:09:28.360 I will give anyone listening 50% off that. 00:09:28.360 --> 00:09:33.580 So I have this really cool tweet and Twitter is broken from what I can tell for everything 00:09:33.580 --> 00:09:34.660 that's not the homepage. 00:09:34.660 --> 00:09:37.200 So something went wrong, but let me describe it. 00:09:37.200 --> 00:09:42.440 So when you look at it in the show notes, you'll be able to see there's a really cool tweet 00:09:42.440 --> 00:09:45.620 from Nick Maul, who was on the guest on the show last week. 00:09:45.620 --> 00:09:46.860 Oh, you got it. 00:09:46.860 --> 00:09:48.080 How can you get this to work? 00:09:48.080 --> 00:09:49.320 You've got some sort of magic. 00:09:49.320 --> 00:09:49.640 All right. 00:09:49.640 --> 00:09:51.520 Well, so thanks for putting on your screen. 00:09:51.520 --> 00:10:00.740 So here we have Will McGuigan showing an animation of basically this really cool, like collapsible 00:10:00.740 --> 00:10:06.620 sidebar and like scrolling within sub windows inside of textual. 00:10:06.760 --> 00:10:06.880 Right. 00:10:06.880 --> 00:10:08.420 We talked about textual as well. 00:10:08.420 --> 00:10:12.520 And it's just such a cool graphic that says like, wow, you can build some pretty amazing 00:10:12.520 --> 00:10:13.660 applications there. 00:10:13.660 --> 00:10:14.140 What do you think? 00:10:14.140 --> 00:10:16.620 Will's just knocking it out of the park with this. 00:10:16.620 --> 00:10:18.880 It's fun to watch him go so fast. 00:10:18.880 --> 00:10:19.520 Absolutely. 00:10:19.520 --> 00:10:20.680 Well done there, Will. 00:10:20.680 --> 00:10:22.500 I'll switch back to mine for a moment. 00:10:22.500 --> 00:10:22.960 Okay. 00:10:22.960 --> 00:10:24.720 Ars Technica works on my computer. 00:10:25.140 --> 00:10:29.960 So remember we did an episode and I titled it something like Flock No or something like 00:10:29.960 --> 00:10:30.300 that. 00:10:30.300 --> 00:10:38.060 So Flock, Federated Learning of Cohorts, is something that Google was trying to do so that they can 00:10:38.060 --> 00:10:40.040 replace third-party cookies. 00:10:40.040 --> 00:10:40.580 Why? 00:10:40.580 --> 00:10:46.840 Because people are running ad blockers or like I am right now, a VPN that at the network 00:10:46.840 --> 00:10:50.520 level blocks all the ad tracking and third-party cookies. 00:10:50.720 --> 00:10:53.900 So they're just basically not working very well anymore. 00:10:53.900 --> 00:10:57.680 So they need to, they're going to cancel third-party cookies from, which means they're 00:10:57.680 --> 00:10:59.580 canceling for the net, the internet. 00:10:59.580 --> 00:11:04.200 And, but because they're Google and they're based on ad revenue primarily, they can't just 00:11:04.200 --> 00:11:05.720 go and we canceled tracking. 00:11:05.720 --> 00:11:06.500 Hooray. 00:11:06.500 --> 00:11:09.500 We're all winning on privacy, right? 00:11:09.500 --> 00:11:14.400 It has to be replaced with some other form of tracking, which they call this Federated Learning 00:11:14.400 --> 00:11:14.860 of Cohorts. 00:11:15.000 --> 00:11:20.140 But the Federated Learning of Cohorts has all these almost more negative consequences. 00:11:20.140 --> 00:11:24.260 And I don't want to go too much into that because we went into quite a lot of detail. 00:11:24.260 --> 00:11:29.400 But for example, you can say, I would like to target, you know, lesbians who just got divorced. 00:11:29.400 --> 00:11:30.780 You run an ad on that. 00:11:30.780 --> 00:11:32.220 People show up on your site. 00:11:32.220 --> 00:11:33.180 They sign up. 00:11:33.180 --> 00:11:33.900 You have an email. 00:11:33.900 --> 00:11:35.240 And now guess what? 00:11:35.240 --> 00:11:38.240 Not only do you know what their email is, you know that they're in this group. 00:11:38.360 --> 00:11:40.720 And maybe this is the very first time you've ever met them, right? 00:11:40.720 --> 00:11:45.020 So really weird, creepy stuff that you could like pull out with this. 00:11:45.020 --> 00:11:49.380 Anyway, the big news is Google delays the rollout till 2023 because you know what? 00:11:49.380 --> 00:11:50.420 People don't like it. 00:11:50.420 --> 00:11:52.400 They're not super keen about it. 00:11:52.400 --> 00:11:56.860 So there's a whole bunch of people who are against that. 00:11:56.860 --> 00:12:00.960 But you're saying they're just delaying it, not stopping it? 00:12:00.960 --> 00:12:02.760 Yes, for now. 00:12:02.760 --> 00:12:06.280 Like, let me, this is a great article that people should check out. 00:12:06.280 --> 00:12:08.420 Like, let me read the first sentence or two. 00:12:08.420 --> 00:12:12.360 Google's plan to up in web advertising and user tracking by dropping third party cookie 00:12:12.360 --> 00:12:13.880 support in Chrome has been delayed. 00:12:13.880 --> 00:12:18.720 Most browsers block third party tracking cookies now as the VPN, like I mentioned. 00:12:18.720 --> 00:12:22.500 But Google, the world's largest advertising company, it wasn't going to follow suit without 00:12:22.500 --> 00:12:24.080 protecting its business model first. 00:12:24.080 --> 00:12:27.180 But there's a lot of challenges with this. 00:12:27.180 --> 00:12:28.920 A lot of people have come out against it. 00:12:28.920 --> 00:12:32.740 And yeah, it's not going to work out super well. 00:12:32.740 --> 00:12:35.520 So they decided to delay it. 00:12:35.640 --> 00:12:36.300 That's what they said. 00:12:36.300 --> 00:12:38.720 Stage two starts mid 2023. 00:12:38.720 --> 00:12:42.280 Google says it's received substantial feedback. 00:12:42.280 --> 00:12:44.040 Including from us. 00:12:44.040 --> 00:12:51.000 And other companies out there are like, we kind of want to keep tracking too, but we're not 00:12:51.000 --> 00:12:52.320 really excited about this. 00:12:52.320 --> 00:12:53.600 So we're just going to not say anything. 00:12:53.600 --> 00:12:56.340 Like Apple, Opera, Mozilla, Microsoft. 00:12:56.340 --> 00:12:58.640 Yeah, they're like, ah, what is I sure about this? 00:12:58.640 --> 00:13:01.020 Anyway, yeah, they've received substantial feedback. 00:13:01.220 --> 00:13:02.740 So hooray, I think for now. 00:13:02.740 --> 00:13:08.140 One thing that we don't talk very often about in Python is what if you want to ship your code 00:13:08.140 --> 00:13:11.600 to somebody and it has sensitive algorithms in it, right? 00:13:12.080 --> 00:13:16.960 It's not that common, but you could get Py2.exe or Py2.app bundle up your code and give it to somebody. 00:13:16.960 --> 00:13:19.880 For example, Dropbox does something to this effect, right? 00:13:19.880 --> 00:13:22.320 Got your Python code running up in your little menu bar. 00:13:22.320 --> 00:13:23.500 There's other ones as well. 00:13:23.500 --> 00:13:26.800 But you might want to encrypt how that works or protect how that works. 00:13:26.800 --> 00:13:30.380 So people can't just open the PY files and look around. 00:13:30.480 --> 00:13:32.060 So there's this thing called Source Defender. 00:13:32.060 --> 00:13:34.800 I'll be clear, this is a paid commercial product. 00:13:34.800 --> 00:13:37.680 I have no affiliation, but they pointed out, they sent me a message. 00:13:37.680 --> 00:13:38.700 Hey, we're doing this thing. 00:13:38.700 --> 00:13:39.200 What do you think? 00:13:39.200 --> 00:13:40.720 It looks kind of interesting. 00:13:40.720 --> 00:13:44.460 I think it's going to be a pretty limited set of people who actually care about this. 00:13:44.460 --> 00:13:47.620 Like if you're running on Docker, you're running on the server, you probably don't care. 00:13:47.620 --> 00:13:49.000 Maybe you do, but probably not. 00:13:49.000 --> 00:13:53.940 But if you'd like to be able to encrypt your source code so it's much harder to see and then 00:13:53.940 --> 00:13:56.900 ship that to somebody, you can use this thing as part of their paid service. 00:13:56.900 --> 00:13:57.740 So that's kind of cool. 00:13:57.740 --> 00:13:58.800 People can check that out. 00:13:59.080 --> 00:14:00.880 Let's see, oh, there's a play noise. 00:14:00.880 --> 00:14:01.320 I don't want that. 00:14:01.320 --> 00:14:05.860 So I was recently interviewed on a day in the life and a work from home Pythonista, which 00:14:05.860 --> 00:14:09.120 is a cool series being done by the folks in the Philippines. 00:14:09.120 --> 00:14:12.920 If you want a tour of the behind the scenes studio and all the work from home stuff, people 00:14:12.920 --> 00:14:13.600 can check that out. 00:14:13.600 --> 00:14:16.400 Python 396 was just released. 00:14:16.400 --> 00:14:19.000 We can check out the change log and see what's happening there. 00:14:19.000 --> 00:14:25.280 There's a security HTTP client about what I think is like a denial of service. 00:14:25.280 --> 00:14:28.880 It sounds like it avoids an infinite loop sort of thing. 00:14:28.880 --> 00:14:30.940 So that might matter to people. 00:14:30.940 --> 00:14:33.040 Probably not, but maybe it does. 00:14:33.040 --> 00:14:38.700 Then a bunch of changes that are happening here, including platform specific ones. 00:14:38.700 --> 00:14:41.360 So if you're running Python 39, and why wouldn't you be? 00:14:41.360 --> 00:14:42.100 Update that. 00:14:42.100 --> 00:14:43.560 Because you're running 310. 00:14:43.560 --> 00:14:44.460 Yes, that's right. 00:14:44.460 --> 00:14:45.440 You're already ahead of the world. 00:14:45.440 --> 00:14:46.600 You're in the future. 00:14:46.660 --> 00:14:54.040 So also, we had Calvin on from six feet up a while ago, and we talked about the conference 00:14:54.040 --> 00:14:55.440 that he was putting together. 00:14:55.440 --> 00:14:59.380 Well, the videos from that conference are out as a YouTube playlist, so people can check 00:14:59.380 --> 00:14:59.700 that out. 00:14:59.700 --> 00:15:01.560 I don't remember how many videos there are. 00:15:01.560 --> 00:15:02.560 Let's click on it and see. 00:15:02.560 --> 00:15:08.300 There are 61 videos, including one on the Python memory deep dive talk that I gave. 00:15:08.300 --> 00:15:10.300 So if you want to check that out, they can. 00:15:10.300 --> 00:15:11.060 Let's see. 00:15:11.060 --> 00:15:12.320 Oh, this one. 00:15:12.320 --> 00:15:13.240 Check this out, Brian. 00:15:13.240 --> 00:15:13.940 Have you seen this? 00:15:14.180 --> 00:15:15.700 Did you know you can pip install Python bytes? 00:15:15.700 --> 00:15:16.340 Yeah. 00:15:16.340 --> 00:15:25.440 You can literally pip install Python bytes because of Scott Stoltzman created this for us as a 00:15:25.440 --> 00:15:25.660 joke. 00:15:25.660 --> 00:15:26.700 He was listening to one of our episodes. 00:15:26.700 --> 00:15:28.340 I can't remember what we talked about. 00:15:28.340 --> 00:15:33.460 This was episode 239, but we must have talked about packaging and pip and things like that. 00:15:33.460 --> 00:15:35.640 So he created a package called Python bytes. 00:15:35.640 --> 00:15:41.140 And what it does is basically you give it a number, like 240, and it would download this 00:15:41.140 --> 00:15:45.060 version as an MP3 file and put it right next to whatever the working directory is. 00:15:45.060 --> 00:15:50.760 So if you want to pip install Python bytes and then python bytes.downloadepisode instead 00:15:50.760 --> 00:15:52.980 of using a podcast player, we're all for that. 00:15:52.980 --> 00:15:53.840 You can check that out. 00:15:53.840 --> 00:15:54.320 Yeah. 00:15:54.320 --> 00:15:54.480 Yeah. 00:15:54.480 --> 00:15:54.920 And that's it. 00:15:54.920 --> 00:15:56.720 That's extra, extra, extra, extra. 00:15:56.720 --> 00:15:57.820 Well, many, many extras. 00:15:57.820 --> 00:15:58.360 Yeah. 00:15:58.440 --> 00:16:03.440 So the Python bytes package is just sort of, it was for fun, but it also, it's really 00:16:03.440 --> 00:16:03.940 small. 00:16:03.940 --> 00:16:09.860 And one of the things I like about it is it's just a really cool example of like with Python, 00:16:09.860 --> 00:16:14.780 you got something that downloads MP3 files off of a feed somewhere. 00:16:14.780 --> 00:16:15.940 It's that easy. 00:16:15.940 --> 00:16:17.360 It's just, that's pretty cool. 00:16:17.360 --> 00:16:18.000 Yeah. 00:16:18.000 --> 00:16:18.620 That's fantastic. 00:16:18.620 --> 00:16:19.100 Absolutely. 00:16:19.100 --> 00:16:20.140 All right. 00:16:20.140 --> 00:16:21.080 Let's see. 00:16:21.080 --> 00:16:22.680 A couple of things from the live stream. 00:16:22.680 --> 00:16:26.820 Sam says, things have happened with Mozilla the last two years that really shook my confidence 00:16:26.820 --> 00:16:27.260 with them. 00:16:27.260 --> 00:16:32.800 I am still a big fan of Firefox and I support their mission, but yeah, it's, I want to see 00:16:32.800 --> 00:16:33.380 them succeed. 00:16:33.380 --> 00:16:34.280 Let's see. 00:16:34.280 --> 00:16:36.000 Another one from the live stream. 00:16:36.000 --> 00:16:38.040 Antonio said, Hey guys, have you mentioned Kivy before? 00:16:38.040 --> 00:16:39.760 Hey, GUIs and Kivy. 00:16:39.760 --> 00:16:40.180 There you go. 00:16:40.180 --> 00:16:42.100 I watched a video about this week. 00:16:42.100 --> 00:16:45.740 It's a GUI that's compatible with many things, including the mobile devices. 00:16:45.740 --> 00:16:46.360 I do. 00:16:46.360 --> 00:16:52.580 My feeling is that Kivy is, is a lot about, it's more about building almost game-like interactions. 00:16:52.580 --> 00:16:56.180 Whereas a lot of GUIs people want, they want like, here's a text box. 00:16:56.260 --> 00:16:57.140 I type in the text box. 00:16:57.140 --> 00:16:58.540 Here's a button I drop in, you know? 00:16:58.540 --> 00:17:00.020 So, but yeah, pretty cool. 00:17:00.020 --> 00:17:01.240 Well, let's see. 00:17:01.240 --> 00:17:06.140 Kim Venwick says, as an aside, shipping a Docker image won't obfuscate the Python. 00:17:06.140 --> 00:17:08.520 The image can be taken apart in files like that. 00:17:08.520 --> 00:17:09.120 That's true. 00:17:09.120 --> 00:17:10.360 They absolutely can. 00:17:10.360 --> 00:17:13.340 I was just thinking like, you're probably just running on like a container service. 00:17:13.340 --> 00:17:15.880 But yeah, if you're shipping it to someone, it's the same. 00:17:15.880 --> 00:17:21.380 Nick Harvey out in the live stream says, could just send the PYC files with no PY. 00:17:21.800 --> 00:17:24.020 It's not foolproof, but it does require more work. 00:17:24.020 --> 00:17:24.320 You're right. 00:17:24.320 --> 00:17:28.640 You'd basically be down to like this dot this and like reading the byte code. 00:17:28.640 --> 00:17:29.920 Yeah, for sure. 00:17:29.920 --> 00:17:31.060 Let's see. 00:17:31.060 --> 00:17:31.900 Final one. 00:17:31.900 --> 00:17:35.820 Rahan says, if it ends up running code on your machine, you can read it. 00:17:35.820 --> 00:17:37.940 It's about putting enough barriers that people won't bother. 00:17:37.940 --> 00:17:39.660 Yeah, that's definitely true. 00:17:39.740 --> 00:17:44.000 I mean, you think of C++ and things like that being completely opaque and yet people take 00:17:44.000 --> 00:17:44.800 that apart all the time. 00:17:44.800 --> 00:17:50.820 But there is also a difference from I'm literally shipping you the source files here to, you know, 00:17:50.820 --> 00:17:53.200 because then you could go in like, oh, here's where the license check is. 00:17:53.200 --> 00:17:55.640 Let's just, you know, command slash comment that out. 00:17:55.640 --> 00:17:56.180 All right. 00:17:56.180 --> 00:17:57.020 Now we're ready to run. 00:17:57.020 --> 00:17:57.440 Yeah. 00:17:57.440 --> 00:17:57.960 Right. 00:17:57.960 --> 00:18:00.460 You want to make it a little bit of a challenge, at least I suspect. 00:18:00.460 --> 00:18:02.600 Anyway, thanks for all the feedback out there, everyone. 00:18:02.600 --> 00:18:06.260 That's the everything extra, extra nine times. 00:18:06.260 --> 00:18:08.460 All right, Chris, what's your first one here? 00:18:08.820 --> 00:18:09.240 All right. 00:18:09.240 --> 00:18:14.980 So the first one is from Andreas Kanz, I think is how you pronounce it. 00:18:14.980 --> 00:18:18.460 And it's a library called Clib, I believe. 00:18:18.460 --> 00:18:21.480 I wasn't sure if it's K-Lib or Clib, but I think it's Clib. 00:18:21.480 --> 00:18:26.060 And it's for automated cleaning of Pandas data frames. 00:18:26.060 --> 00:18:28.940 I guess I should even say it's a little bit more than just cleaning. 00:18:28.940 --> 00:18:30.300 It's automated analysis. 00:18:30.300 --> 00:18:36.080 And, you know, I'll be the first to say I'm a little skeptical about some things that try 00:18:36.080 --> 00:18:37.940 and automate the process. 00:18:37.940 --> 00:18:39.260 But I was playing around with it. 00:18:39.260 --> 00:18:42.240 And there's some pretty cool things that it does. 00:18:42.240 --> 00:18:49.180 The documentation, probably the best way to learn about it is the Towards Data Science article 00:18:49.180 --> 00:18:53.700 that he wrote, which gives a pretty nice overview of what it does. 00:18:53.700 --> 00:19:01.040 It has some, as I mentioned, some pretty nice cleaning features as well as analysis features. 00:19:01.040 --> 00:19:07.440 So I was going to kind of go through a couple of the, describe a couple of things. 00:19:07.660 --> 00:19:15.140 The first one that I thought was really interesting is the, there's this function called data cleaning. 00:19:15.140 --> 00:19:18.580 And it essentially does, you can control what it does. 00:19:18.580 --> 00:19:20.180 So it can clean the column names. 00:19:20.180 --> 00:19:21.620 It can convert data types. 00:19:21.620 --> 00:19:22.540 It can drop missing. 00:19:23.380 --> 00:19:32.980 So one of the things that pandas does is it's not really aggressive about the size or the data types that it uses. 00:19:32.980 --> 00:19:39.900 So when you read in data, it will just kind of assign it maybe to a float or, you know, an object. 00:19:40.580 --> 00:19:42.640 But if you want, you can get in there. 00:19:42.640 --> 00:19:48.820 And if it's a value, if it's a column, let's say that has only values from, you know, less than 100. 00:19:48.820 --> 00:19:51.800 If you convert it to an integer, it saves memory. 00:19:51.800 --> 00:19:55.280 If you save enough memory, then you can actually speed up your code. 00:19:55.280 --> 00:20:04.500 And so this goes behind the scenes and takes your data frame and converts it essentially to the smallest value, NumPy value that it can store. 00:20:05.140 --> 00:20:18.380 And then, you know, I took a random data set and sure enough, it did reduce the memory footprint quite a bit, which I thought was pretty interesting because it's one of those things that is very tedious to do on your own by hand. 00:20:18.380 --> 00:20:20.780 Does it do like if you have the same string? 00:20:20.780 --> 00:20:24.720 Does it just create a pointer to one copy instead of having that many times, stuff like that? 00:20:24.720 --> 00:20:28.440 It can do that by converting it to a category type. 00:20:28.440 --> 00:20:31.220 That's essentially what pandas is doing when you create a category. 00:20:31.220 --> 00:20:36.460 It does that to kind of string to like a list conversion. 00:20:36.460 --> 00:20:38.800 And it's, you know, it's pretty effective. 00:20:38.800 --> 00:20:49.380 And yeah, I've used the category piece before, but I haven't actually gone in and tried to, you know, shorten up the numeric columns, which is really useful. 00:20:49.380 --> 00:20:50.720 The other thing... 00:20:50.720 --> 00:20:52.720 You just convert them to all the integers and then it'll just be shorter. 00:20:52.720 --> 00:20:54.220 So you don't have to worry about the size. 00:20:54.220 --> 00:20:55.120 I'm just teasing. 00:20:55.120 --> 00:20:55.600 Yeah. 00:20:55.600 --> 00:21:04.320 But I mean, it does even do, it's like, it can do even like int 16s or int 32s or int 8. 00:21:04.320 --> 00:21:04.920 Oh yeah, interesting. 00:21:04.920 --> 00:21:08.480 Like it'll shrink to the size that'll like, oh, these are all under 256. 00:21:08.480 --> 00:21:10.020 So we'll go to like one byte. 00:21:10.020 --> 00:21:10.860 Exactly. 00:21:10.860 --> 00:21:11.700 Exactly. 00:21:11.700 --> 00:21:18.840 You know, and I haven't looked at the code to see, you know, how it actually figures it out, but I had a fairly large data frame and it was pretty quick. 00:21:19.640 --> 00:21:22.700 The other one that was interesting is the clean column names. 00:21:22.700 --> 00:21:29.060 So I think there are some other libraries out there that will like strip spaces or special characters from column names. 00:21:29.240 --> 00:21:37.680 But what this one will actually do is actually, if you have a column name that has, let's say camel case, it'll convert it to all underscore. 00:21:37.680 --> 00:21:47.760 Or it will just essentially normalize all of your column names, which, you know, you could have a debate about whether you want to do that. 00:21:47.760 --> 00:21:53.800 But when you have a data frame that has a lot of columns and you're just looking at it the first time, that can really be helpful. 00:21:53.800 --> 00:22:04.440 And then the other function that it does that works pretty well is for cleaning duplicate data or empty data. 00:22:04.440 --> 00:22:15.480 So if you have a lot of columns that have no values in it or just, you know, maybe 90% of the values are empty, you can set thresholds and just clean that all out. 00:22:15.480 --> 00:22:18.640 So I was playing around with it and I was pretty impressed. 00:22:18.640 --> 00:22:30.260 And I kind of wanted to call it out because the documentation right now is mostly around the Jupyter notebooks that he has. 00:22:30.260 --> 00:22:35.980 So I think, you know, it would be nice if we could get some more docs in there and some more examples. 00:22:36.660 --> 00:22:40.100 But overall, I was really impressed with the library. 00:22:40.100 --> 00:22:47.460 And I think people should kind of take a look at it and see if it's something they want to use for some of their own processes. 00:22:47.460 --> 00:22:48.080 Yeah. 00:22:48.080 --> 00:22:52.060 Some of them sound interesting, even if you don't have to trust it, right? 00:22:52.060 --> 00:22:57.760 Like the shrink to the smallest data set, data type, for example, or normalized column names. 00:22:57.760 --> 00:23:01.040 Those don't seem as risky as, you know, clean it up. 00:23:01.040 --> 00:23:02.060 Exactly. 00:23:02.060 --> 00:23:02.940 Find the wrong data. 00:23:03.460 --> 00:23:03.860 Exactly. 00:23:03.860 --> 00:23:08.500 And then I forgot to mention, it also has some nice correlation plots. 00:23:08.500 --> 00:23:18.300 And some of these things you can already do with Seaborn or Matplotlib, but I found that it gives you a little more control and it's just a little bit easier to do it. 00:23:18.300 --> 00:23:22.180 There are certainly other tools out there that do this as well. 00:23:22.960 --> 00:23:35.420 So, oh, and then the categorical data plots I thought was a nice summary of the data and gives you some nice graphs and it helps you understand where you've got some missing values. 00:23:35.420 --> 00:23:36.380 But yeah. 00:23:36.380 --> 00:23:36.600 Yeah. 00:23:36.600 --> 00:23:39.220 The visualizing the missing data is a really interesting feature. 00:23:39.600 --> 00:23:39.780 Yeah. 00:23:39.780 --> 00:23:44.880 And there is another Pandas data frame called MissingNo that does this and does it well. 00:23:44.880 --> 00:23:53.500 But I think this is a unique combination, especially some of the data, the memory saving features that it has are pretty neat. 00:23:53.500 --> 00:23:53.900 Yeah. 00:23:53.900 --> 00:23:58.540 So the cleaning features, though, have a lot of, there's a lot of parameters to it. 00:23:58.580 --> 00:24:00.080 So it looks like you have a lot of control. 00:24:00.080 --> 00:24:05.660 And one of the, I mean, again, this is open source, so it isn't that magical. 00:24:05.660 --> 00:24:07.760 You can just look at the source and see what it's doing. 00:24:07.760 --> 00:24:08.800 So it looks. 00:24:08.800 --> 00:24:09.880 Exactly. 00:24:09.880 --> 00:24:10.200 Yeah. 00:24:10.200 --> 00:24:14.840 And that was one of the things I was looking at is like data cleaning, I think is kind of the top level. 00:24:14.840 --> 00:24:17.260 And you can just run that wide open and it'll do everything. 00:24:17.260 --> 00:24:20.780 And it actually prints out a pretty nice summary of what it does. 00:24:20.780 --> 00:24:28.420 But you can also go in there and specify parameters, like you said, to control it so that maybe it doesn't rename the columns. 00:24:28.420 --> 00:24:30.600 Or drop some of the missing data. 00:24:30.600 --> 00:24:37.140 The other thing that I tried to play with that seemed really interesting is this pool duplicate subsets. 00:24:37.140 --> 00:24:45.480 And essentially what it tries to do, and I had a little bit of trouble with this because I think I put too much data at it. 00:24:45.480 --> 00:24:53.320 But it tries to, maybe if you have 10 columns of data, it says, well, you know what, four or five of them are very heavily correlated. 00:24:53.600 --> 00:24:58.940 So we're going to drop them and just give you the four or five that are actually most useful. 00:24:58.940 --> 00:25:05.520 And so I think that's some interesting tools to use when you get some data that maybe you haven't worked with before. 00:25:05.520 --> 00:25:06.120 Yeah. 00:25:06.120 --> 00:25:06.640 Yeah. 00:25:06.640 --> 00:25:07.140 Very nice. 00:25:07.140 --> 00:25:08.640 What a good find. 00:25:08.640 --> 00:25:09.820 Brian, you got the next one? 00:25:09.820 --> 00:25:10.440 Sure. 00:25:10.440 --> 00:25:10.880 Yeah. 00:25:10.880 --> 00:25:12.020 Just a second. 00:25:12.960 --> 00:25:17.840 So I wanted to remind people to every once in a while, look at functools. 00:25:18.640 --> 00:25:27.180 Because I've experienced functools as kind of an interesting library that's built in. 00:25:27.180 --> 00:25:30.240 That is, it kind of grows with you. 00:25:30.240 --> 00:25:34.140 So if you're new to Python and you look at it, it's going to be confusing. 00:25:34.140 --> 00:25:37.080 It's like all intermediate stuff in there. 00:25:38.680 --> 00:25:48.740 But as you learn and experience more Python programming, come back to it every once in a while because there's stuff in there that you'll use that you didn't think about before. 00:25:48.740 --> 00:25:51.520 So I'm going to go through a few things. 00:25:51.520 --> 00:25:59.720 And actually, I wanted to call out, there was an article by Martin Hines that I read that kind of reminded me to go through and look this. 00:25:59.720 --> 00:26:01.060 So I want to shout out to him. 00:26:01.060 --> 00:26:01.480 Thanks. 00:26:02.520 --> 00:26:04.300 We've talked about some of this stuff before. 00:26:04.300 --> 00:26:14.140 We talked about function overloading and using single dispatch as one of the ways you can do function overloading in Python, which is cool. 00:26:14.140 --> 00:26:15.340 And that's part of functools. 00:26:15.340 --> 00:26:18.660 And hopefully people are familiar with wraps. 00:26:18.660 --> 00:26:24.940 Wraps is a way to create decorators that act like the thing that you decorated. 00:26:24.940 --> 00:26:29.220 And so if you're writing decorators, make sure you check out wraps. 00:26:29.220 --> 00:26:30.860 And then caching as well. 00:26:30.860 --> 00:26:33.780 I think, I'm sure we've talked about LRU cache. 00:26:33.780 --> 00:26:35.040 I'm sure we have, yeah. 00:26:35.040 --> 00:26:35.620 Yeah. 00:26:35.620 --> 00:26:39.140 So that's in functools, the caching. 00:26:39.140 --> 00:26:44.360 And new in 3.9, there's just a simple cache. 00:26:44.360 --> 00:26:45.960 You don't have to say LRU cache. 00:26:45.960 --> 00:26:49.220 And it's just a convenience wrapper around LRU cache. 00:26:49.220 --> 00:26:52.160 But it also, there's no max size. 00:26:52.160 --> 00:26:56.680 So you don't want to do that for things that you actually want to throw items away. 00:26:57.280 --> 00:26:59.220 But caching is super cool. 00:26:59.220 --> 00:27:01.200 So check that out. 00:27:01.200 --> 00:27:01.900 And then I didn't. 00:27:01.900 --> 00:27:05.640 When I first saw the LRU cache, I'm like, whoa, I got to go figure out what this LRU is. 00:27:05.640 --> 00:27:08.320 And it's not like, rather than just like cache the response. 00:27:08.320 --> 00:27:09.320 Yeah. 00:27:09.520 --> 00:27:14.740 I guess the other question, though, you might be is like, well, what if you pass two variable or two arguments or sets of arguments? 00:27:14.740 --> 00:27:15.320 How do those? 00:27:15.320 --> 00:27:15.700 Yeah. 00:27:15.700 --> 00:27:19.260 So either way, it's kind of not 100% totally obvious what's going to happen. 00:27:19.260 --> 00:27:20.040 Yeah, it's very cool. 00:27:20.380 --> 00:27:20.460 Yeah. 00:27:20.460 --> 00:27:24.400 So there's a bunch of caching stuff in there, like the LRU cache. 00:27:24.400 --> 00:27:26.120 But then you can also cache a property. 00:27:26.120 --> 00:27:29.600 And actually, the property one I hadn't used before. 00:27:29.600 --> 00:27:31.840 But I was playing with it this morning. 00:27:31.840 --> 00:27:33.320 And it's really cool. 00:27:33.320 --> 00:27:46.900 So like, for instance, if you've got a data class or any class that has a bunch of stuff, and some of the things, you have an expensive read on one of those because you have to calculate the value. 00:27:46.900 --> 00:27:53.720 You can throw a cache property on it, and it looks pretty cool. 00:27:54.560 --> 00:28:00.620 One of the neat things about it is, so it only reads it once, and then it caches the value of the property. 00:28:00.620 --> 00:28:08.400 And if you need it to refresh, you call delete on it, which is kind of a weird but kind of cool also. 00:28:08.400 --> 00:28:14.640 But it's odd to call delete on something that you want to still be there, and it'll just reread it next time. 00:28:14.640 --> 00:28:15.760 So that's how that works. 00:28:15.760 --> 00:28:16.820 That is weird. 00:28:16.820 --> 00:28:17.880 That's definitely weird. 00:28:17.880 --> 00:28:20.740 Total ordering, I didn't realize, was there. 00:28:21.900 --> 00:28:31.660 So if you have some data type that you want to be able to compare, you can use total ordering to define equal and one other operator. 00:28:31.660 --> 00:28:35.360 And then you get all of the comparison operators to show up. 00:28:35.360 --> 00:28:36.180 You can use that. 00:28:36.180 --> 00:28:50.440 And then the last one I wanted to highlight is partial method, which partial and partial method, which these are kind of neat in that, like, let's say you've got a function that takes a whole bunch of arguments. 00:28:50.440 --> 00:28:56.980 But you want to pre-fill some of those in and create a new function that has some of the arguments pre-filled in. 00:28:56.980 --> 00:29:00.620 You can do that with this, and it's pretty neat. 00:29:00.620 --> 00:29:01.460 Yeah. 00:29:01.460 --> 00:29:02.500 Okay, interesting. 00:29:02.500 --> 00:29:05.440 I see you partially supply some of the arguments, but not all of them. 00:29:05.440 --> 00:29:06.200 Yeah. 00:29:06.200 --> 00:29:12.780 So, yeah, just a shout out to this, that, yeah, these are intermediate or advanced topics. 00:29:12.780 --> 00:29:21.980 But, but there, as, so as you learn more Python, come back to this every once in a while and you, you might use, find it useful. 00:29:21.980 --> 00:29:22.500 Yep. 00:29:22.500 --> 00:29:22.920 Indeed. 00:29:22.920 --> 00:29:26.940 I was like, how did I miss this hashed property thing? 00:29:26.940 --> 00:29:33.680 Like, surely I would have paid attention to that because what, so often these properties that are like computed things, but they, you know, often don't change. 00:29:33.740 --> 00:29:34.120 Right. 00:29:34.120 --> 00:29:35.340 You get something back from the database. 00:29:35.340 --> 00:29:37.500 You want to, it has time sorted in seconds. 00:29:37.500 --> 00:29:38.840 You want to know how many days it is. 00:29:38.840 --> 00:29:40.840 So if something happens, you might have a day's property. 00:29:40.840 --> 00:29:41.060 Right. 00:29:41.060 --> 00:29:41.920 But that's probably not good. 00:29:41.920 --> 00:29:43.680 So having that hashed is cool. 00:29:43.680 --> 00:29:45.280 If you're sure it's not going to change. 00:29:45.280 --> 00:29:46.760 But I'm like, how did I miss it? 00:29:46.760 --> 00:29:47.520 It's new in 3.8. 00:29:47.520 --> 00:29:47.960 Yeah. 00:29:47.960 --> 00:29:49.120 So it's not, it's not super old. 00:29:49.480 --> 00:29:59.060 And like Chris said, one of the reasons to revisit a lot of these things and pay attention to the news on Python is because the language changes like this. 00:29:59.060 --> 00:29:59.820 So, yeah. 00:29:59.820 --> 00:30:00.680 Yeah, for sure. 00:30:00.680 --> 00:30:04.940 Kim out there in live streams says, also worth looking at intertools from time to time. 00:30:04.940 --> 00:30:05.540 Definitely. 00:30:05.540 --> 00:30:06.160 Great. 00:30:06.160 --> 00:30:06.400 Indeed. 00:30:06.400 --> 00:30:09.900 It's in the same level of complexity, but for collection. 00:30:09.900 --> 00:30:13.560 It's kind of like, you wouldn't first go there, but eventually like, oh yeah, this is what I wanted. 00:30:13.560 --> 00:30:14.280 I just didn't know it. 00:30:14.280 --> 00:30:18.940 Speaking of things you didn't know it, let me scare you all a little, make you all delighted. 00:30:19.140 --> 00:30:19.400 I don't know. 00:30:19.400 --> 00:30:20.460 You tell me how you take to this. 00:30:20.460 --> 00:30:22.480 So let me set the stage. 00:30:22.480 --> 00:30:24.600 GitHub has a little bit of source code. 00:30:24.600 --> 00:30:26.840 Much of it actually public, right? 00:30:26.840 --> 00:30:29.440 Like it's public repos and whatnot. 00:30:29.440 --> 00:30:36.000 So it can be analyzed and talked about and shared or used to train an artificial intelligence, which is pretty crazy. 00:30:36.000 --> 00:30:43.320 And if you look at the artificial intelligence around text, there's the GPT-3 stuff, which is like scary, good text-based AI. 00:30:43.320 --> 00:30:48.760 Well, they decided, what if, you know, our parent company also makes this editor? 00:30:48.800 --> 00:30:58.720 What if we hit an AI based on understanding the source code from GitHub, like all the source code from GitHub and put it into VS Code and then it did stuff? 00:30:58.720 --> 00:30:59.660 Have you all seen this? 00:30:59.660 --> 00:31:00.940 It's called GitHub Copilot. 00:31:00.940 --> 00:31:01.620 Yeah. 00:31:01.620 --> 00:31:02.300 Yeah. 00:31:02.300 --> 00:31:02.560 Yeah. 00:31:02.560 --> 00:31:03.680 I haven't done it yet. 00:31:03.680 --> 00:31:06.400 I was going to put the link in there and you beat me to it. 00:31:06.400 --> 00:31:07.120 Oh yeah. 00:31:07.120 --> 00:31:08.400 I was on top of it. 00:31:08.400 --> 00:31:14.540 So if you go over here, there actually works for TypeScript, Go, Ruby, Python, a couple other languages. 00:31:14.540 --> 00:31:18.620 It says it works for many languages, but it's best on those, of course. 00:31:18.620 --> 00:31:27.840 But if you just look like at their homepage, the copilot.github.com, they've got this little animation and it says, I'm going to write a function that says parse expenses. 00:31:27.840 --> 00:31:29.200 And it takes some kind of text. 00:31:29.460 --> 00:31:32.460 And you put a doc string, literally a doc string in Python. 00:31:32.460 --> 00:31:37.400 It says, parse the list of expenses and return the list of tuples, date, value, currency. 00:31:37.400 --> 00:31:39.120 Ignore a line starting with hash. 00:31:39.120 --> 00:31:40.520 Parse using date time. 00:31:40.520 --> 00:31:41.660 Here's some examples. 00:31:41.660 --> 00:31:42.420 Tab. 00:31:42.420 --> 00:31:46.200 And then it writes the code that does that. 00:31:46.560 --> 00:31:48.460 And let's see, what is it going to do? 00:31:48.460 --> 00:31:50.260 It says, it's in the middle of animation. 00:31:50.260 --> 00:31:52.100 It creates a list of expenses. 00:31:52.100 --> 00:31:53.900 It goes through each line on split. 00:31:53.900 --> 00:31:56.820 It says, if the line starts with hash, this is all Python code. 00:31:56.820 --> 00:31:58.500 Continue on your loop. 00:31:58.500 --> 00:32:01.200 Otherwise, date, value, currency equals split it. 00:32:01.200 --> 00:32:03.840 And then it knows how to parse the date one. 00:32:03.840 --> 00:32:07.900 Convert the value to a float and then store the currency as a string. 00:32:07.900 --> 00:32:10.020 And it's not just that sometimes it'll do this. 00:32:10.020 --> 00:32:16.500 You can actually get alternate implementations by tabbing through its recommended solution, which is, 00:32:16.500 --> 00:32:17.260 pretty crazy. 00:32:17.260 --> 00:32:20.820 So this is powered by OpenAIs. 00:32:20.820 --> 00:32:24.280 It's called Codec or something like that. 00:32:24.280 --> 00:32:25.760 I don't see it right here right now. 00:32:25.760 --> 00:32:27.320 Anyway, I'll probably run across it in a second. 00:32:27.320 --> 00:32:28.560 That's what it's powered by. 00:32:28.560 --> 00:32:31.320 And it says things like, you're the pilot. 00:32:31.320 --> 00:32:33.700 So with GitHub Copilot, you're always in charge. 00:32:33.700 --> 00:32:40.700 You can cycle through alternative suggestions and choose which to accept or reject and then manually edit the suggested code. 00:32:40.700 --> 00:32:41.300 Oh, yeah. 00:32:41.300 --> 00:32:42.180 And it learns from you. 00:32:42.180 --> 00:32:44.080 So I don't know. 00:32:44.080 --> 00:32:46.000 This is wild, y'all. 00:32:46.100 --> 00:32:47.380 This is pretty wild stuff here. 00:32:47.380 --> 00:32:48.100 What do you think? 00:32:48.100 --> 00:32:51.720 I think it's really impressive. 00:32:51.720 --> 00:32:55.940 I mean, it will be interesting to see what it's like when you use it in real life. 00:32:55.940 --> 00:32:59.740 And I think that there could certainly be limitations. 00:32:59.740 --> 00:33:02.320 But I don't know about you. 00:33:02.320 --> 00:33:10.020 But whenever I'm programming, there's always these things I just need to go and look at the documentation or look at Stack Overflow to refresh my memory. 00:33:10.020 --> 00:33:11.300 You've got to connect to SQLAlchemy. 00:33:11.300 --> 00:33:15.120 And I totally forgot how to do those three steps for that connection string sequence. 00:33:15.120 --> 00:33:15.500 Right? 00:33:15.500 --> 00:33:16.240 Exactly. 00:33:16.240 --> 00:33:16.920 Yeah. 00:33:17.280 --> 00:33:27.120 And I've seen, I saw on Twitter where someone was throwing a little shade at that example that you're walking through because they said, well, why are you storing the currency as a float? 00:33:27.120 --> 00:33:31.960 It should be a decimal because if you store a currency as a float, you're going to have all the rounding issues. 00:33:31.960 --> 00:33:33.340 So... 00:33:33.340 --> 00:33:36.060 That's how Superman makes all his money or the evil villain in Superman. 00:33:36.060 --> 00:33:36.580 What was it? 00:33:36.580 --> 00:33:37.400 One of those shows. 00:33:37.400 --> 00:33:37.880 Yeah. 00:33:37.880 --> 00:33:41.480 Richard Pryor in one of the original Superman. 00:33:41.480 --> 00:33:41.880 Yeah. 00:33:41.880 --> 00:33:44.300 Yeah. 00:33:44.300 --> 00:33:45.960 And it's not just based on the doc string. 00:33:45.960 --> 00:33:51.660 Like the example I first spoke about was you wrote complex doc string and then say, do that thing. 00:33:51.660 --> 00:33:55.220 But you can do it based just on function name. 00:33:55.220 --> 00:33:57.620 You can just type a meaningful function name. 00:33:57.620 --> 00:34:00.300 What was the example they used? 00:34:00.300 --> 00:34:00.880 I can't remember. 00:34:00.880 --> 00:34:09.000 But yeah, so you basically just write a doc string, a comment, a function name, or even some code to give more context to it. 00:34:09.000 --> 00:34:10.160 And then off it goes. 00:34:10.160 --> 00:34:11.560 So yeah, pretty neat. 00:34:11.560 --> 00:34:14.660 Codex, that's the name of the AI system behind it. 00:34:14.660 --> 00:34:19.440 So basically, this is a plugin for VS Code, but a really nice one. 00:34:19.440 --> 00:34:21.240 So here's some examples we'll all be familiar with. 00:34:21.240 --> 00:34:22.240 So fetch tweets. 00:34:22.240 --> 00:34:29.740 And the example here is you literally write def fetch underscore tweets underscore from underscore user have. 00:34:29.740 --> 00:34:33.260 And then what it auto completes with is, oh, you're going to need to pass the username in. 00:34:33.260 --> 00:34:37.400 And then here's how you authorize with Tweepy, set up the API credentials. 00:34:37.400 --> 00:34:39.160 And then here's the code you write. 00:34:39.160 --> 00:34:39.680 Oh, yeah. 00:34:39.680 --> 00:34:40.920 And here's your return. 00:34:40.920 --> 00:34:42.500 Or I want to do a scatter plot. 00:34:42.660 --> 00:34:48.000 And you write import, import matplotlib dot pyplot as plot, draw scatter plot, have. 00:34:48.000 --> 00:34:49.460 And then boom, there it is. 00:34:49.460 --> 00:34:50.600 Or memoization. 00:34:50.600 --> 00:34:53.140 I wanted to point this one out because of what you're covering, Brian. 00:34:53.140 --> 00:35:00.280 It says, oh, here's how you memoize a function, which is to, if it's past a set of arguments, it's always going to return the same answer. 00:35:00.280 --> 00:35:01.660 So just give that answer. 00:35:01.760 --> 00:35:04.540 Like, remember, these arguments equal this return value once it's run. 00:35:04.620 --> 00:35:11.320 And it shows how to create a complex decorator that is going to have a function that remembers the values using caching. 00:35:11.320 --> 00:35:14.720 It could just go at funktools dot cache. 00:35:14.720 --> 00:35:15.700 You know what I mean? 00:35:15.700 --> 00:35:18.100 Like, so there's things like that that is missing, right? 00:35:18.100 --> 00:35:23.880 Because you could achieve the exact same outcome with funktools cache decorators, right? 00:35:23.880 --> 00:35:26.100 Instead of trying to write a bunch of code that reimplements that. 00:35:26.260 --> 00:35:28.680 But anyway, pretty, pretty wild thing. 00:35:28.680 --> 00:35:29.900 I don't know really how to feel about that. 00:35:29.900 --> 00:35:30.740 I've been thinking about this today. 00:35:30.740 --> 00:35:32.860 It's kind of freaking me out, but it's also kind of cool. 00:35:32.860 --> 00:35:33.460 Yeah. 00:35:33.460 --> 00:35:45.780 I wanted to point out a comic that people have been pointing out with relation to this is, is the, you know, I wish we could have just specify what we wanted to do, the computer to do. 00:35:45.780 --> 00:35:47.620 And, and it just does it. 00:35:47.620 --> 00:35:49.180 And we already have that. 00:35:49.180 --> 00:35:50.000 It's called code. 00:35:50.000 --> 00:35:50.920 So anyway. 00:35:50.920 --> 00:35:52.320 Yeah. 00:35:52.320 --> 00:35:56.040 I, I don't think this, you know, people often say things like. 00:35:56.040 --> 00:35:58.200 I remember hearing this 20 years ago. 00:35:58.200 --> 00:36:02.980 Oh, this low code thing where you create these little boxes that do stuff and you drag and drop between them. 00:36:02.980 --> 00:36:04.700 We're not going to need programmers anymore. 00:36:04.700 --> 00:36:06.720 We're all just going to become drag or droppies. 00:36:06.720 --> 00:36:08.660 And then like you programmers won't be needed. 00:36:08.660 --> 00:36:11.640 The business people will just drag you drop either way the future. 00:36:11.640 --> 00:36:13.860 And that never, ever happened. 00:36:13.860 --> 00:36:14.160 Right. 00:36:14.160 --> 00:36:15.940 Because people got to put them in production. 00:36:15.940 --> 00:36:17.000 They've got to debug them. 00:36:17.000 --> 00:36:18.140 They've got to scale them. 00:36:18.140 --> 00:36:19.020 And so on. 00:36:19.020 --> 00:36:20.200 Yeah. 00:36:20.200 --> 00:36:20.720 Yeah. 00:36:20.720 --> 00:36:21.520 I think the same thing here. 00:36:21.520 --> 00:36:22.300 Like sure. 00:36:22.300 --> 00:36:25.960 It wrote it once, but you can't have a write only experience. 00:36:25.960 --> 00:36:26.880 For your code. 00:36:26.880 --> 00:36:30.220 You have to understand your code and be able to evolve your code and work with. 00:36:30.220 --> 00:36:33.880 So this might power you into a solution faster. 00:36:33.880 --> 00:36:37.520 But I don't think it escapes the need of people doing meaningful software work. 00:36:37.520 --> 00:36:39.220 The person that pointed out. 00:36:39.220 --> 00:36:43.140 And several people pointed out the example of using money. 00:36:43.140 --> 00:36:44.320 Floats and money. 00:36:45.080 --> 00:36:49.120 That does highlight one of the problems with something like this, though, that everybody 00:36:49.120 --> 00:36:51.940 needs to be careful of is the code that's generated. 00:36:51.940 --> 00:36:57.940 Now you have to like, you were already creating, carefully thinking about it when you were creating 00:36:57.940 --> 00:36:58.140 it. 00:36:58.140 --> 00:37:02.180 But if something else creates it, you've got to scrutinize that to make sure that's really 00:37:02.180 --> 00:37:03.280 doing the right thing. 00:37:03.280 --> 00:37:09.420 And so you're code reviewing some AI code while you're coding your own stuff. 00:37:09.420 --> 00:37:11.160 That's just a different part of your brain. 00:37:11.240 --> 00:37:13.060 You've got to make sure that you're really paying attention. 00:37:13.060 --> 00:37:13.600 Yeah. 00:37:13.600 --> 00:37:13.940 Yeah. 00:37:13.940 --> 00:37:17.280 And even I was looking at that matplotlib example. 00:37:17.280 --> 00:37:21.660 And I would even argue that's not really the way you should do a scatterplot in matplotlib. 00:37:21.660 --> 00:37:26.200 Because you should use the object-oriented interface in matplotlib. 00:37:26.200 --> 00:37:27.740 I mean, the code will work. 00:37:27.740 --> 00:37:30.680 But I wouldn't advocate that you use that code. 00:37:31.100 --> 00:37:37.780 And so to your point, I think it will be interesting to see if it does learn on your own coding style. 00:37:37.780 --> 00:37:42.820 So does it start to recognize those things that you're always, you know, like you said, connecting 00:37:42.820 --> 00:37:47.580 to a database or fetching a file or doing a certain pandas function? 00:37:47.580 --> 00:37:49.820 Will it start to learn that? 00:37:49.820 --> 00:37:51.700 And then that would be really interesting. 00:37:51.700 --> 00:37:55.200 I think there's something about it adapting to you and it learning from what you're doing. 00:37:55.200 --> 00:37:57.040 But I have no idea what that actually means. 00:37:57.040 --> 00:37:57.580 Yeah. 00:37:57.580 --> 00:37:58.660 Hopefully it's paying attention. 00:37:58.660 --> 00:38:03.980 So if it generates something and you change it to the different method and everybody else 00:38:03.980 --> 00:38:08.940 is doing that also, maybe they'll stop suggesting the old one and start suggesting the new one. 00:38:09.200 --> 00:38:09.380 Yeah. 00:38:09.380 --> 00:38:13.780 I'll, you know, Chris, your point about having to, maybe it was you, Brian, sorry. 00:38:13.780 --> 00:38:17.200 Whoever said about, you've got to like criticize this and you didn't write it. 00:38:17.200 --> 00:38:20.940 So you basically have to study it and then, then understand or understand it and study it 00:38:20.940 --> 00:38:22.680 to make sure it's doing the right thing. 00:38:22.680 --> 00:38:27.700 You know, I, a couple of years ago, I don't know, a while ago, I was river floating and 00:38:27.700 --> 00:38:31.320 broke my hand on some rocks, broke my finger in a bunch of places. 00:38:31.320 --> 00:38:35.240 And like my fingers were completely wrapped up all the way to the very tips. 00:38:35.240 --> 00:38:38.360 There was no like, oh, little pecking typing while my hand healed. 00:38:38.460 --> 00:38:41.820 It was like, nope, no one handed, really slow. 00:38:41.820 --> 00:38:47.660 So to keep things going, I used voice to text to try to like, at least keep email flowing 00:38:47.660 --> 00:38:49.060 for a month or something, you know? 00:38:49.060 --> 00:38:52.480 And what I found was I could write pretty decent emails. 00:38:52.480 --> 00:38:56.040 It's hard to like stop and think in whole sentences the way the little tools like it to 00:38:56.040 --> 00:38:57.960 work, but you can get it to work pretty well. 00:38:57.960 --> 00:39:03.760 But the mistakes it makes that are phonetically correct, but actually what you mean wrong, like 00:39:03.760 --> 00:39:08.440 they and they, or, or something that sounds like what you said, but is actually not 00:39:08.440 --> 00:39:10.920 what you mean to say is incredibly hard. 00:39:10.920 --> 00:39:14.100 It's much harder to understand and edit than you would think. 00:39:14.100 --> 00:39:18.020 And so things like this, like, well, I wanted it to do that and I hit tab and I, okay, it's 00:39:18.020 --> 00:39:20.820 doing, I feel like there's going to be a lot of blind spots. 00:39:20.820 --> 00:39:21.400 Yeah. 00:39:21.400 --> 00:39:22.320 Well, it did what it says. 00:39:22.320 --> 00:39:22.660 It didn't. 00:39:22.660 --> 00:39:24.380 I typed the thing and it seems right. 00:39:24.380 --> 00:39:26.260 And like, how do you really, really know? 00:39:26.320 --> 00:39:31.400 I, it just seems like in the same type of situation, it's going to be harder than normal code to 00:39:31.400 --> 00:39:31.640 check. 00:39:31.640 --> 00:39:34.020 Cause you didn't have to think through it to create it. 00:39:34.020 --> 00:39:34.440 You know? 00:39:34.440 --> 00:39:35.300 Yeah. 00:39:35.300 --> 00:39:36.600 A couple of comments from the live stream. 00:39:36.900 --> 00:39:39.360 Ray Han, don't, don't give him ideas. 00:39:39.360 --> 00:39:42.780 It says, greeting Dr. Falcon. 00:39:42.780 --> 00:39:45.200 Gosh, where have you been? 00:39:45.200 --> 00:39:48.540 Let's play thermonuclear war as a dock string. 00:39:49.380 --> 00:39:54.120 And Nick says, I can't help but think of Microsoft pay, which Microsoft pay was this really cool 00:39:54.120 --> 00:39:58.040 bot that was like super good at adapting this stuff and they put it on Twitter, but people 00:39:58.040 --> 00:40:01.560 decided to like be mean to, to, to it instead of teach it. 00:40:01.560 --> 00:40:07.980 I think in like Japanese Twitter, it became a very kind and intelligent bot, but on like English 00:40:07.980 --> 00:40:12.820 Twitter, it got turned into like a racist, horrible creature like right away. 00:40:12.820 --> 00:40:14.280 And they actually had to cancel the project. 00:40:14.280 --> 00:40:14.840 So yeah. 00:40:16.360 --> 00:40:21.860 And then Arthur says, next April fool, April, April fool's day prank. 00:40:21.860 --> 00:40:24.380 Everyone start writing terrible code that influences the AI. 00:40:24.380 --> 00:40:27.560 And this, this is why English day went down the tubes. 00:40:27.560 --> 00:40:30.660 Let's see. 00:40:30.660 --> 00:40:34.400 And then Sam, for goodness sakes, don't trade it on GitHub code. 00:40:34.400 --> 00:40:37.220 It'll arbitrarily turn on debug mode. 00:40:37.220 --> 00:40:37.960 Yeah, perhaps. 00:40:37.960 --> 00:40:39.220 Yeah. 00:40:39.220 --> 00:40:42.380 Kim thinks this is both very impressive and vaguely unsettling. 00:40:42.380 --> 00:40:44.320 And that captures what I was thinking. 00:40:44.980 --> 00:40:47.760 Rahen, will it go and talk to the marketing people for me? 00:40:47.760 --> 00:40:48.620 No. 00:40:48.620 --> 00:40:50.600 I'm good with people. 00:40:50.600 --> 00:40:51.440 That's what I do. 00:40:51.440 --> 00:40:52.180 Yeah. 00:40:52.180 --> 00:40:52.520 Okay. 00:40:52.520 --> 00:41:00.060 Another thing that's not mentioned here explicitly, but I think is interesting is this code is coming 00:41:00.060 --> 00:41:00.560 from GitHub. 00:41:00.560 --> 00:41:01.100 Yeah. 00:41:01.100 --> 00:41:08.280 When I go and I'm saying like, I'm working on super secret commercial project for large organization 00:41:08.280 --> 00:41:10.820 that has lots of people trying to scrutinize it. 00:41:10.820 --> 00:41:13.120 And I hit memoize tab. 00:41:13.120 --> 00:41:14.800 It's going to write some amazing code. 00:41:14.800 --> 00:41:16.420 Oh, by the way, was that GPL? 00:41:16.420 --> 00:41:17.820 Where did that code come from? 00:41:17.820 --> 00:41:18.180 Right? 00:41:18.180 --> 00:41:20.820 Like what's the license of the code that was on GitHub? 00:41:20.820 --> 00:41:25.980 Did I just now all of a sudden grab something that turned, you know, like if I was doing this 00:41:25.980 --> 00:41:28.640 on Windows and I hit tab is Windows now open source. 00:41:28.640 --> 00:41:29.460 I don't know. 00:41:29.800 --> 00:41:30.940 That's a really interesting point. 00:41:30.940 --> 00:41:34.960 And you would think if it was a small startup, someone will probably sue them. 00:41:34.960 --> 00:41:37.420 But, you know, this is Microsoft now. 00:41:37.420 --> 00:41:37.680 Yeah. 00:41:37.680 --> 00:41:38.560 Exactly. 00:41:38.560 --> 00:41:38.960 Yeah. 00:41:38.960 --> 00:41:39.420 Yeah. 00:41:39.420 --> 00:41:42.100 Anyway, so I agree with Kim. 00:41:42.100 --> 00:41:43.100 This is both very impressive. 00:41:43.660 --> 00:41:45.860 If this is the start, like where will it go? 00:41:45.860 --> 00:41:46.680 It would be very amazing. 00:41:46.680 --> 00:41:49.560 But it's also vaguely unsettling at the same time. 00:41:49.560 --> 00:41:51.160 And I don't know how I feel about it. 00:41:51.160 --> 00:41:53.940 Other than I wish it was in PyCharm so I could play with it more often. 00:41:53.940 --> 00:41:56.300 All right. 00:41:56.300 --> 00:41:57.440 Chris, you got the last one? 00:41:57.440 --> 00:41:58.300 I do. 00:41:58.300 --> 00:42:02.680 So this is another library called Cats. 00:42:02.680 --> 00:42:06.760 And it's a time series analysis library. 00:42:06.760 --> 00:42:08.640 And it's made by the same. 00:42:08.640 --> 00:42:09.780 Well, it's from Facebook. 00:42:10.540 --> 00:42:16.280 And a lot of people may have heard of Profit, which is a library for time series forecasting. 00:42:16.280 --> 00:42:24.060 And one of the things that's interesting to me about Profit and Cats is I think time series 00:42:24.060 --> 00:42:27.540 forecasting is something that's really common in the business world. 00:42:27.540 --> 00:42:34.040 I mean, you think about trying to forecast sales or maybe inventory movements or stock prices, 00:42:34.040 --> 00:42:37.220 a whole bunch of different use cases for it. 00:42:37.220 --> 00:42:45.560 And I think in general, most organizations don't have a group of PhDs that are really sophisticated 00:42:45.560 --> 00:42:46.680 in their analysis. 00:42:46.680 --> 00:42:50.340 So people use Excel and kind of come up with their own approaches. 00:42:50.340 --> 00:42:53.080 And that's why I thought Profit was interesting. 00:42:53.080 --> 00:42:57.300 And I think this is interesting because it does come from Facebook. 00:42:57.300 --> 00:43:01.840 And you have to assume that they've got a lot of smart people that are doing a lot of forecasting. 00:43:02.220 --> 00:43:09.340 And they've taken some of the things that Profit was good at and added some additional tools. 00:43:09.340 --> 00:43:16.580 So before I go too much into Cats, one thing I wanted to mention is I did write an article 00:43:16.580 --> 00:43:17.360 about Profit. 00:43:17.360 --> 00:43:27.400 But I think other people, this gentleman, Peter Cotton, wrote an article about Profit and essentially 00:43:27.400 --> 00:43:29.180 questioning how good it was. 00:43:29.180 --> 00:43:34.240 And this is a really long, really well thought out article. 00:43:34.240 --> 00:43:37.500 And some of the math and some of the concepts are way over my head. 00:43:37.500 --> 00:43:41.980 But I do encourage people, if you're looking at time series forecasting, take a look at this. 00:43:41.980 --> 00:43:49.260 But what Cats does is instead of just doing forecasting with Profit, it has a couple of different models 00:43:49.260 --> 00:43:50.260 that you can use. 00:43:50.260 --> 00:43:58.480 You can also do some more just basic time series analysis with it to detect seasonality patterns 00:43:58.480 --> 00:44:00.500 and change points and other trends. 00:44:00.500 --> 00:44:06.020 There's also, if you want to incorporate this in some of your other machine learning algorithms 00:44:06.020 --> 00:44:11.140 to pull out features from your time series data, you can do that with this library as well. 00:44:11.500 --> 00:44:18.360 And there's a whole bunch of other libraries or utilities to build like ensemble models and 00:44:18.360 --> 00:44:20.740 other approaches for time series forecasting. 00:44:20.740 --> 00:44:24.800 This is another one where it is relatively new. 00:44:24.800 --> 00:44:31.140 So there's not a whole lot of documentation, but it's a whole bunch of different Python notebooks, 00:44:31.140 --> 00:44:32.580 Jupyter notebooks, I mean. 00:44:33.120 --> 00:44:38.280 And like one of the things I think is interesting is from a forecasting perspective, you can use Profit, 00:44:38.280 --> 00:44:46.280 but use the same API and use Cerema, I think, Cerema and Holt Winners, as well as some other 00:44:46.280 --> 00:44:47.400 ensemble models. 00:44:47.400 --> 00:44:50.440 You can backtest, you can tune your hyperparameters. 00:44:50.440 --> 00:44:57.760 And then you can also, it's got several of these other algorithms for change point detection. 00:44:58.400 --> 00:45:05.440 And, you know, a lot of this, like I said, is I'm not an expert on the math, but I am interested in how you 00:45:05.440 --> 00:45:10.540 figure out how to take these tools and apply them to those real world business problems. 00:45:10.540 --> 00:45:15.840 And so I think it's really great when we have some of these libraries out there that are developed by 00:45:15.840 --> 00:45:20.500 really smart people that do understand the state of the art, that can maybe make it a little simpler 00:45:20.500 --> 00:45:23.020 for others to apply to their own unique challenges. 00:45:23.020 --> 00:45:25.480 Yeah, this looks really nice to bundle these all together. 00:45:25.840 --> 00:45:29.280 What's a type of problem you might answer with this? 00:45:29.280 --> 00:45:30.320 I think so. 00:45:30.320 --> 00:45:39.040 One example could be help me figure out what my blog or my website traffic is going to look like in six months 00:45:39.040 --> 00:45:39.480 from now. 00:45:39.480 --> 00:45:48.300 So I need to figure out, do I need to, you know, resize my servers or upgrade my disk space or... 00:45:48.300 --> 00:45:50.400 What's my AWS bandwidth bill going to be? 00:45:50.780 --> 00:45:51.300 Exactly. 00:45:51.300 --> 00:45:57.100 You know, the other one that I think it's probably used a lot in inventory. 00:45:57.100 --> 00:46:00.780 So trying to figure out, okay, what do I think sales is going to look like? 00:46:00.780 --> 00:46:07.060 What do I need to reorder so that I actually have enough product so that we don't stock out? 00:46:07.060 --> 00:46:11.000 I think those are some pretty common use cases. 00:46:11.260 --> 00:46:15.340 A lot of the examples here are the airline flight data. 00:46:15.340 --> 00:46:22.320 So anything that you have that's over a period of time, typically kind of on a daily basis over multiple years, 00:46:22.320 --> 00:46:27.620 you can then start to forecast out what the future, what those future numbers would look like. 00:46:27.620 --> 00:46:31.240 Then you have this magic prediction power for the executives. 00:46:31.680 --> 00:46:32.120 Exactly. 00:46:32.120 --> 00:46:38.740 And I think what's interesting about it is most of these, I think most times when people do prediction in Excel, 00:46:38.740 --> 00:46:43.140 it's kind of you put the numbers in there and kind of do your linear line. 00:46:43.140 --> 00:46:48.360 But these tend to give you more error bars so you can give a range. 00:46:48.360 --> 00:46:58.180 So I think a prediction like this is much more valuable when you say it could be between, you know, 100 and 110 versus it's going to be 101.5. 00:46:58.180 --> 00:47:02.900 And when you do that, it conveys a lot more precision than is really there. 00:47:02.900 --> 00:47:04.560 Yeah, that makes a lot of sense. 00:47:04.560 --> 00:47:06.860 Comment from the live stream, Sam Morley says, 00:47:06.860 --> 00:47:14.280 when I was experimenting with time series data, I managed to get better results with a fairly simple, naive Dharma model than I did using profit. 00:47:14.280 --> 00:47:23.400 And I think that's exactly what this article, I don't know if he's read this article, but this, the article from Dr. Cotton, 00:47:23.400 --> 00:47:29.160 that's essentially what he says is some of the more simple models did outperform profit. 00:47:29.160 --> 00:47:30.180 Yeah. Interesting. 00:47:30.180 --> 00:47:30.820 Well, cool. 00:47:30.820 --> 00:47:31.360 All right. 00:47:31.360 --> 00:47:32.820 Brian, is that it for all of our items? 00:47:32.820 --> 00:47:33.680 It is. 00:47:33.680 --> 00:47:35.600 Got any extra stuff you want to throw out there? 00:47:35.600 --> 00:47:38.500 Oh, I just had a, just a quick one. 00:47:38.500 --> 00:47:43.380 Somebody on Twitter last week asked, why did I write a second edition of the book? 00:47:43.380 --> 00:47:46.340 And so I thought, well, that's a reasonable question. 00:47:46.340 --> 00:47:51.500 So at pytestBook.com, you can go and I've added a, why a second edition section. 00:47:51.500 --> 00:47:53.000 So I think I'll read that. 00:47:53.000 --> 00:48:02.340 New built-in fixtures, new flags, package scope features, f-strings, types, all sorts of good things are available that weren't available then, right? 00:48:02.340 --> 00:48:02.840 Yep. 00:48:02.840 --> 00:48:03.320 Yeah. 00:48:03.320 --> 00:48:05.100 There's all sorts of reasons. 00:48:05.100 --> 00:48:07.220 Always good to see Pathlib there. 00:48:07.220 --> 00:48:08.220 I love Pathlib. 00:48:08.220 --> 00:48:10.760 It makes my life so much easier when dealing with files. 00:48:10.760 --> 00:48:11.480 Yeah. 00:48:11.480 --> 00:48:13.220 I finally made the move. 00:48:13.220 --> 00:48:16.980 I've put down os.path and I'm now all about the Pathlib. 00:48:16.980 --> 00:48:17.540 Loving it. 00:48:17.540 --> 00:48:18.160 That's good. 00:48:18.160 --> 00:48:18.660 Yeah. 00:48:18.660 --> 00:48:20.500 Chris, anything else you want to throw out there? 00:48:20.500 --> 00:48:22.420 I was going to throw out one other. 00:48:22.420 --> 00:48:32.200 I was doing some research for working with units of measure. 00:48:32.200 --> 00:48:32.680 UNYT. 00:48:32.680 --> 00:48:32.680 UNYT. 00:48:32.680 --> 00:48:33.880 UNYT. 00:48:33.880 --> 00:48:34.200 UNYT. 00:48:34.200 --> 00:48:35.520 UNYT. 00:48:35.520 --> 00:48:35.960 UNYT. 00:48:35.960 --> 00:48:40.480 That allows you to do things like convert kilometers to miles. 00:48:41.480 --> 00:48:48.100 it works with numpy it works with all the scientific stack and that was um i i hadn't 00:48:48.100 --> 00:48:51.380 heard of that one i thought it was kind of interesting and want to put that out there 00:48:51.380 --> 00:48:57.180 next time you need to actually do something with units and convert uh back and forth might want to 00:48:57.180 --> 00:49:02.560 consider that and then it looks like there's a lot of like physics and chemistry type things like 00:49:02.560 --> 00:49:07.480 the mass of the earth the radius of the earth as constants probably pi and e and all those things 00:49:07.480 --> 00:49:12.660 yes exactly and i think it's when you start getting into it there's probably temptation just 00:49:12.660 --> 00:49:17.640 to code it all yourself just put those constants in there but uh when it starts to get more 00:49:17.640 --> 00:49:22.980 complicated i think something like this could be really uh useful and then there's another approach 00:49:22.980 --> 00:49:29.540 called pint which also works there we go uh which also works with units and it has a little bit 00:49:29.540 --> 00:49:34.820 different uh approach and so i i think it's good to take a look at both of them and if you have a need 00:49:34.820 --> 00:49:39.840 then you can decide which api is going to work best for your unique situation yeah these are cool i 00:49:39.840 --> 00:49:46.300 haven't looked at unit but i love pint and i think the name is so good yes uh because my wife asked me 00:49:46.300 --> 00:49:53.080 like how many else's are in a pound or how many pints are in a liter i'll be like or even in i don't know 00:49:53.080 --> 00:49:59.300 a court or vice versa i'm like i have no idea i just these are such messed up volume measures and so 00:49:59.300 --> 00:50:03.040 it's like here's the thing that takes the thing you don't really know about and allows you to convert it 00:50:03.040 --> 00:50:07.460 to the others in a safe way it's good exactly i got one more quick extra to throw out as well for 00:50:07.460 --> 00:50:12.360 you chris there i forgot to mention that you were the author of the move to from excel to python 00:50:12.360 --> 00:50:17.320 with pandas course over at talk python trainings which is a really popular course basically it's 00:50:17.320 --> 00:50:22.740 a intro to pandas course disguised as solving problems you might with excel right exactly yeah 00:50:22.740 --> 00:50:27.460 yeah no thanks and uh i've had a lot of good feedback from folks so hopefully it's uh 00:50:27.460 --> 00:50:32.460 interesting to the listeners that haven't had a chance to check it out i might have to buy that for 00:50:32.460 --> 00:50:39.560 my boss maybe you can get that discount code yeah yeah the discount code all right you ready for 00:50:39.560 --> 00:50:45.680 some jokes my twitter came back so i can show the twitter joke now yeah all right so dean who is 00:50:45.680 --> 00:50:51.400 often but i don't see him this day today on the live stream sent a joke over and said do you know how 00:50:51.400 --> 00:50:59.860 they say async in italian async yo or async io which i thought was a pretty good one async io i think you 00:50:59.860 --> 00:51:05.460 love italian all right um you guys got another one out there i saw one in the notes another joke but i 00:51:05.460 --> 00:51:12.120 didn't see who put it there i i've got one um so does anyone know why cryptocurrency engineers 00:51:12.500 --> 00:51:19.720 allowed to vote no i don't know because they're minors oh that's a good dad joke yeah it is 00:51:19.720 --> 00:51:27.180 absolutely well on that high note let's uh let's call a show what do you what do you say 00:51:27.180 --> 00:51:31.800 yep right thanks as always chris thanks for joining us this time thank you very much really appreciate 00:51:31.800 --> 00:51:36.780 it yeah bye everyone thank you for listening to python bytes follow the show on twitter via at 00:51:36.780 --> 00:51:44.380 python bytes that's python bytes as in b-y-t-e-s and get the full show notes at pythonbytes.fm if you 00:51:44.380 --> 00:51:49.660 have a news item you want featured just visit pythonbytes.fm and send it our way we're always on the lookout for 00:51:49.660 --> 00:51:54.820 sharing something cool on behalf of myself and brian aukin this is michael kennedy thank you for 00:51:54.820 --> 00:51:57.440 listening and sharing this podcast with your friends and colleagues