Summary
Jackie Kazil has led a distinguished and varied career with a strong focus on providing information and tools that empower others. This includes her work in data journalism, as a presidential innovation fellow, co-founding 18F, co-authoring a book, and being elected to the board of the Python Software Foundation. In this episode she shares these stories and more with us and how Python has helped her along the way.
Brief Introduction
- Hello and welcome to Podcast.__init__, the podcast about Python and the people who make it great.
- I would like to thank everyone who has donated to the show. Your contributions help us make the show sustainable.
- When you’re ready to launch your next project you’ll need somewhere to deploy it. Check out Linode at linode.com/podcastinit and get a $20 credit to try out their fast and reliable Linux virtual servers for running your application.
- You’ll want to make sure that your users don’t have to put up with bugs, so you should use Rollbar for tracking and aggregating your application errors to find and fix the bugs before your users notice they exist. Use the link rollbar.com/podcastinit to get 90 days and 300,000 errors for free on their bootstrap plan.
- Visit our site to subscribe to our show, sign up for our newsletter, read the show notes, and get in touch.
- To help other people find the show you can leave a review on iTunes, or Google Play Music, and tell your friends and co-workers
- Join our community! Visit discourse.pythonpodcast.com to join other listeners of the show and share ideas for how to make it better.
- Your host as usual is Tobias Macey and today I’m interviewing Jackie Kazil about her work with 18F, writing Data Wrangling with Python, and her career with Python.
Interview with Jackie Kazil
- Introductions
- How did you get introduced to Python?
- Looking at your background it shows that you got your start in Journalism and that you are now working on an additional degree in Computational Social Science. Can you share a bit about that journey and what set you on that path?
- What is computational social science and what has your particular focus been within that field?
- How has your work in news media prepared you for your current role?
- One of your many notable achievements is co-founding 18F. Can you start by explaining what that organization is and how you got involved in the efforts to build it?
- What are some of the notable uses of Python at 18F?
- In what ways did your experience working with 18F differ from the work you have done at companies outside of government?
- You recently co-wrote and published Data Wrangling with Python through O’Reilly Media. What kind of subject matter do you cover in the book and who is the target audience?
- There are a number of resources available to learn the various tools for working with data in Python. What is the gap that this book is aiming to fill and how did you get started with it?
- What are some of the most interesting things that you learned while working on the book?
Keep In Touch
Picks
- Tobias
- Jackie
Links
The intro and outro music is from Requiem for a Fish The Freak Fandango Orchestra / CC BY-SA
[00:00:14]
Unknown:
Hello, and welcome to podcast dot in it, the podcast about Python and the people who make it great. I'd like to thank everyone who has donated to the show. Your contributions help us make the show sustainable. When you're ready to launch your next project, you'll need somewhere to deploy it. So you should check out linode@linode.com/podcastin it to get a $20 credit for trying out their fast and reliable Linux virtual servers for running your application. You'll also want to make sure that your users don't have to put up with any bugs, so you should use Rollbar for tracking and aggregating your application errors to find and fix the bugs before your users notice they exist. Use the link rollbar.com/podcast in it to get 90 days and 300, 000 errors tracked for free on the bootstrap plan. You can visit our site, subscribe to the show, sign up for the newsletter, read the show notes, and get in touch. And to help other people find the show, you can leave a review on Itunes or Google Play Music and tell your friends and coworkers.
You can also join the community at discourse.pythonpodcast.com to join other listeners of the show and share ideas for how to make it better. Your host as usual is Tobias Macy, and today, I'm interviewing Jackie Kaisel about her work with 18 f, writing data wrangling with Python, and her career in general with Python. So, Jackie, could you please introduce yourself?
[00:01:16] Unknown:
Yes. Hi. I'm I'm Jackie Hazel. I've been doing Python for about 8 years now, give or take. I started, while working in journalism, and then I moved to government. And now I'm in the, private sector. I also, sit on the board of the Python Software Foundation and do a lot of work with, PyLadies, both locally and internationally. And I wrote a book on data wrangling with Python.
[00:01:45] Unknown:
And how did you first get introduced to Python?
[00:01:48] Unknown:
Well, that's that's quite interesting. So okay. So in the mid 2000, there was, Django was created, and it was 1 of the things, that was the thing to learn amongst journalists who were programming in journalism to create data applications. And at the time, I was working for an organization called Investigative Reporters and Editors and the National Institute For Computer Assisted Reporting, and that was while I was at the University of Missouri. And some of the, I believe it was Adrian Holovatti who created Django ended up, he was at the University of Missouri, and then he ended up in, Lawrence, Kansas.
And there was a lot of, excitement around, Python as it related to Django and data applications. So when I was working on my master's, at University of Missouri, I started to pick it up then, but I didn't really learn it until I learn it learn it until I was working on my master's project at the Washington Post having to do with exit poll data and the 2008 elections.
[00:02:56] Unknown:
Yeah. Polling and elections are something that is on everyone's mind these days, I'm sure. So I'm curious to get your thoughts on the your juxtapositions between the 2008 elections and what we recently experienced in terms of the use of polling and the way that the data was interpreted.
[00:03:14] Unknown:
Sure. So I don't know that I follow that that closely. 1 of the things I will point out because we tried to do this at The Washington Post was we tried to take they had all the data, historically from polling. And since the polling questions and responses and sometimes the scales changed over the years, there's no consistent way to compare polling from 1 year to another year without inserting a lot of assumptions on what those questions mean and the intent of the answers to try to get the scales and questions to match up. Because we had we had, like, at The Washington Post, there was, like, 30 years or something of polling data, and they wanted to show sort of sentiment and how people felt over time.
But we kept on running into this problem of this is not this question is not the same as this question. And sometimes there was even can inconsistencies in the questioning from quarter to quarter. So it's really it's really hard to, to compare the 2 as apples to apples.
[00:04:18] Unknown:
Yeah. Another thing that compounds the issues with trying to attribute any accuracy to the polls is the fact that the technological means by which the polls are conducted has shifted a lot in recent years because it used to be that you would call somebody up on their landline telephone to, you know, ask them the polling questions. But now people increasingly don't have landline telephones, and so we have to try and approximate that same level of coverage by other means. And it introduces a lot of bias into the data collection process, so it makes it really difficult to be able to get any sort of meaningful assumptions out of the other end.
[00:04:53] Unknown:
Yeah. That's that's a great point, and, I'll add to that too that I do work with a group in, DC called the ByteBack, b y, t e, Back. And the purpose of this group is to teach, digital literacy. And the first task that they give when somebody comes in the door is take an email and attach something to it and send it. And some of the folks are are unable to do that and don't have that level of digital literacy. And so if you're depending on electronic means, you're completely canceling folks who didn't grow up with computers, who might not have been exposed to technology, might not be able to afford technology, and, participate in the same way that you or I might when we're, you know, on computers all day long.
[00:05:38] Unknown:
So going back to your history in journalism, when I was looking at your background, it showed that you are also working on an additional degree in computational social science. So I'm wondering if you could share a bit about your journey along that path and, what got you started.
[00:05:55] Unknown:
Sure. So I was at the, Washington Post, and I had a friend there who, was a little bit of a mentor of mine as well. His name was Ryan O'Neil. And so I got my master's in journalism, and Ryan decided to start his PhD in operations research. And, you know, I said to him that I was unfulfilled, that I wanted to do more. And he told me about operations research, and we talked about some of the problems he was working on. And I was like, yeah. That's great. But I think there I started to talk about problems that were too complex to solve with equations and talked about the sort of the human aspect of systems, which the operations research doesn't necessarily account for. And he pointed me towards this other program at the same university he was at in Virginia called computational social science. I wasn't necessarily interested in in working towards a PhD, but, it kind of accidentally happened.
I started taking the program, and I I really loved it. It kind of takes the humor human behavior aspect of some of the things that a journalist accidentally learns just by talking to people and learning about people, and adds the computational aspect of creating large so computational social sciences is the study of complex systems, includes things like network analysis, agent based modeling. And when you look at the systems, instead of sort of taking individual people, it takes it models a system based on individual decisions and the effect those individual decisions can have on each other. So it's just kind of a, you know, interesting way to to explore and grow in the next step. For me, because I I was like, okay. Well, you know, I've learned, you know, I got my journalism degree. I learned Python.
I wanna learn something now, and what is that thing? And so that's how I fell in love with computational social science.
[00:07:54] Unknown:
And how do you think your work with news media prepared you for that particular field? And are there any particular lessons that you're able to draw from your work history that you can apply to your current research?
[00:08:08] Unknown:
A lot of the modeling that happens in computational science specific to agent based modeling is about modeling individual behavior that's not always with humans. It could be with animals. It could be with other entities like cells. But, a lot of the work that I do is, about humans, and I will say that journalism taught me a couple of things. 1, the ability to actually talk to other people, even if it was by force because the assignments that I had, you know, you have to you have to be able to complete your assignment. And so, therefore, you get that extra perspective, that might might be unlike your own. It also taught me a little bit about how assumptions and ethics. So 1, never make assumptions.
And if you make assumptions, then you test those assumptions and make sure that the assumptions or the, you know, they're not necessarily assumptions, maybe they're hypotheses. And you test those hypotheses, and then you change the model to adjust the findings of your test. But always assume that whatever you're building is not necessarily correct. And the second thing is ethics and that there are sometimes real humans behind the thing that you are building. Meaning, you know, there are class projects or whatever, but if you are making the assumption, if, a, you talk to someone and they provide information to you and you use that information to, direct your model, that in some way that information could be tied to that person and making sure that they're protected in some way so that the information they provided doesn't somehow backfire against them, and that the model itself does not create some output that would be, harmful to them as a result of that.
[00:09:55] Unknown:
And for the work that you're doing in the realm of computational social science, are you still using some of the standard Python data analysis tools like pandas and NumPy and scipy?
[00:10:07] Unknown:
Yes. And, in addition to that, I built with a colleague of mine a agent based library in Python, and we were so, you know, for social network analysis, I use mostly network x. For agent based modeling now, I've been working towards building this my own library. It's had so far 26 contributors, I believe. And interestingly, I discovered the main time or the main point at which someone becomes a contributor is, during sprints. It is an amazing time to push a library forward, made great progress. There are 3 core contributors right now, and we are, I think, on the verge of sort of making it a solid 1.
I say that, but these things take time. So when I say on the verge, let's say within the next year, not like next week. But yeah. So that's that's sort of my tooling of choice. And what's the name of that library? The name of the library is Mesa. It's under project Mesa on GitHub. The there's a bunch of links in there that show at the bottom a tutorial docs, email list. It's listing on PyPI and,
[00:11:20] Unknown:
how to contribute back. So you mentioned that through your work in journalism, you started getting involved with government. And as part of that, you ended up being 1 of the cofounders of the 18 f organization. So I'm wondering if you can explain a bit about what 18 f is, what it does, and how you got involved in the efforts to get it started.
[00:11:41] Unknown:
Sure. So after I left The Washington Post, I joined a contract at the Library of Congress, and I worked on a a couple of projects relating to old newspapers. I was an advisor on the remake of congress.gov, which was something that I was congressional data was something I was using as a journalist on the outside. For me, moving from journalism to government was it was a scary jump because it's a a complete industry shift. And if you have sort of 1 knowledge set of or sort of 1 safe area that you're comfortable with, it's difficult to, like, completely switch industries to another 1 and saying, well, you know, will I be happy doing this, you know, thing, and how is it different from where I'm coming from and so on and so forth. But, you know, the thing thing that connected journalism to government for me was the sort of civic I guess, in that particular case, you could say civic duty slash empowering citizens to be able to, make better decisions.
And it was it it felt good too to be also to be able to change some of the things that were frustrating about government from the inside as opposed to being 1 of the people on the outside who was like, well, that needs to be fixed, and that needs to be fixed, and that needs to be fixed, and then and then moving on. It was it was really fun to go be on the inside and be like, oh, that needs to be fixed. Let's fix it. So after the Library of Congress, I became a presidential innovation fellow, and that focused on disaster response and recovery and, working on a joint project with, FEMA and the National Geo Spatial Intelligence Agency. When some of the PIFs PIF is the short for Presidential Innovation Fellow. Some of the PIFs ended up finishing, I think, in January of 2014, December or January. So they got some of the early start on 18 f.
I continued my fellowship because mine was extended until May or June of 2014. And throughout that, we kind of worked together. Even though I was working on the fellowship, I was still, helping to stand up, 18 f. And the purpose of it was that, at the time, the Presidential Innovation Fellows were stationed often 1 person at an agency. Sometimes there were 2 or 3 people working on a project, but it wasn't scalable to take 1 person and have them fix all the things in 1 year at a particular agency. We were empowered, but we weren't, you know, we weren't the head of the agency. And and and if we were the head of the agency, we have more things to worry about than how technology works within that agency. So I ended up in a pretty good position with my Presidential Innovation Fellows project where I was able to build it out. It took off. National Geospatial Intelligence Agency, was able to run with it, and it's actually been covered in the news and I believe presented to president Obama. It's called, GeoQ. I just included a a link for you. And, but often, you know, there was this somebody would build something out or push something forward and the support wasn't didn't exist for offboarding because it was this, you know, very we're gonna give you a presidential innovation fellow, and they're gonna have an impact. But, like, it's hard to sort of have that impact and also tie up everything perfectly within 6 months to a year. So we stood up 18 f to be able to make this idea of innovation and technology more scalable within government. And 18f provides 2 major, 2 major services, at least when I when I left. And I I believe there's continuing to do that at this 0.1 of those is direct to build out. So they build something out for you, and then they work with you to hand off the the project and create a support structure internally for you to be able to manage it. And then the second thing is consulting. The consulting is not so much consulting, but it's more like a little bit of consulting, a little bit of education, where they sort of help build the capacity, for you to deliver on your own. And some of this is also it overlaps a little bit with USDS, United States Digital Services, and what they do, but the 2 the missions between the 2 are slightly different.
Actually, you know, they they are different. But at the same time, they often work together because they're going to be working in the same agencies and and and doing different things. But sometimes they overlap, like US people who are in USDS and people who are in 18 f both gathered round and, worked in the early days around health care dot gov when the the issues with health care dot gov were happening.
[00:16:17] Unknown:
And what are some of the notable uses of Python within 18 f that you're aware of?
[00:16:22] Unknown:
Well, Python at 18 f was used for, the FOIA portal project. Just paste a link for you. Openfoya.gov. Python's also used in the, FEC API and that is here. So before that, the FEC is the Federal Election Commission. The data was, less than optimal for people to use the data programmatically. And so part of that, the creation of the FEC API made it a lot easier for journalists and political campaigns and lobbyists and whoever else to understand where money was going to in Washington when it came to elected officials. And some of those were, some of the spellings and names and everything was were were difficult to even match, and so this API helps provide a more normalized dataset.
I think also not at 18 f. There's a gentleman named Sean Herron, who did his presidential innovation fellowship at the FDA, and he created an API for adverse side effects of drugs. And I believe
[00:17:36] Unknown:
Python was used for that as well. And how have your experiences working in government and with a t and f differed from the experiences that you've had now that you've gotten back into industry?
[00:17:49] Unknown:
So, you know, I think it's a little different because 1 of the things that's different is going from industry before to government and then now going from government back to industry. The the difference between the 2 is before and now is a little little difficult to compare because, I'm at 2 different points in my career. I will say that I think a lot of the things that I learned about working in government, especially being in an empowered position is made me stronger in technology because I'm able to now understand more of, the, I guess, the people dynamics and why people make the decisions that they make.
I will say, overall, 1 of the reasons why I left government, If I had never done 18 f, if I never been a presidential innovation fellow, I would totally be applying, and I would I would be totally applying to both of those places right now. I think everybody should do it. But at the time, I had been in government for a couple of years, and I felt that I was losing some of my ability to I was becoming a little complacent at points. And I don't wanna say complacent like I didn't wanna do things, but I think there was there was 1 conversation I had where somebody said, we wanna launch this thing in 5 years, so we're gonna do a code freeze in 3, and which means that there would be 2 years of frozen code. And I remember at that point, I had said, well, that's ridiculous.
We can code freeze in in 4 years. And then I walked out of the room, and I'm like, what? That is ridiculous. We're gonna have a year of frozen code? What did I just say? And that's when I was like, I need to get out of here. I I have I have started to sort of, I brought in these these great sort of private sector practices, but then I started to adopt, some practices of risk aversion. And suddenly, I was too afraid to I was I was still not quite like, well, we need to have like a 5 year code freeze or something, but, I I wasn't that bad. But I I could see myself, like, when when I said that thing in a meeting, And then I went back afterwards, and I was like I wrote a whole thing, and I was like, this is not like, we don't need to code freeze this long. This is why you have all these great practices like CICD and testing and so on and so forth. And but yeah. So that that's kind of the the difference for me. I I I decided, yes. This is probably the best move for me, and I ended up moving to Capital 1. And and Capital 1 is it was that was another difficult move for me. It was it was something like the journalism to government was that, you know, that jump, like, is government really gonna make me happy? Journalism make me happy, so on and so forth. And so the same thing I went through the same evaluation with moving from government to Capital 1.
I don't know that I would have moved for any private sector company, but everything to me as I was evaluating Capital 1 as a company, all the signals to me sent that this is a healthy company where they care about people, and they they not only care about their employees, but they care about their customers, and they want to empower people to make better decisions. And so for me, that was that was a little bit of the sort of empowering message that happened with my transition of government to government and that you're empowering citizens to make better decisions, and it carries on this, sort of per personal public service mission I have. I don't know that I would have gone to any private sector company, though.
[00:21:30] Unknown:
So you mentioned during your introduction that you have also cowritten a book called data wrangling with Python. So I'm wondering if you can describe a bit about what kind of subject matter you cover there and who you targeted for the audience of that book.
[00:21:43] Unknown:
Yeah. So the the book was originally written, with journalists becoming, journalists who want to get into data in mind. That was the sort of original I don't wanna say the original audience because it quickly shifted, and then it moved to just general people who want to learn more about using data. And what I had noticed as somebody who was self taught, what I had noticed was that there were a lot of great books out there, when it came to specific scientific libraries, but not any that sort of dealt with data and Python and what 80 or 90 percent of what you do with data is. And so I started writing the book. I think it was it took 2 or 3 years because it was also during the same time as the Presidential Innovation Fellowship and, 18 f. And so it was a little bit of a it was a a lot going on during that time.
And, the idea being though that when somebody says to someone else, you should learn how to use Python to do data stuff, or you should use Pandas or something like that. If they pick up a book on pandas, that they're not confusing what is, like, what is Python versus what is pandas. And being able to give them the building blocks of, you know, basic data structures to parsing different data types to APIs to doing a little bit of larger data stuff. And the the later chapters tend to kind of dabble in a little bit of stuff and then say, okay. If this is the topic you're interested, go check out this other book. And if this if this is the topic you wanna dig more into, go check out this other book because we weren't trying to write all the things, but give them give folks an entry point into a variety of topics. My co author, Catherine Jarmel, who is amazing, she wasn't my co author at first, but I pitched the I pitched the book. I got the contract with O'Reilly, and, there were just too many things happening. And, I think I was on chapter 4, maybe 5, which was the PDF chapter. And I was really feeling down because parsing data from PDFs.
I just felt like I kept on writing and writing and there was there was no writing out of that. It was like I was always gonna be parsing data from PDFs, and I was never gonna finish the book because the PDF chapter was gonna eat me alive. And, the other first the first couple of chapters, like, were nothing. And I don't know. I think I was about somewhere between 70 to a 100 pages into the book, and I realized I needed, a co author to hold me accountable. And I made a list of people. Catherine was my first, and I I asked her if she wanted to do it, and she's amazingly passionate and, and intense at times.
We, she was living in Germany and we would I'd wake up at, like, 5 or 6 sometime to have meetings with her in the morning, and it was already, like, midday for her. And she would be like, good morning. And it was it was it was, it was it was fun. It was fun. But she's also she's also the, founder of the original Pie Ladies Group in LA, 1 of the founders out there, and before she moved to Germany. And, so we had known each other through PyLadies. But before that, we had known each other because we were both, sort of data journalists at the Washington Post. She was working in the local section, and I was a generalist across the whole newspaper. And we we had done a lot of work together before, and so it was great to just form this partnership again after many years. Whenever we felt like we had lost our way as far as, like, how do we write this, who is this for, we always went went back to the the journalist who might know how to use Excel or maybe Access or something else and wants to be able to do more, but can't because their tools are limiting them. And so that was always our target audience even if it did even if it was widely more widely applicable to a general
[00:25:31] Unknown:
audience. Yeah. It's definitely useful to have resources available for people who are getting into a new field because in a lot of cases, the available amount of knowledge is so broad. It's really hard to be able to know what you don't know. And having a having an accessible resource that you can look to that will at least give you the introductory aspects of a number of different tools so that you at least are familiar with what some of the options are and what you would use them for is it's a very useful resource to have because it's easy to say, oh, everybody's talking about pandas. So now I I think I should just use pandas for everything, not really understanding it, but well enough to know when it's not the right choice and when you should pick up something like NumPy or something like just using the Excel read and Excel write libraries for being able to interface with Excel data.
[00:26:18] Unknown:
Yeah. Abs absolutely. And when we started the book too, we started with, like, let's not even use some of those tools, but let's talk about let's just use CSD reader to be able to show where some of the the basic building blocks are. And then after that, say, well, yeah, you can you can do this in NumPy, you can do this in Pandas, that kind of thing to show them how these things work together. 1 of the biggest decisions in the book, it's still something we go back and forth on when we talk about, you know, the next version is Python 27 plus versus 3. And when we talk about Python 27 versus 3, we publish that using Python 2 7. Now before it made it to publication, it was February of 2016.
And by that time, all the work that I was doing at 18 f and all the work that I was doing in my personal life, the the modeling library that we had, that I've been working on, all of that was Python 3. But we purposely chose Python 27 for the book because Python 27 is default comes default on Mac, and we provide instructions to doing Python 3. And we would love to have everyone doing Python 3, but for a beginner, we felt like it could be very frustrating and prohibitive to being able to actually do something if you spend, you know, the first hour or a couple of hours or maybe days trying to figure out how to set up Python 3. And also, doing things on your system that you might have never done before, which might be, also very scary when you're doing stuff on the command line, and you're like, wait, what am I typing? Is this gonna destroy my machine? So that was that was something we always go back on, and I I want to figure out a way to up update the book to Python 3. And and I've been trying to figure out who from Apple I could get to you don't have to use it in the OS. Just make it standard on Mac that Python 3 is available just so I can I can update the book with without discussion?
[00:28:13] Unknown:
The startup cost for Python and a number of other programming languages is often so high that it's discouraging for people to even want to proceed with it at all because as soon as they open up their computer and they say, okay. Now how do I actually do this? And they say, you know, there's 15 steps, then they say, okay. Forget it. I'll just go do something else.
[00:28:32] Unknown:
Exactly. Exactly. And now you still have to execute some of those steps when you're on a Windows machine. But if you have a Mac, you don't have to execute those steps because you could just type in Python and you have a prompt. And it's so powerful to be able to show somebody that, and they can have a file on their machine, and all of a sudden they are reading a CSV a couple lines later and playing with the data. And they they're they're taken from, oh, I wanna do something in Python to actually doing in something in Python within, you know, 30 minutes.
[00:29:03] Unknown:
And what are some of the most interesting things that you learned while you were working on the book?
[00:29:07] Unknown:
Well, you know, there's nothing like having to explain something to someone else that solidifies your ability to know it yourself. And so 1 of the things I always recommend when I am doing meetups and stuff like that and somebody is like I asked somebody, I was like, oh, could you teach a class on x? And they say, oh, well, I know how to do x, but I don't know if I can teach it. And I was like, do you wanna become an expert in x? And they're like, that would be great. And I was like, okay, then you should teach the class. Because what you're gonna do in the process of preparing for that is try to learn it as best you can. So when learn it well enough to be able to process it and explain it in a way that makes sense. So you need to understand the topic first before you're able to explain it, because you need to make sure that the the way you're explaining it is in an order that makes sense. And so that was 1 of the things that was really helpful. Another thing that I learned, I think, was about the the publishing process. I have a coworker right now. His name is Azat.
He has written a few notebooks. When I say a few, I think he's up to 15. And, I think when I started at my current job, he was at 11. I don't I don't know. And then 1 day I introduced him and he's like I said, yeah, missus Azad, he's written 11 books. And he's like, no, actually 13. And then, like, 2 months later, I'm like, yeah, missus Azat, he's written 13 books. And he's like, no, 15. And I was like, I've written 1 book. What like, how did you write 15? And, like, I barely made it through 1. So 1 of the things I had learned is that there are a couple of different types of book. This is more of like the introduction book, so it has all the things, but it's actually, I think sometimes nicer and more digestible to write smaller books that touch on 1 topic. I think some of this I think this book could have probably been been divided into 2 books and or maybe it could have stopped around chapter 9, which was data exploration and analysis, and then the rest of that could have been turned into maybe micro books or something. So that was that was another thing. I think another thing I learned about this process was the actual comparison of languages mentioned. When I started, I wasn't thinking of including that, but then I realized that when I was explaining the book to folks, people were just confused by words that were being used, who didn't have a lot of exposure with Python. So the idea being that, you know, what is JavaScript versus Python? Why should I use 1 of the over the other? Should why should I use you know, people said I should use r. Other people I should said I should use Python. Why would I use 1 of the other? My friend is using Java or c or c plus plus, and why that, you know, versus Python? You know, why HTML versus Python? I thought I should learn HTML.
So, yeah, the comparison of languages mentioned, that was something that I wasn't expecting to include. But, when I started talking to some of the folks and bouncing who are target audience members and bouncing ideas off of them, That was 1 thing that came up over and over again. Also, this is a side note, but the book was written by 2 women. It was edited by 2 women and then target audience reviewed by 2 women, which is really exciting, especially for women in technology.
[00:32:21] Unknown:
Yeah. That's definitely a very notable achievement, and I appreciate you calling that out. So you also mentioned in your introduction that you're on the board of the PSF. So I'm wondering if you can talk a bit about your experience with that so far and some of the goals that you have during your time on the board.
[00:32:38] Unknown:
Sure. So the board has been really, really interesting. I was excited to join because I wanted to focus more on I think when I went in, I was looking to expand more efforts around diversity. I run I'm the chair of the grants committee as well, and we provide grants around the world. So grants go through the board or they go through the grants committee. The grants committee handles grants in order to because the board meets monthly, the grants committee can keep grants rolling and process them faster. And after being on the grants committee for about a year, I was really excited about sort of expanding it more, and that's expanding not the grants more, but expanding diversity efforts more. And I think that's the platform I ran on. In the end, what I ended up becoming more interested in was actually the, you know, the PSF as an organization.
And I sit on 3 boards, and it was interesting to see how the different boards operate and what each of them do differently. You know, whether they do fundraising, whether they don't do fundraising, what kind of governance processes they have, how accountability works versus, you know, having employees not having employees. So functioning as an organization became very interesting to me and being able to make sure that the PSF is the best it could be. And the the folks with the PSF, there are 4 employees. I think 1 of them is 1 of them is part time, so I don't know how to count that person, but they are all they're just they're fantastic at making the PSF run like a machine and and keeping things on point and herding cats.
But the, as as a result of it, I did speak at a couple conferences this year, and I'll speak, I believe at, PyCon PyCon Caribbean in February. And 1 of the conferences I spoke at this year was actually PyCon, cz in the Czech Republic. And my family was Czech, so I was very excited about this. And when I went to the Czech Republic, I learned quite a few things there about the Python community that was very interesting to me, such as there is a librarian for the Python community in the Czech Republic, which has a bunch of tech books, And he goes to all the meetups in the Czech Republic every month, and you can check out a book from him and then return it, like, at a later meetup. I just thought I thought that was amazing, and that's that showed a lot of dedication to me, because it's like a 2 to 3 hour train ride, you know, to the meet up in another city and then back. In addition to this, I also discovered through the process of being on the PSF board is that the embassies, American embassies abroad have programs called the small grants program, and while we provide grants often, the grants that the PSF provides is not going to be, especially for a PyCon or a larger event, is not going to be enough to be able to fund the whole budget, and so what the small grants do that are provided by US embassies are depending the embassy itself creates the rules and the deadlines, I think, and the topics that they support. But the 1 in the Czech Republic, for example, looked into, supporting science and technology and also wanted to, support women. And the the grants they give, I think, are somewhere between 1, 000 to 20, 000. And, so I think PyCon CZ was going to apply for 1 of those grants. Well then, I looked into it further and found out that the US State Department has issues basically these small grants all over small grants for embassies to to distribute all around the world, and the mandates come from the state department, so there's some similarities.
So, essentially, in this process, and I I still have to I still have to write this up because I wanted to encourage, Python groups around the world to take advantage of this. Essentially, this could be a extra source of, funding for PyCons around the world, taking advantage of these small grants with the US,
[00:36:39] Unknown:
state from the US state department. I probably went a little off track there, but Totally fine. That's definitely useful information for people to be aware of because it's often hard to understand where you can turn to for support for endeavors such as that. So being able to disseminate that information is definitely useful.
[00:36:57] Unknown:
Yeah. And and I'll say too with the grants that the PSF gives out, it's everything. They'll cover meetup fees. They with the grant proposal process, it's very, low hanging. It's could be depending on how much detail. Of course, you can find more detail. We'll ask less questions. It's like half a page to 2 pages. And if you submit and you don't have enough detail, then we will come back to you and say provide x y z. It's not like some grant processes where you submit and if you don't use the right font, they will reject your conferences.
[00:37:40] Unknown:
So are there any other topics that you think we should cover before we close out? No. I I think I'm, I think I'm good. For anybody who wants to follow what you're up to and get in touch, what would be the best ways for them to do that?
[00:37:53] Unknown:
So you can follow me on Twitter. So, basically, I use 1 handle across everything. So as long as you know that handle, you can find me anywhere. And that's Jackie Kazel, j a c kiekazil. So my Twitter handle's at jackiekazel. My email address is jackiekazel@gmail.
[00:38:12] Unknown:
So that's the that's the way to get ahold of me. Excellent. So with that, I will move us into the picks. And my pick today, I recently watched the latest Jason Bourne movie, and all those movies are fairly entertaining and, good distraction for the end of your day. So if you haven't seen any or all of them, I definitely recommend giving them a look. And with that, I will pass it to you. Do you have any picks for us today?
[00:38:36] Unknown:
So, my pick for today is going to be, Czech dumpling dough. And the reason is, my mom makes fruit dumplings, but she also makes meat dumplings out of the same dough. And, it's just this amazing amazing versatile dough like pasta,
[00:38:55] Unknown:
but like the Eastern European version of pasta. I'll see if I can find a recipe that you can include. Okay. Great. Well, I'm sure the audience will appreciate if you can find a recipe that we can put in the show notes. Otherwise, I'll see what I can come up with. Okay. Awesome. Alright. Well, I appreciate you taking the time out of your day to share some of your history and some of your experiences working with Python in various environments. I hope you enjoy the rest of your day. Yeah. Thank you so much.
Hello, and welcome to podcast dot in it, the podcast about Python and the people who make it great. I'd like to thank everyone who has donated to the show. Your contributions help us make the show sustainable. When you're ready to launch your next project, you'll need somewhere to deploy it. So you should check out linode@linode.com/podcastin it to get a $20 credit for trying out their fast and reliable Linux virtual servers for running your application. You'll also want to make sure that your users don't have to put up with any bugs, so you should use Rollbar for tracking and aggregating your application errors to find and fix the bugs before your users notice they exist. Use the link rollbar.com/podcast in it to get 90 days and 300, 000 errors tracked for free on the bootstrap plan. You can visit our site, subscribe to the show, sign up for the newsletter, read the show notes, and get in touch. And to help other people find the show, you can leave a review on Itunes or Google Play Music and tell your friends and coworkers.
You can also join the community at discourse.pythonpodcast.com to join other listeners of the show and share ideas for how to make it better. Your host as usual is Tobias Macy, and today, I'm interviewing Jackie Kaisel about her work with 18 f, writing data wrangling with Python, and her career in general with Python. So, Jackie, could you please introduce yourself?
[00:01:16] Unknown:
Yes. Hi. I'm I'm Jackie Hazel. I've been doing Python for about 8 years now, give or take. I started, while working in journalism, and then I moved to government. And now I'm in the, private sector. I also, sit on the board of the Python Software Foundation and do a lot of work with, PyLadies, both locally and internationally. And I wrote a book on data wrangling with Python.
[00:01:45] Unknown:
And how did you first get introduced to Python?
[00:01:48] Unknown:
Well, that's that's quite interesting. So okay. So in the mid 2000, there was, Django was created, and it was 1 of the things, that was the thing to learn amongst journalists who were programming in journalism to create data applications. And at the time, I was working for an organization called Investigative Reporters and Editors and the National Institute For Computer Assisted Reporting, and that was while I was at the University of Missouri. And some of the, I believe it was Adrian Holovatti who created Django ended up, he was at the University of Missouri, and then he ended up in, Lawrence, Kansas.
And there was a lot of, excitement around, Python as it related to Django and data applications. So when I was working on my master's, at University of Missouri, I started to pick it up then, but I didn't really learn it until I learn it learn it until I was working on my master's project at the Washington Post having to do with exit poll data and the 2008 elections.
[00:02:56] Unknown:
Yeah. Polling and elections are something that is on everyone's mind these days, I'm sure. So I'm curious to get your thoughts on the your juxtapositions between the 2008 elections and what we recently experienced in terms of the use of polling and the way that the data was interpreted.
[00:03:14] Unknown:
Sure. So I don't know that I follow that that closely. 1 of the things I will point out because we tried to do this at The Washington Post was we tried to take they had all the data, historically from polling. And since the polling questions and responses and sometimes the scales changed over the years, there's no consistent way to compare polling from 1 year to another year without inserting a lot of assumptions on what those questions mean and the intent of the answers to try to get the scales and questions to match up. Because we had we had, like, at The Washington Post, there was, like, 30 years or something of polling data, and they wanted to show sort of sentiment and how people felt over time.
But we kept on running into this problem of this is not this question is not the same as this question. And sometimes there was even can inconsistencies in the questioning from quarter to quarter. So it's really it's really hard to, to compare the 2 as apples to apples.
[00:04:18] Unknown:
Yeah. Another thing that compounds the issues with trying to attribute any accuracy to the polls is the fact that the technological means by which the polls are conducted has shifted a lot in recent years because it used to be that you would call somebody up on their landline telephone to, you know, ask them the polling questions. But now people increasingly don't have landline telephones, and so we have to try and approximate that same level of coverage by other means. And it introduces a lot of bias into the data collection process, so it makes it really difficult to be able to get any sort of meaningful assumptions out of the other end.
[00:04:53] Unknown:
Yeah. That's that's a great point, and, I'll add to that too that I do work with a group in, DC called the ByteBack, b y, t e, Back. And the purpose of this group is to teach, digital literacy. And the first task that they give when somebody comes in the door is take an email and attach something to it and send it. And some of the folks are are unable to do that and don't have that level of digital literacy. And so if you're depending on electronic means, you're completely canceling folks who didn't grow up with computers, who might not have been exposed to technology, might not be able to afford technology, and, participate in the same way that you or I might when we're, you know, on computers all day long.
[00:05:38] Unknown:
So going back to your history in journalism, when I was looking at your background, it showed that you are also working on an additional degree in computational social science. So I'm wondering if you could share a bit about your journey along that path and, what got you started.
[00:05:55] Unknown:
Sure. So I was at the, Washington Post, and I had a friend there who, was a little bit of a mentor of mine as well. His name was Ryan O'Neil. And so I got my master's in journalism, and Ryan decided to start his PhD in operations research. And, you know, I said to him that I was unfulfilled, that I wanted to do more. And he told me about operations research, and we talked about some of the problems he was working on. And I was like, yeah. That's great. But I think there I started to talk about problems that were too complex to solve with equations and talked about the sort of the human aspect of systems, which the operations research doesn't necessarily account for. And he pointed me towards this other program at the same university he was at in Virginia called computational social science. I wasn't necessarily interested in in working towards a PhD, but, it kind of accidentally happened.
I started taking the program, and I I really loved it. It kind of takes the humor human behavior aspect of some of the things that a journalist accidentally learns just by talking to people and learning about people, and adds the computational aspect of creating large so computational social sciences is the study of complex systems, includes things like network analysis, agent based modeling. And when you look at the systems, instead of sort of taking individual people, it takes it models a system based on individual decisions and the effect those individual decisions can have on each other. So it's just kind of a, you know, interesting way to to explore and grow in the next step. For me, because I I was like, okay. Well, you know, I've learned, you know, I got my journalism degree. I learned Python.
I wanna learn something now, and what is that thing? And so that's how I fell in love with computational social science.
[00:07:54] Unknown:
And how do you think your work with news media prepared you for that particular field? And are there any particular lessons that you're able to draw from your work history that you can apply to your current research?
[00:08:08] Unknown:
A lot of the modeling that happens in computational science specific to agent based modeling is about modeling individual behavior that's not always with humans. It could be with animals. It could be with other entities like cells. But, a lot of the work that I do is, about humans, and I will say that journalism taught me a couple of things. 1, the ability to actually talk to other people, even if it was by force because the assignments that I had, you know, you have to you have to be able to complete your assignment. And so, therefore, you get that extra perspective, that might might be unlike your own. It also taught me a little bit about how assumptions and ethics. So 1, never make assumptions.
And if you make assumptions, then you test those assumptions and make sure that the assumptions or the, you know, they're not necessarily assumptions, maybe they're hypotheses. And you test those hypotheses, and then you change the model to adjust the findings of your test. But always assume that whatever you're building is not necessarily correct. And the second thing is ethics and that there are sometimes real humans behind the thing that you are building. Meaning, you know, there are class projects or whatever, but if you are making the assumption, if, a, you talk to someone and they provide information to you and you use that information to, direct your model, that in some way that information could be tied to that person and making sure that they're protected in some way so that the information they provided doesn't somehow backfire against them, and that the model itself does not create some output that would be, harmful to them as a result of that.
[00:09:55] Unknown:
And for the work that you're doing in the realm of computational social science, are you still using some of the standard Python data analysis tools like pandas and NumPy and scipy?
[00:10:07] Unknown:
Yes. And, in addition to that, I built with a colleague of mine a agent based library in Python, and we were so, you know, for social network analysis, I use mostly network x. For agent based modeling now, I've been working towards building this my own library. It's had so far 26 contributors, I believe. And interestingly, I discovered the main time or the main point at which someone becomes a contributor is, during sprints. It is an amazing time to push a library forward, made great progress. There are 3 core contributors right now, and we are, I think, on the verge of sort of making it a solid 1.
I say that, but these things take time. So when I say on the verge, let's say within the next year, not like next week. But yeah. So that's that's sort of my tooling of choice. And what's the name of that library? The name of the library is Mesa. It's under project Mesa on GitHub. The there's a bunch of links in there that show at the bottom a tutorial docs, email list. It's listing on PyPI and,
[00:11:20] Unknown:
how to contribute back. So you mentioned that through your work in journalism, you started getting involved with government. And as part of that, you ended up being 1 of the cofounders of the 18 f organization. So I'm wondering if you can explain a bit about what 18 f is, what it does, and how you got involved in the efforts to get it started.
[00:11:41] Unknown:
Sure. So after I left The Washington Post, I joined a contract at the Library of Congress, and I worked on a a couple of projects relating to old newspapers. I was an advisor on the remake of congress.gov, which was something that I was congressional data was something I was using as a journalist on the outside. For me, moving from journalism to government was it was a scary jump because it's a a complete industry shift. And if you have sort of 1 knowledge set of or sort of 1 safe area that you're comfortable with, it's difficult to, like, completely switch industries to another 1 and saying, well, you know, will I be happy doing this, you know, thing, and how is it different from where I'm coming from and so on and so forth. But, you know, the thing thing that connected journalism to government for me was the sort of civic I guess, in that particular case, you could say civic duty slash empowering citizens to be able to, make better decisions.
And it was it it felt good too to be also to be able to change some of the things that were frustrating about government from the inside as opposed to being 1 of the people on the outside who was like, well, that needs to be fixed, and that needs to be fixed, and that needs to be fixed, and then and then moving on. It was it was really fun to go be on the inside and be like, oh, that needs to be fixed. Let's fix it. So after the Library of Congress, I became a presidential innovation fellow, and that focused on disaster response and recovery and, working on a joint project with, FEMA and the National Geo Spatial Intelligence Agency. When some of the PIFs PIF is the short for Presidential Innovation Fellow. Some of the PIFs ended up finishing, I think, in January of 2014, December or January. So they got some of the early start on 18 f.
I continued my fellowship because mine was extended until May or June of 2014. And throughout that, we kind of worked together. Even though I was working on the fellowship, I was still, helping to stand up, 18 f. And the purpose of it was that, at the time, the Presidential Innovation Fellows were stationed often 1 person at an agency. Sometimes there were 2 or 3 people working on a project, but it wasn't scalable to take 1 person and have them fix all the things in 1 year at a particular agency. We were empowered, but we weren't, you know, we weren't the head of the agency. And and and if we were the head of the agency, we have more things to worry about than how technology works within that agency. So I ended up in a pretty good position with my Presidential Innovation Fellows project where I was able to build it out. It took off. National Geospatial Intelligence Agency, was able to run with it, and it's actually been covered in the news and I believe presented to president Obama. It's called, GeoQ. I just included a a link for you. And, but often, you know, there was this somebody would build something out or push something forward and the support wasn't didn't exist for offboarding because it was this, you know, very we're gonna give you a presidential innovation fellow, and they're gonna have an impact. But, like, it's hard to sort of have that impact and also tie up everything perfectly within 6 months to a year. So we stood up 18 f to be able to make this idea of innovation and technology more scalable within government. And 18f provides 2 major, 2 major services, at least when I when I left. And I I believe there's continuing to do that at this 0.1 of those is direct to build out. So they build something out for you, and then they work with you to hand off the the project and create a support structure internally for you to be able to manage it. And then the second thing is consulting. The consulting is not so much consulting, but it's more like a little bit of consulting, a little bit of education, where they sort of help build the capacity, for you to deliver on your own. And some of this is also it overlaps a little bit with USDS, United States Digital Services, and what they do, but the 2 the missions between the 2 are slightly different.
Actually, you know, they they are different. But at the same time, they often work together because they're going to be working in the same agencies and and and doing different things. But sometimes they overlap, like US people who are in USDS and people who are in 18 f both gathered round and, worked in the early days around health care dot gov when the the issues with health care dot gov were happening.
[00:16:17] Unknown:
And what are some of the notable uses of Python within 18 f that you're aware of?
[00:16:22] Unknown:
Well, Python at 18 f was used for, the FOIA portal project. Just paste a link for you. Openfoya.gov. Python's also used in the, FEC API and that is here. So before that, the FEC is the Federal Election Commission. The data was, less than optimal for people to use the data programmatically. And so part of that, the creation of the FEC API made it a lot easier for journalists and political campaigns and lobbyists and whoever else to understand where money was going to in Washington when it came to elected officials. And some of those were, some of the spellings and names and everything was were were difficult to even match, and so this API helps provide a more normalized dataset.
I think also not at 18 f. There's a gentleman named Sean Herron, who did his presidential innovation fellowship at the FDA, and he created an API for adverse side effects of drugs. And I believe
[00:17:36] Unknown:
Python was used for that as well. And how have your experiences working in government and with a t and f differed from the experiences that you've had now that you've gotten back into industry?
[00:17:49] Unknown:
So, you know, I think it's a little different because 1 of the things that's different is going from industry before to government and then now going from government back to industry. The the difference between the 2 is before and now is a little little difficult to compare because, I'm at 2 different points in my career. I will say that I think a lot of the things that I learned about working in government, especially being in an empowered position is made me stronger in technology because I'm able to now understand more of, the, I guess, the people dynamics and why people make the decisions that they make.
I will say, overall, 1 of the reasons why I left government, If I had never done 18 f, if I never been a presidential innovation fellow, I would totally be applying, and I would I would be totally applying to both of those places right now. I think everybody should do it. But at the time, I had been in government for a couple of years, and I felt that I was losing some of my ability to I was becoming a little complacent at points. And I don't wanna say complacent like I didn't wanna do things, but I think there was there was 1 conversation I had where somebody said, we wanna launch this thing in 5 years, so we're gonna do a code freeze in 3, and which means that there would be 2 years of frozen code. And I remember at that point, I had said, well, that's ridiculous.
We can code freeze in in 4 years. And then I walked out of the room, and I'm like, what? That is ridiculous. We're gonna have a year of frozen code? What did I just say? And that's when I was like, I need to get out of here. I I have I have started to sort of, I brought in these these great sort of private sector practices, but then I started to adopt, some practices of risk aversion. And suddenly, I was too afraid to I was I was still not quite like, well, we need to have like a 5 year code freeze or something, but, I I wasn't that bad. But I I could see myself, like, when when I said that thing in a meeting, And then I went back afterwards, and I was like I wrote a whole thing, and I was like, this is not like, we don't need to code freeze this long. This is why you have all these great practices like CICD and testing and so on and so forth. And but yeah. So that that's kind of the the difference for me. I I I decided, yes. This is probably the best move for me, and I ended up moving to Capital 1. And and Capital 1 is it was that was another difficult move for me. It was it was something like the journalism to government was that, you know, that jump, like, is government really gonna make me happy? Journalism make me happy, so on and so forth. And so the same thing I went through the same evaluation with moving from government to Capital 1.
I don't know that I would have moved for any private sector company, but everything to me as I was evaluating Capital 1 as a company, all the signals to me sent that this is a healthy company where they care about people, and they they not only care about their employees, but they care about their customers, and they want to empower people to make better decisions. And so for me, that was that was a little bit of the sort of empowering message that happened with my transition of government to government and that you're empowering citizens to make better decisions, and it carries on this, sort of per personal public service mission I have. I don't know that I would have gone to any private sector company, though.
[00:21:30] Unknown:
So you mentioned during your introduction that you have also cowritten a book called data wrangling with Python. So I'm wondering if you can describe a bit about what kind of subject matter you cover there and who you targeted for the audience of that book.
[00:21:43] Unknown:
Yeah. So the the book was originally written, with journalists becoming, journalists who want to get into data in mind. That was the sort of original I don't wanna say the original audience because it quickly shifted, and then it moved to just general people who want to learn more about using data. And what I had noticed as somebody who was self taught, what I had noticed was that there were a lot of great books out there, when it came to specific scientific libraries, but not any that sort of dealt with data and Python and what 80 or 90 percent of what you do with data is. And so I started writing the book. I think it was it took 2 or 3 years because it was also during the same time as the Presidential Innovation Fellowship and, 18 f. And so it was a little bit of a it was a a lot going on during that time.
And, the idea being though that when somebody says to someone else, you should learn how to use Python to do data stuff, or you should use Pandas or something like that. If they pick up a book on pandas, that they're not confusing what is, like, what is Python versus what is pandas. And being able to give them the building blocks of, you know, basic data structures to parsing different data types to APIs to doing a little bit of larger data stuff. And the the later chapters tend to kind of dabble in a little bit of stuff and then say, okay. If this is the topic you're interested, go check out this other book. And if this if this is the topic you wanna dig more into, go check out this other book because we weren't trying to write all the things, but give them give folks an entry point into a variety of topics. My co author, Catherine Jarmel, who is amazing, she wasn't my co author at first, but I pitched the I pitched the book. I got the contract with O'Reilly, and, there were just too many things happening. And, I think I was on chapter 4, maybe 5, which was the PDF chapter. And I was really feeling down because parsing data from PDFs.
I just felt like I kept on writing and writing and there was there was no writing out of that. It was like I was always gonna be parsing data from PDFs, and I was never gonna finish the book because the PDF chapter was gonna eat me alive. And, the other first the first couple of chapters, like, were nothing. And I don't know. I think I was about somewhere between 70 to a 100 pages into the book, and I realized I needed, a co author to hold me accountable. And I made a list of people. Catherine was my first, and I I asked her if she wanted to do it, and she's amazingly passionate and, and intense at times.
We, she was living in Germany and we would I'd wake up at, like, 5 or 6 sometime to have meetings with her in the morning, and it was already, like, midday for her. And she would be like, good morning. And it was it was it was, it was it was fun. It was fun. But she's also she's also the, founder of the original Pie Ladies Group in LA, 1 of the founders out there, and before she moved to Germany. And, so we had known each other through PyLadies. But before that, we had known each other because we were both, sort of data journalists at the Washington Post. She was working in the local section, and I was a generalist across the whole newspaper. And we we had done a lot of work together before, and so it was great to just form this partnership again after many years. Whenever we felt like we had lost our way as far as, like, how do we write this, who is this for, we always went went back to the the journalist who might know how to use Excel or maybe Access or something else and wants to be able to do more, but can't because their tools are limiting them. And so that was always our target audience even if it did even if it was widely more widely applicable to a general
[00:25:31] Unknown:
audience. Yeah. It's definitely useful to have resources available for people who are getting into a new field because in a lot of cases, the available amount of knowledge is so broad. It's really hard to be able to know what you don't know. And having a having an accessible resource that you can look to that will at least give you the introductory aspects of a number of different tools so that you at least are familiar with what some of the options are and what you would use them for is it's a very useful resource to have because it's easy to say, oh, everybody's talking about pandas. So now I I think I should just use pandas for everything, not really understanding it, but well enough to know when it's not the right choice and when you should pick up something like NumPy or something like just using the Excel read and Excel write libraries for being able to interface with Excel data.
[00:26:18] Unknown:
Yeah. Abs absolutely. And when we started the book too, we started with, like, let's not even use some of those tools, but let's talk about let's just use CSD reader to be able to show where some of the the basic building blocks are. And then after that, say, well, yeah, you can you can do this in NumPy, you can do this in Pandas, that kind of thing to show them how these things work together. 1 of the biggest decisions in the book, it's still something we go back and forth on when we talk about, you know, the next version is Python 27 plus versus 3. And when we talk about Python 27 versus 3, we publish that using Python 2 7. Now before it made it to publication, it was February of 2016.
And by that time, all the work that I was doing at 18 f and all the work that I was doing in my personal life, the the modeling library that we had, that I've been working on, all of that was Python 3. But we purposely chose Python 27 for the book because Python 27 is default comes default on Mac, and we provide instructions to doing Python 3. And we would love to have everyone doing Python 3, but for a beginner, we felt like it could be very frustrating and prohibitive to being able to actually do something if you spend, you know, the first hour or a couple of hours or maybe days trying to figure out how to set up Python 3. And also, doing things on your system that you might have never done before, which might be, also very scary when you're doing stuff on the command line, and you're like, wait, what am I typing? Is this gonna destroy my machine? So that was that was something we always go back on, and I I want to figure out a way to up update the book to Python 3. And and I've been trying to figure out who from Apple I could get to you don't have to use it in the OS. Just make it standard on Mac that Python 3 is available just so I can I can update the book with without discussion?
[00:28:13] Unknown:
The startup cost for Python and a number of other programming languages is often so high that it's discouraging for people to even want to proceed with it at all because as soon as they open up their computer and they say, okay. Now how do I actually do this? And they say, you know, there's 15 steps, then they say, okay. Forget it. I'll just go do something else.
[00:28:32] Unknown:
Exactly. Exactly. And now you still have to execute some of those steps when you're on a Windows machine. But if you have a Mac, you don't have to execute those steps because you could just type in Python and you have a prompt. And it's so powerful to be able to show somebody that, and they can have a file on their machine, and all of a sudden they are reading a CSV a couple lines later and playing with the data. And they they're they're taken from, oh, I wanna do something in Python to actually doing in something in Python within, you know, 30 minutes.
[00:29:03] Unknown:
And what are some of the most interesting things that you learned while you were working on the book?
[00:29:07] Unknown:
Well, you know, there's nothing like having to explain something to someone else that solidifies your ability to know it yourself. And so 1 of the things I always recommend when I am doing meetups and stuff like that and somebody is like I asked somebody, I was like, oh, could you teach a class on x? And they say, oh, well, I know how to do x, but I don't know if I can teach it. And I was like, do you wanna become an expert in x? And they're like, that would be great. And I was like, okay, then you should teach the class. Because what you're gonna do in the process of preparing for that is try to learn it as best you can. So when learn it well enough to be able to process it and explain it in a way that makes sense. So you need to understand the topic first before you're able to explain it, because you need to make sure that the the way you're explaining it is in an order that makes sense. And so that was 1 of the things that was really helpful. Another thing that I learned, I think, was about the the publishing process. I have a coworker right now. His name is Azat.
He has written a few notebooks. When I say a few, I think he's up to 15. And, I think when I started at my current job, he was at 11. I don't I don't know. And then 1 day I introduced him and he's like I said, yeah, missus Azad, he's written 11 books. And he's like, no, actually 13. And then, like, 2 months later, I'm like, yeah, missus Azat, he's written 13 books. And he's like, no, 15. And I was like, I've written 1 book. What like, how did you write 15? And, like, I barely made it through 1. So 1 of the things I had learned is that there are a couple of different types of book. This is more of like the introduction book, so it has all the things, but it's actually, I think sometimes nicer and more digestible to write smaller books that touch on 1 topic. I think some of this I think this book could have probably been been divided into 2 books and or maybe it could have stopped around chapter 9, which was data exploration and analysis, and then the rest of that could have been turned into maybe micro books or something. So that was that was another thing. I think another thing I learned about this process was the actual comparison of languages mentioned. When I started, I wasn't thinking of including that, but then I realized that when I was explaining the book to folks, people were just confused by words that were being used, who didn't have a lot of exposure with Python. So the idea being that, you know, what is JavaScript versus Python? Why should I use 1 of the over the other? Should why should I use you know, people said I should use r. Other people I should said I should use Python. Why would I use 1 of the other? My friend is using Java or c or c plus plus, and why that, you know, versus Python? You know, why HTML versus Python? I thought I should learn HTML.
So, yeah, the comparison of languages mentioned, that was something that I wasn't expecting to include. But, when I started talking to some of the folks and bouncing who are target audience members and bouncing ideas off of them, That was 1 thing that came up over and over again. Also, this is a side note, but the book was written by 2 women. It was edited by 2 women and then target audience reviewed by 2 women, which is really exciting, especially for women in technology.
[00:32:21] Unknown:
Yeah. That's definitely a very notable achievement, and I appreciate you calling that out. So you also mentioned in your introduction that you're on the board of the PSF. So I'm wondering if you can talk a bit about your experience with that so far and some of the goals that you have during your time on the board.
[00:32:38] Unknown:
Sure. So the board has been really, really interesting. I was excited to join because I wanted to focus more on I think when I went in, I was looking to expand more efforts around diversity. I run I'm the chair of the grants committee as well, and we provide grants around the world. So grants go through the board or they go through the grants committee. The grants committee handles grants in order to because the board meets monthly, the grants committee can keep grants rolling and process them faster. And after being on the grants committee for about a year, I was really excited about sort of expanding it more, and that's expanding not the grants more, but expanding diversity efforts more. And I think that's the platform I ran on. In the end, what I ended up becoming more interested in was actually the, you know, the PSF as an organization.
And I sit on 3 boards, and it was interesting to see how the different boards operate and what each of them do differently. You know, whether they do fundraising, whether they don't do fundraising, what kind of governance processes they have, how accountability works versus, you know, having employees not having employees. So functioning as an organization became very interesting to me and being able to make sure that the PSF is the best it could be. And the the folks with the PSF, there are 4 employees. I think 1 of them is 1 of them is part time, so I don't know how to count that person, but they are all they're just they're fantastic at making the PSF run like a machine and and keeping things on point and herding cats.
But the, as as a result of it, I did speak at a couple conferences this year, and I'll speak, I believe at, PyCon PyCon Caribbean in February. And 1 of the conferences I spoke at this year was actually PyCon, cz in the Czech Republic. And my family was Czech, so I was very excited about this. And when I went to the Czech Republic, I learned quite a few things there about the Python community that was very interesting to me, such as there is a librarian for the Python community in the Czech Republic, which has a bunch of tech books, And he goes to all the meetups in the Czech Republic every month, and you can check out a book from him and then return it, like, at a later meetup. I just thought I thought that was amazing, and that's that showed a lot of dedication to me, because it's like a 2 to 3 hour train ride, you know, to the meet up in another city and then back. In addition to this, I also discovered through the process of being on the PSF board is that the embassies, American embassies abroad have programs called the small grants program, and while we provide grants often, the grants that the PSF provides is not going to be, especially for a PyCon or a larger event, is not going to be enough to be able to fund the whole budget, and so what the small grants do that are provided by US embassies are depending the embassy itself creates the rules and the deadlines, I think, and the topics that they support. But the 1 in the Czech Republic, for example, looked into, supporting science and technology and also wanted to, support women. And the the grants they give, I think, are somewhere between 1, 000 to 20, 000. And, so I think PyCon CZ was going to apply for 1 of those grants. Well then, I looked into it further and found out that the US State Department has issues basically these small grants all over small grants for embassies to to distribute all around the world, and the mandates come from the state department, so there's some similarities.
So, essentially, in this process, and I I still have to I still have to write this up because I wanted to encourage, Python groups around the world to take advantage of this. Essentially, this could be a extra source of, funding for PyCons around the world, taking advantage of these small grants with the US,
[00:36:39] Unknown:
state from the US state department. I probably went a little off track there, but Totally fine. That's definitely useful information for people to be aware of because it's often hard to understand where you can turn to for support for endeavors such as that. So being able to disseminate that information is definitely useful.
[00:36:57] Unknown:
Yeah. And and I'll say too with the grants that the PSF gives out, it's everything. They'll cover meetup fees. They with the grant proposal process, it's very, low hanging. It's could be depending on how much detail. Of course, you can find more detail. We'll ask less questions. It's like half a page to 2 pages. And if you submit and you don't have enough detail, then we will come back to you and say provide x y z. It's not like some grant processes where you submit and if you don't use the right font, they will reject your conferences.
[00:37:40] Unknown:
So are there any other topics that you think we should cover before we close out? No. I I think I'm, I think I'm good. For anybody who wants to follow what you're up to and get in touch, what would be the best ways for them to do that?
[00:37:53] Unknown:
So you can follow me on Twitter. So, basically, I use 1 handle across everything. So as long as you know that handle, you can find me anywhere. And that's Jackie Kazel, j a c kiekazil. So my Twitter handle's at jackiekazel. My email address is jackiekazel@gmail.
[00:38:12] Unknown:
So that's the that's the way to get ahold of me. Excellent. So with that, I will move us into the picks. And my pick today, I recently watched the latest Jason Bourne movie, and all those movies are fairly entertaining and, good distraction for the end of your day. So if you haven't seen any or all of them, I definitely recommend giving them a look. And with that, I will pass it to you. Do you have any picks for us today?
[00:38:36] Unknown:
So, my pick for today is going to be, Czech dumpling dough. And the reason is, my mom makes fruit dumplings, but she also makes meat dumplings out of the same dough. And, it's just this amazing amazing versatile dough like pasta,
[00:38:55] Unknown:
but like the Eastern European version of pasta. I'll see if I can find a recipe that you can include. Okay. Great. Well, I'm sure the audience will appreciate if you can find a recipe that we can put in the show notes. Otherwise, I'll see what I can come up with. Okay. Awesome. Alright. Well, I appreciate you taking the time out of your day to share some of your history and some of your experiences working with Python in various environments. I hope you enjoy the rest of your day. Yeah. Thank you so much.
Introduction and Guest Introduction
Jackie's Journey into Python
Challenges in Polling Data
Transition from Journalism to Computational Social Science
Python Tools and Mesa Library
Founding 18F and Government Work
Returning to Industry
Data Wrangling with Python Book
Experience on the PSF Board
Closing Remarks
