Transcript #167: Cheating at Kaggle and uWSGI in prod

00:00 Hello and welcome to Python Bytes, where we deliver Python news and headlines directly to

00:04 your earbuds. This is episode 167, recorded January 29th, 2020. I'm Michael Kennedy,

00:11 and Brian Okken is away. We miss you, Brian, but I have a very special guest to join me,

00:16 Vicki Boykus. Welcome, Vicki.

00:18 Thanks for having me.

00:19 Yeah, it's great to have you here. I'm excited to get your take on this week's Python news. I

00:23 know you found some really interesting and controversial ones that we're going to jump

00:27 into, and that would be great. Also, great is Datadog. They're sponsoring this episode,

00:31 so check them out at pythonbytes.fm/Datadog. I'll tell you more about them later. Vicki,

00:37 let me kick us off with command line interface libraries, for lack of a better word. So back

00:44 on episode 164, so three episodes ago, I talked about this thing called Typer. Have you heard of

00:50 Typer? T-Y-P-E-R?

00:51 I have not, but I've heard of Click, so I'm curious to see how this differs from that even.

00:56 Yeah, yeah. So this is sort of a competitor to Click. Typer is super cool because what it does is it uses

01:03 native Python concepts to build out your CLI rather than attributes where you describe everything. So,

01:11 for example, you can have a function and you just say this function takes a name, colon,

01:16 stir to give it a type or an int or whatever, and then Typer can automatically use the type

01:23 type and the name of the parameters and stuff to generate like your help and the inbound arguments

01:28 and so on. So that's pretty cool, right? Yeah. Seems like a great excuse to start using type

01:32 annotations if you haven't yet. Yeah, exactly. Very, very nice that it leverages type annotations,

01:37 hence the name Typer, right? So our listeners are great. They always send in stuff that we haven't

01:41 heard about or you're like, I can't believe you didn't talk about this other thing. So Marcello sent in a

01:47 message and says, hey, you should talk about CLI's, C-L-I-Z-E, which turns functions into command line

01:54 interface. So CLI's is really cool. And it's very similar in regard to how it works as to Typer. So what you do is you

02:04 create functions, you give them variables, you don't have to use the types in the sense that Typer does. But you have

02:10 positional arguments, and you have keyword only arguments. And you know, Python has that syntax that very few people use, but it's cool.

02:18 If you want to enforce it, where you can say here are some parameters to a function comma star comma, here are some more and the

02:26 stuff after the star has to be addressed as a keyword argument, right? Yeah, so it leverages that kind of stuff. So you can say, like their

02:34 example says, here's a Hello World function, and it takes a name, which has a default of none, and then star comma, no

02:40 capitalize is false, and it gives it a default value. So all you got to do to run it is basically import CLI's dot run and call

02:48 run on your function. And then what it does is it verifies those arguments about whether or not they're required. And then it'll

02:56 convert the keyword arguments to like dash dash, this or that. So like dash dash, no capitalize, will pass true to no capitalize. If you admit it, it'll

03:06 pass, you know, whatever the default is, I guess, so false. So there's like positional ones where you don't say the name, but then also this cool way of

03:13 adding on these, these --capitalize and so on. So it seems like a really cool, and pretty extensive library for building command line

03:21 interfaces. Yeah, so this seems like it'd be good if you have a lot of parameters that you have to pass in. I'm thinking,

03:26 specifically of some of the work that you would do in the cloud, like in the AWS command line. Yeah, yeah. Or similar?

03:31 Yeah, for sure. Another thing that's cool is it will take your doc strings and use those as help messages.

03:37 Oh, that's neat. Yeah, so you know, and like some editors, you can type triple quote, enter, and it'll generate, you know, here's the

03:44 summary of the method, and then here's the arguments, and you can put this, or you can just write them out, of course. And then here's the

03:50 descriptions about each parameter, those become help messages about each command in there. So it's really

03:58 nice. And I like how it uses just pure Python, sort of similar to Typer in that regard, that you don't put

04:04 like three or four levels of decorators on top of things and then reference other parts of that. You just say,

04:09 here's some Python code, I want to treat it as a command line interface, clys.run.

04:14 That is pretty cool. Yeah.

04:16 So there's now a lot of choices if you want to do command line interfaces.

04:19 Yeah, yeah, definitely. And click is good, and it's very popular in argparse as well. But I'm kind of a

04:26 fan of these pure Python ones that don't require me to go do a whole bunch of extra stuff. So yeah,

04:30 definitely loving that. You know what, I bet that Kaggle's not loving what you're talking about next.

04:35 Before we get into this.

04:37 Well, I think they might be, but...

04:39 Yeah, we'll see. Okay. Tell us about Kaggle and what the big news here is.

04:43 Yeah. So there was a dust up at Kaggle a couple weeks ago. So just as a little bit of background,

04:48 Kaggle is a platform that's now owned by Google that allows data scientists to find data sets,

04:54 to learn data science. And most importantly, it's probably known for letting people participate

04:59 in machine learning competitions. That's kind of how it gained its popularity and notoriety.

05:03 Yeah, that's how I know it.

05:04 Yep. And so people can sharpen their data science and modeling skills on it. So they recently,

05:09 I want to say last fall, hosted a competition that was about analyzing pet shelter data. And this

05:16 resulted in enormous controversy. So what happened is there's this website that's called petfinder.my

05:23 that helps people find pets to rescue in Malaysia from shelters. And in 2019, they announced a

05:30 collaboration with Kaggle to create a machine learning predictor algorithm, which pets would be most

05:36 likely to be adopted based on the metadata descriptions on the site. So if you go to petfinder.my,

05:42 you'd see that they'll have a picture of the pet and then a description, how old they are,

05:47 and some other attributes about them.

05:49 Right. Were they vaccinated or things like that, right? Sort of, you might think, well,

05:54 if they're vaccinated or they're neutered or spayed, they may be more likely to be adopted,

05:59 but you don't necessarily know, right? So that was kind of some, like, what are the important factors

06:03 was this whole competition, right?

06:04 Yeah. The goal was to help the shelters write better descriptions so that pets would be adopted

06:10 more quickly. So after several months, so they held the competition for several months and there was a

06:16 contestant that won and he was previously what was called a Kaggle Grandmaster. So he'd won a lot of

06:21 different stuff on Kaggle before and he won $10,000 in prize money. But then what happened is they

06:27 started to validate all of his data. Because when you do a Kaggle competition, you then submit

06:33 all of your data and all of your results and your notebooks and your code.

06:37 Like how you trained your models and stuff like that, right?

06:40 Yeah.

06:40 Okay.

06:41 Yeah. All of that stuff. And then what happened was a pet finder wanted to put this model into

06:46 production. So you initially have something like a Jupyter or a Colab notebook in this case.

06:51 And the idea is that now you want to be able to integrate it into the pet finder website.

06:56 So they can actually use these predictors to fine tune how they post the content. And so when a

07:03 volunteer who was Benjamin Minnicoffer offered to put the algorithm into production and he started

07:09 looking at it, he found that there was a huge discrepancy between the first and second place

07:14 entrance in the contest. And so what happened was, so a little to get more into the technical

07:20 aspect, the data they gave to the contestants asked them to predict the speed at which a pet

07:24 would be adopted from one to five and included some of the features you talked about, like animal

07:28 breed coloration, all that stuff. The initial training set had 15,000 animals. And then after a

07:35 couple of months, the contestants were given 4,000 animals that had not been seen before as a test of

07:41 how accurate they were. So what the winner did was he actually scraped basically most of the website so

07:48 that he got that 4,000 set, the validation set also. And he had the validation set in his notebook.

07:56 So basically what he did was he used the MD5 library to create a hash for each unique pet. And then he looked

08:05 up the adoption score for each of those pets, basically when they were adopted from that external

08:10 data set. And there were about 3,500 that had overlaps with the validation set. And then he did a column

08:18 manipulation in pandas to get at the hidden prediction variable for every 10th pet, not every single pet,

08:23 but every 10th pet. So it didn't look too obvious.

08:25 Right. So he gave himself like a 10% head start or advantage or something like that.

08:30 Exactly. And he replaced the prediction with, that should have been generated by the algorithm with

08:36 the actual value. And then he did a dictionary lookup between the initial MD5 hash and the value

08:42 of the hash. And this was all of obfuscated in a separate function that happened in his data.

08:48 Wow. And so they must've been looking at this going, what does the MD5 hash of the pet attributes

08:53 have to do with anything? You know what I mean? Right. It's the, the hashes are meant to obscure

08:58 stuff, right?

08:59 Right. Yeah.

09:00 So what was the fallout?

09:01 So the fallout was this guy worked at h2o.ai. And so he was fired from there and Kaggle also issued

09:08 an apology where they explained exactly what happened. And they expressed the hope that this didn't mean

09:14 that every contest going forward would be viewed with suspicion for more openness and for collaboration

09:20 going forward. Wow. And it was an amazing catch. Yeah. That's such a good catch. I'm so,

09:24 so glad that Benjamin did that. I've got the whole deal here. Now did Kaggle actually end up paying

09:31 him the 10,000 before they caught it? Is there like some sort of waiting period? Unfortunately,

09:36 I think the money had already been dispersed by that point. Yeah. I can easily see something. Well,

09:42 you know, like the prize money will be sent out after a, you know, very deep, it may change the

09:49 timing of that for sure in the future, who knows, but wow, that's crazy. Do you know why he was fired?

09:54 I mean, they're just like, we don't want you to say, I mean, h2o.ai, they're kind of a,

10:01 we'll help you with your AI story. So I guess, you know, they're probably just like, we don't want

10:07 any of the negativity of that on our product. Yeah, I think that's essentially it. And it was

10:13 a pretty big competition in the data science community. And I think also once they'd started

10:17 to look into it in other places, previously, he talked about just basically scraping data to gain

10:24 competitions as well. So all of that stuff started to come out as well. I think they wanted to distance

10:29 themselves. Yeah, I can imagine. Yeah. Okay. Well, thank you for sharing that. Now, before we get to the

10:33 next one, let me tell you about this week's sponsor, Datadog. They're a cloud scale monitoring

10:38 platform that unifies metrics, logs and traces. Monitor your Python applications in real time,

10:44 find bottlenecks with detailed flame graphs, trace requests as they travel across service boundaries,

10:49 and they're tracing client auto instruments, popular frameworks like Django, AsyncIO, Flask,

10:56 so you can quickly get started monitoring the health and performance of your Python apps.

11:00 Do that with a 14 day free trial and Datadog will send you a complimentary t-shirt, cool little

11:06 Datadog t-shirt. So check them out at pythonbytes.fm/Datadog. This next one kind of hits home for me

11:12 because I have a ton of services and a lot of servers and websites and all these things working

11:17 together, running on micro WSGI, UWSGI. And I've had it running for quite a few years. It's got a lot of

11:25 traffic, you know, we do like, I don't know, 14 terabytes of traffic a month or maybe even more

11:31 than that. So quite a bit of traffic going around these services and whatnot. So it's been working

11:37 fine. But I ran across this article by the engineers at Bloomberg. So they talked about this thing called

11:43 configuring micro WSGI for production deployment. And I actually learned a lot from this article. So

11:50 I don't feel like I was doing too many things wrong, but there was a couple of things I'm like,

11:53 oh yeah, I should probably do that. And other stuff just that is really nice. So I just want to run you

11:58 through a couple of things that I learned. And if you want to hear more about how we're using micro WSGI,

12:03 you can check that out on Talk Python 215. Dan Bader and I swap stories about how we're running our

12:10 thing, our various things, you know, Talk Python training and realpython.com and whatnot.

12:16 So this is guidance from Bloomberg's engineering structured products application group.

12:22 That's quite the title. And they decided to use micro WSGI because it's really, you know,

12:27 good for performance, easy to work with. However, they said micro WSGI is, as it's maturing,

12:34 some of the defaults that made sense when it was new, like in 2008, don't make sense anymore.

12:40 The reason is partly just because the way people use these sites is different or these servers is

12:46 different. For example, doing proxies up in front of micro WSGI with say Nginx, that used to not be

12:53 so popular. So they made these defaults built into the system that maybe don't make sense anymore.

12:58 And so what they did is we're going to go through and they said, we're going to go through and talk

13:02 about all the things that we're going to override the defaults for. And why unbit the developer

13:07 micro WSGI is going to fix all of these bad defaults in the 2.1 release. But right now it's 2.0

13:14 as of this recording. So you're going to have to just, you know, hang in there or apply some of these

13:20 changes. Now, I do want to point out one thing. When I switched on a lot of these, I did them one at a

13:25 time. And the way you get it to reload its config is you say, relaunch the process, restart the process

13:31 with a system CTL, like a daemon management thing from Linux. And one of their recommendations

13:37 is to use this flag die on term, which is for it to die on a different signal that it receives.

13:44 And for whatever reason, maybe I'm doing it wrong. But whenever I turn that on, it would just lock up

13:49 and it would take about two minutes to restart the server because it would just hang until it

13:54 eventually timed out. It was like forcefully killed. So that seems bad. So I'm not using that.

13:58 But I'll go quickly over the settings that I use that I thought were cool here. So there's,

14:03 you've got these complicated config files. If you want to have make sure everything's validated,

14:07 you can say strict equals true. That's cool. That will verify that everything that's typed in the

14:11 file is accurate and is valid because it's kind of forgiven at the moment. Master is true is a good

14:17 setting because this allows it to create worker processes and recycle them based on number of requests

14:22 and so on. Something that's interesting, I didn't even realize you could do. Maybe tell me if you knew

14:28 this was possible in Python apps. You can disable the GIL, the global interpreter lock. You can say,

14:33 you know what, for this Python interpreter, let's not have a GIL.

14:36 Wow. How does that work?

14:37 Yeah. Well, it's, I mean, people talk about having no GIL. It's like, oh, you can do all this cool

14:41 concurrency and whatnot. But what it really means is you're basically guaranteeing you can only have

14:46 one thread. So if you try to launch, say, a background job on a micro WSGI server and you don't

14:52 pass enable threads, it's true. It's just going to not run because there's no GIL and there's no way to start it.

14:57 That's something you want to have on. Vacuum equals true. This one I had off and I turned it on.

15:03 Apparently this cleans up like temporary files and so on. Also a single interpreter. It used to be that

15:07 micro WSGI was more of an app server that might have different versions of Python and maybe Ruby as well.

15:12 And this will just say, no, no, it's just the one version. A couple other ones. You can specify the

15:17 name that shows up in like top or glances. So it'll say like, you can give it, say your website name,

15:24 and it'll say things like that thing worker process one or that thing master process or whatnot. And so

15:30 there's just a bunch of cool things in here with nice descriptions of why you want these features.

15:35 So if you are out there and you're running micro WSGI, give this a quick scan. It's really cool.

15:40 Now, this next one also is pretty neat. So this one comes from the people who did Spacey, right? What do

15:46 they got going on?

15:47 Yep, that's right. So this was just released a couple of days ago and it's called Think and they bill it as

15:54 a functional take on deep learning. And so basically what it there's, if you're familiar with deep

15:59 learning, there's kind of two big competing frameworks right now. TensorFlow and PyTorch and

16:05 MXNet is also in there. So the idea of this library is that it abstracts away some of the boilerplate

16:11 that you have to write for both TensorFlow and PyTorch. PyTorch has a little bit less. TensorFlow

16:15 with Paris on top also has a little bit less, but you end up writing a lot of the same kind of stuff.

16:22 And there's also some stuff that's obfuscated away from you, specifically some of the matrix

16:28 operations that go on under the hood. And so what Think does is, so it already runs on Spacey, which is

16:36 an NLP library under the covers. So what the team did was they surfaced it so that other people could

16:42 use it more generically in their projects. And so it has that favorite thing that we love. It has type

16:49 checking, which is particularly helpful for tensors when you're trying to get stuff and you're not sure

16:55 why it's not returning things. It has classes for PyTorch wrappers and for TensorFlow. And you can

17:01 intermingle the two if you want to, if you have two libraries that bridge things. It has deep support

17:07 for NumPy structures, which are the kind of the underlying structures for deep learning.

17:12 It operates in batches, which is also a common feature of deep learning projects. So they process

17:19 features and data in batches. And then it also sometimes a problem that you have with deep

17:25 learning is you're constantly tuning hyperparameters or the variables that you put into your model to

17:31 figure out how long you're going to run it for, how many training epochs you're going to have, what size

17:36 your images are going to be. Usually those are those clustered in the beginning of your file is kind of

17:41 like a dump or a dictionary or whatever. It has a special structure to handle those as well. So it basically

17:47 hopes to make it easier and more flexible to do deep learning, especially if you're working with

17:53 two different libraries and it offers a nice higher level abstraction on top of that. And the other

17:58 cool thing is, is they have already released all the examples and code that are available in Jupyter

18:04 notebooks on their GitHub repo. So I'm definitely going to be taking a closer look at that.

18:08 Yeah, that's really cool. They have all these nice examples there and even buttons to open them in

18:12 Colab, which is, yeah, that's pretty awesome. This looks great. And it looks like it's doing some work

18:19 with FastAPI as well. I know they hired a person who's maintaining FastAPI, which is cool. Also,

18:25 their Prodigy project. So yeah, this looks like a really nice library that they put together. Cool. And

18:31 Ines has been on the show before. Ines from Explosion AI here as a guest co-host as well. Super

18:37 cool. That's awesome. Yeah. This next one I want to talk about, you know, I'd love to get your opinion

18:41 because you're more on the data science side of things, right? Yeah. Yeah. So this next one, I want

18:45 to tell folks about, this is another one from listeners. You know, we talked about something that

18:51 validates Panda Panda's and like, oh, you should also check out this thing. So this comes from

18:55 Jacob Deppin. Thank you, Jacob, for sending us in. And so it's pandas dash vet. And what it is,

19:02 is a plugin for Flake 8 that checks Panda's code. And it's this opinionated take on how you should use

19:09 Panda's. They say one of the challenges is that the, if you go and search on Stack Overflow or other

19:16 tutorials, or even maybe video courses, they might show you how to do something with Panda's,

19:20 but maybe that's a deprecated way of working with pandas or some sort of old API. And there's,

19:26 there's a better, better way. So the idea is to make pandas more friendly for newcomers by trying

19:32 to focus on best practices and saying, don't do it that way. Do it this way. You know, read CSV. It

19:36 has so many parameters. What are you doing? Here's how you use it. Things like that. So this is based on a

19:43 talk or this linter was created. The idea was sparked by a talk by Ania Kupsinski.

19:50 Sorry, I'm sure I blew that name bad, but at PyCascades 2019 in Seattle,

19:55 I want your code responsibly. So I'll link to that as well. So it's kind of cool to see the evolution.

19:59 Like Ania gave a talk at PyCascades and then this person's like, oh, this is awesome. I'm going to

20:05 actually turn this into a Flake 8 plugin and so on. What are your thoughts on this? Do you like this idea?

20:10 Yeah, I'm a huge fan of it. I think in general, there's been kind of like this,

20:13 I wouldn't want to say culture war about whether notebooks are good or bad. And there was recently

20:18 a paper released, I want to say, not a paper, but a blog post a couple of days ago about how you should

20:24 never use notebooks. There was a talk by Joel Gruse last year about all the things that notebooks are

20:31 bad with. I think they have their place. And I think this is one of the ways you can have,

20:36 I want to say, guardrails around them and help people do things. I like the very opinionated

20:42 warning that they have here, which is that DF is a bad variable name. Be kinder to yourself,

20:47 because that's always true. You always start with the default of DF and then you end up with

20:51 34, 35 of them. I joke about this on Twitter all the time, but it's true. So that's a good one.

20:56 The loc and the .ix and loc, iloc is always a point of confusion. So it's good that they have that.

21:03 And then the pivot table one is preferred to pivot or on stack. So there's a lot of places. So Pandas is

21:08 fantastic, but there's a lot of these places where you have old APIs, you have new APIs, you have people

21:14 who usually are both new to Python and programming at the same time coming in and using these. So this

21:20 is a good set of guardrails to help write better code if you're writing it in a notebook.

21:24 Oh yeah, that's super cool. Do you know, is there a way to make Flake 8 run in the notebook

21:28 automatically? I don't know. You probably can, yeah. It probably wouldn't be too hard.

21:32 Yeah, but I don't know. Yeah, but I think it's interesting that you ask that because that's not

21:36 generally that's something you would do with notebooks. But maybe this kind of stuff will push it in the

21:41 direction of being more like what we consider quote unquote mainstream or just web dev or

21:48 backend programming. Yeah, cool. Well, I definitely think it's nice if I were getting started with

21:52 Pandas, give this a check. You also, if you're getting started with Pandas, you may also be getting

21:57 started with NumPy, right? Yep. So NumPy is the backbone of numerical computing in Python.

22:03 So I talked about TensorFlow, PyTorch, machine learning in the previous stories. All of that

22:08 kind of rests on the work and the data structures that NumPy are created. So Pandas, Scikit-Learn,

22:15 TensorFlow, PyTorch, they all lean heavily, if not directly depend on the core concepts, which include

22:19 matrix operation through the NumPy array, also known as an ND array. The problem was with ND arrays is

22:27 they're fantastic, but the documentation was a little bit hard for newcomers. So Anne Bonner wrote

22:32 a whole new set of documentation for people that are both new to Python and scientific programming,

22:38 and that's included in the NumPy docs themselves. Before, if you wanted to find out what arrays were,

22:45 how they worked, you could go to the section and you could find out the parameters and attributes and

22:50 all the methods of that class. But you wouldn't find out how or why you would use it. And so this

22:55 documentation is fantastic because it has an explanation of what they are. It has visuals of

23:00 what happens when you perform certain operations on arrays. And it has a lot of really great resources if

23:05 you're just getting started with NumPy. The strong recommend for me, if you're doing any type of data

23:10 work in Python, especially with Pandas, that you become familiar with NumPy arrays. And this makes it

23:15 really easy to do so. Yeah, nice. It has things like, how do I convert a 1D array to a 2D array?

23:21 Or what's the difference between a Python list and a NumPy array and whatnot? Yeah, it looks really

23:28 helpful. I like the why. That's often missing. You'll see like, you do this, use this function for

23:34 this and here are the parameters. Sometimes they'll describe them, sometimes not. And then it's just like,

23:40 well, maybe this is what I want. Stack Overflow seemed to indicate this is what I want. I'm not

23:44 sure. I'll give it a try. Right. So I like the little extra guidance behind it. That's great.

23:47 Yeah, it does a really good job of orienting you. Cool. All right. Well, Vicki, those are our main

23:52 topics for the week. But we got a few extra quick items just to throw in here at the end. I'll let you

23:58 go first with yours. Sure. This is just a bit of blatant self-promotion about who I am. So I am a

24:04 data scientist on the side. I write a newsletter that's called Norm Cortac. And it's about all the things that I'm

24:10 not seeing covered in the mainstream media. And it's just a random hodgepodge of stuff. It ranges from

24:15 anything like machine learning, how the data sets got created initially for NLP. I've written about Elon Musk

24:23 memes. I wrote about the recent raid of the Nginx office in great detail and what happened there. So there's a free

24:30 version that goes out once a week and paid subscribers get access to one more paid newsletter per week. But really, it's more

24:36 about the idea of supporting in-depth writing. So it's just vicki.substack.com.

24:41 Cool. Well, that's a neat newsletter. And I'm a subscriber. So very, very nice. I have a quick one for you all

24:48 out there. And maybe two, actually. One, pip20.0 was released. So not a huge change. Obviously,

24:57 this is compatible with the stuff that I did before and whatnot. But it does a couple of nice things. And I

25:04 think this is going to be extra nice for beginners because it's so challenging. You go to a tutorial

25:10 and it says, all right, the first thing you got to do to run whatever, I want to run Flask or I want to

25:15 run Jupyter is you say pip install Flask or pip install Jupyter. And it says you do not have permission to,

25:21 you know, write to wherever those are going to install, right? Depending on your system. And so

25:27 if that happens now in pip20, it will install as if --user was passed into the user profile.

25:35 That's cool, huh?

25:36 That's really neat.

25:37 Yeah, yeah. So that's great. And cache wheels are built from GitHub requirements and a couple of

25:42 other things. So yeah, nothing major, but nice to have that there. And then also, I'd previously

25:48 gone on a bit of a rant saying I was bugged that homebrew, which is how I put Python on my Mac,

25:54 was great for installing Python 3 until 3.7. So if you just, it's even better because if you just say,

26:01 brew install Python, that means Python 3, which not legacy Python, which is great.

26:07 But that sort of stopped working. It still works, but it installs Python 3.7. So that was,

26:12 that was kind of like a, oh, sad face. But I'm sorry, I forget the person who sent this over on

26:18 Twitter. But when the listener sent in a message said, you can brew install Python at 3.8. And that

26:24 works. Why? That's not...

26:25 Is it safe to brew again? I've just started downloading directly from Python.

26:29 I know, exactly. Exactly. So I'm trying it today. And so far, it's going well. So I'm

26:35 really excited that on macOS, we can probably get the latest Python. Even if you got to say the

26:40 version, I just have an alias that re-aliases what Python means in my ZSHRC file. And it'll just say,

26:47 you know, if you type Python, that means Python 3.8 for now. Anyway, I'm pretty...

26:51 Yeah, fingers crossed. So it looks like it's good. And that's nice. Hopefully,

26:55 it just keeps updating itself. I suspect it will, at least within the 3.8 branch.

26:58 All right. You ready to close this out with a joke?

27:00 Yeah.

27:01 Yeah. So I'm sure you've heard the type of joke, you know, a mathematician and a physicist walk into a

27:07 bar and, right, well, some weird thing about numbers and space ensues. So this one is kind of like

27:14 that one. It's about search engine optimization. So an SEO expert walks into a bar, bars, pub,

27:21 public house, Irish pub, tavern, bartender, beer, liquor, wine, alcohol, spears, and so on.

27:27 It's bad, huh?

27:30 I like that. That's nice.

27:32 Yeah, it's so true. Like, you remember how blatant websites used to be like 10 years ago,

27:37 they would just have like a massive bunch of just random keywords at the bottom. Just, you know,

27:42 like it seems like...

27:43 Yeah. And sometimes they would be in white and white text.

27:45 Yes, exactly. White and white.

27:46 You can see them, but then if you highlight it, it would be like whole three paragraphs.

27:49 Here's where the SEO hacker went. I don't think that works so well anymore. But yeah,

27:55 it's a good joke nonetheless. And Vicky, it's been great to have you here. Thanks so much for

28:00 filling in for Brian and sharing the data science view of the world with us.

28:04 Thanks for having me.

28:05 You bet. Bye.

28:05 Bye.

28:06 Thank you for listening to Python Bytes. Follow the show on Twitter via at Python Bytes. That's

28:11 Python Bytes as in B-Y-T-E-S. And get the full show notes at pythonbytes.fm. If you have a news

28:17 item you want featured, just visit pythonbytes.fm and send it our way. We're always on the lookout for

28:22 sharing something cool. On behalf of myself and Brian Okken, this is Michael Kennedy. Thank you for

28:27 listening and sharing this podcast with your friends and colleagues.