Python Project

Julian 0:00

I want a cool project first of all, and it's solving a real need. That's what I love about it. You're solving a real world problem, right. It's the first project that we talked about first is cool, but you know that stuff exists. This doesn't exist, so you could make some real impact and really help people. So that is very impressive. Hello and welcome to the PyBytes podcast. We talk about Python, career and mindset. We're your hosts. I'm Julian Sequeira and I am Bob Valdobos. If you're looking to improve your Python, your career and learn the mindset for success, this is the podcast for you. Let's get started. Hey everyone, welcome back to the PyBytes podcast. I am Julian. I am here with a very special guest. I know I say this every time we have a guest, but this time I mean it. Sorry, everyone else, I'm here with Victoria Colmack. Victoria, welcome to the podcast.

Victoria 0:56

Thank you. Thank you, julian. I'm good, I'm glad to be here, Very excited.

Julian 1:01

And I believe this is your first podcast.

Victoria 1:04

It is. I'm a little nervous, but you know what? We'll? Just we'll go with the flow, see what happens. Yeah, very keen, very keen to be here in the presence of a great podcast leader in the world in the field.

Julian 1:18

Where.

Victoria 1:18

Wherever he is, I'm not sure if he's asked.

Julian 1:21

All right, no, thank you. I'm stoked to have you here. This is a very everyone listening and watching. This is a very special podcast to me, this episode because victoria is actually my cousin technically she's my second cousin, uh, but my cousin nonetheless and one of the reasons I wanted to have her on here aside from the fact that she's just awesome and extremely intelligent way more intelligent than me but it's because she uses Python. So we were actually at her birthday party last weekend and we got into a great conversation about how she's using Python at university and beyond and all the different things she's doing with it, and I said you have to come on the podcast. So here we are following through on that, but I don't want to steal your thunder. Victoria, do you want to do on the podcast? So here we are following through on that, but I don't want to steal your thunder. Victoria, do you want to do a quick introduction for everyone who doesn't know you and just tell us about yourself before we jump in?

Victoria 2:13

Yes, yeah, thank you. So, yeah. So, as Julian mentioned, I am first and foremost his cousin. No, no, but I. So a bit about me. I'm a university student currently. I do biomedical engineering and computer science. I study at University of Sydney and, yeah, I'm. My passion really is in medical technology. I'm all about bringing technological innovation into hospitals you know, private medical centers, things like AI, all these new electronics, things like that. I'm really all about bringing that into the workforce and that's why I chose the degrees I'm doing, and I think Python is a great way to implement that. So I've been having a lot of fun programming inside of uni, outside of uni at work and, yeah, so keen to share about some of the stuff I've been doing is.

Julian 3:05

That's super cool. You know, it kills me sometimes. I'm like the technology that, um, people in university and high school have access to now. God, I would. I would have loved to have that when I was back, when I was your age, just a youngin, just a young whippersnapper. It would have been cool. So it's so cool that you can tie your you know curiosity for the world that we live in your scientific mind and all that stuff, and you can tie it together with the technology that you have at your fingertips is eager to build and build, change and explore and discover, and you have those tools on your laptop in your own house. You know it's, it's super cool, so, all right.

Julian 3:52

So let's, let's dive into this. Um, now you let's start with the, the university piece, right? You've, you've built some incredible stuff through university. Um, now, everyone, this is going to be pretty technical, so pay attention while you're driving your cars and sitting there and, you know, focus. All right, because I'm learning along with you. I read through some notes Victoria sent me before and it went straight over my head. So why don't we start with the first project? It was the blood oximeter, right?

Victoria 4:21

Yes, yeah, pulse oximeter.

Julian 4:23

Pulse oximeter See, I even got the name wrong.

Victoria 4:26

They're close enough. It works with the blood component with that.

Julian 4:31

That's all right. It's right there, so you go go for it. What's it about?

Victoria 4:35

Yeah, so you'll see these little devices a lot in hospitals, so they're the things that will, like you, kind of clip them onto your finger or sometimes your toe, um, but basically they're just measuring your heart rate, your blood oxygen levels, uh, measurements like that. Those are the kind of two main ones, and so for I'm in my fourth year now, so two years ago we had to build one um as a university project, uh, with a few others in a team, and yeah, so we had to build both the hardware and software for this device. We had to kind of start from scratch. So, you know, we got the microcontrollers out. I believe we used an ESP32. For those who are, you know, big hardware firmware lovers, you know ESP32 is really good.

Victoria 5:24

And so, yeah, we coded up the hardware and see, we had to um, do a bunch of signal processing with the data we were collecting from our sensor. We, we did all that signal processing in python and we, we got to make a really pretty uh gui um with a cool python packet, cool python api, um I forget the name, I think it's like pi gui or something like with a cool Python API. I forget the name, I think it's like PyGUI or something like that, but it was really. Yeah, it was really cool and yeah, that was basically the project. We were putting together sensors, hardware communicating with it in Python over Bluetooth serial, and, yeah, in the end we got a good mark. I will say, though, got a good mark, but we stayed. I will say, though, like funny story, we stayed up till like 3am the night before of the demo at uni, so the demo was at 9am. I was in the lab at like 3am the night before making sure everything worked, but it paid off.

Julian 6:18

Wow, yeah, that's, that's super cool. So this the, just for my information, with this project. Did you like? What made you choose this specific device to create?

Victoria 6:32

yeah, so well we were. Actually, we all had to make the same device, so our for that subject. Yeah, our teachers were like oh, we want you to all make a pulse oximeter, but the way at which we approached it, technically, I mean, and like what signal processing techniques would pick, etc. That was all up to us. So there was little room for creativity so that's.

Julian 6:55

That's really cool. Um, coming into this and this, this is where I find this really interesting. So you've this. This was part of which degree uh, my biomedical engineering degree.

Julian 7:06

This is the biomedical one. Um, now, coming into this, did you have experience with technology to that to that level, like micro? You're talking about microcontrollers here, python, bluetooth modules. You would have to write your own functions there to some extent. There's a lot of stuff to learn here, let alone managing the actual medical side of it. You're talking building something from scratch, coding it, making sense of the data that the sensors are bringing in. There's a whole it's almost like a full stack app right of stuff that you have to figure out. So what was your experience before you jumped into this? Did you have any experience with this stuff, or was this learning it all from scratch?

Victoria 7:54

yeah, so. So I was in my second year at the time. Before that subject, I had never worked with software and hardware together, if that made sense. So I had, yeah. So in my first year I did a introduction to python course. That's what I had up my sleeve at that point and I, um, and I had, I knew a little bit of c from first year as well, but they were they were very separate concepts at the start, right like that.

Victoria 8:22

The whole subject with the pulse oximeter was about bringing things together, bring it like watching like separate parts become like a full stack, like in my opinions, which was really cool. Um, but yeah, I at that point all I had was the python knowledge and I knew some object-oriented programming for my computer science degree. Yeah, um, and so I guess, yeah, so when I was building my gui and all the signal processing stuff, I actually did it in an object-oriented framework. So I had all these, you know, but but that was all for my comp sci degree.

Victoria 8:57

A lot of people I was studying with weren't doing computer science and I was thinking about it. They the, the of, I guess knowledge of that programming, especially in the C part, was quite different, like I felt that there was a bit of a gap and I kind of felt bad for those who hadn't been experienced in computer science. I was lucky enough that I was, but some of them didn't know what classes and functions fully were or how to implement them for this project. So I'm lucky in the fact that I had that experience.

Julian 9:26

Okay, so that's the catch, right? I mean, there's a huge learning curve here and I'll say fortunate. But I also think you're slightly crazy for doing both biomedical and comp sci at the same time. Yeah.

Victoria 9:44

I mean, I tortured myself a little with that one.

Julian 9:47

So okay, so you had that experience, but you still, like there was still a lot of learning involved here, right, bringing that stuff over, integrating it all together, was there a process that you went through to figure things out? So I guess, and I think, thinking about our audience people watching and listening to this, right, a lot of people are interested in building stuff and let's just very generically say this is kind of like an IoT project, right? Yeah, it's not connected to the network and whatever, but still you're working with a little device. You're building something out of little microcontrollers and coding it yourself, maybe using whatever version of Python, micropython, whatever it might be right, but you're putting something together to make a little device that performs a task, right?

Julian 10:38

So what did you do? What process did you go to to learn that? Did you go to to learn that and I'm phrasing this question in such a way that I want to inspire people who want to do these kinds of things, maybe just in their own house with a raspberry pi, or pick up arduino or whatever, it is right. So what did you do to kind of learn what you had to to get this over the line?

Medical Sensor Project With Python

Victoria 10:58

yeah, that's a good question, I think so our teachers were really helpful in the fact that they encouraged us to build system diagrams, just like to visually see first where's the information flow going to go from, so from one point in your code to another. Like just visually map out where are things going, because you know there's so many connections and flows of information. Map out where are things going because you know there's so many connections and flows of information and like we physically I think I I sent it in the paper to you we had a system diagram, all these different boxes and you know, uh, like separating the hardware software. This part of the hardware is going to go to this part of your code. Mapping all that out was really helpful for me.

Julian 11:43

Yep.

Victoria 11:44

So yeah, system block diagrams. Whenever I build a big system now for a project, I always draw one up and I refine it as I'm building, because sometimes, you know, things change the way you implement things.

Julian 11:57

Well, as you're learning and things break and you're like, oh, actually, I need to implement something else in between, I've got to redo this diagram right. It's actually really cool that you mentioned that. I didn't know you were going to say that. So, by the way, this is completely ad hoc, this interview. We haven't discussed anything in advance. With our coaching program, the very first thing we do is we get people to essentially mind map their entire project so we can visualize the data flow, and it ends up being a system diagram in a way, which is pretty cool. So that's awesome. And so when it comes to the Python side of it and I hate to push on the Python, but this is a Python podcast, so what was some of the highlights with Python in this project? You said use Python to create the GUI, the GUI, but you also had to interact with a Bluetooth module, right, yeah? So was all this written in Python? Everything?

Victoria 12:51

Yeah, so everything that was not on the hardware was written in Python. So, as I mentioned, all my Python code was written in an object-oriented type of framework. And for me, I did that because at that point in time I had been doing it a lot at uni and I found it to be quite efficient. So I had a Bluetooth module which would connect over serial to the microcontroller. That class would interact with my GUI module and all those GUI modules had those functions to plot the graphs in real time. Yeah, so in that way Python was really good because everything was really encapsulated well. And yeah, plotting and signal processing in Python there's a lot of good packages. For example, if you want to do an FFT in Python, a lot of good packages to do that. So it was really helpful, that's cool, nice.

Julian 13:47

So with the overall project. Now I want to move on to the next project, but with this one, what was? If you can share it, I don't know if you're allowed to talk about this stuff. What was the end result? I mean, you obviously got a grade for it, but does this get used somewhere, or is it just purely for the sake of this project? And now it's done.

Victoria 14:10

Yeah. So our team and I, we got a high distinction for the project, which was really really good to see all the hard work paid off. It hasn't been used for anything else because it was just a group project, yeah, and it kind of gets done every year. So I don't think we were technically allowed to remove it out of our year. So I could maybe publish the results online or something. I don't know if I pay people to no.

Julian 14:40

Fair enough, that's fair. Well, what I like is these projects that you've built and the things that you're working on are getting progressively way more advanced. So that was the I want to say, the simplest out of the three things that you've shown me that we'll dive into. So the second one I don't even know how to pronounce it, probably, but anyway, it's for sleep apnea, right? So what? What was tell us about this second project that you built?

Victoria 15:07

yeah, so this one was much more recent. This was, uh, another sensor that I built with a larger team at uni, so six others last year in about october, and so that that medical sensor it's a it's a thoracic pressure sensor. So, for context, your thorax is kind of a region in your body around your chest area, including your lungs, your heart. I don't know if you can really see here, but it's kind of like this area here. Okay, so that's your thoracic cavity and so a bit of medical knowledge, because I'm an enthusiast and I will sprinkle it here and there.

Victoria 15:47

But your thoracic cavity, as you're breathing in and out, it's going to have a bunch of pressure changes inside, like air pressure, naturally, as you inhale and exhale. But with people with sleep apnea those pressure changes are much more drastic, right, because if you're sleeping at night and all of a sudden your airway is cut off temporarily, you're going to have this big surge of air pressure in and out when it comes back, and that's actually really not good for the heart. So the reason so, for this project we actually collaborated with a clinical school at UCEDD so someone that was separate to like this academic learning, and they said that there is actually no like mainstream way to detect these pressure swings in the thoracic cavity um, not invasively so. This sensor is was something that we designed to just put on the skin, not, uh, invasively. The the most common way is to like stick a tube down the throat. That's a bit too uncomfortable, like no one wants to go through that overnight for a sleep study.

Julian 16:51

Oh my gosh, yeah, that sounds terrible.

Victoria 16:54

Yeah, no, it's pretty invasive, like we were looking at the pictures of what we were working with initially and, yeah, they put like a whole balloon down there into a diaphragm and that's how they measure the pressure. Yeah, you guys can search that up if you'd like. But that was the motivation to build a sensor that was able to record these pressure swings non-invasively, just on the skin. And yeah, we wanted to measure them because they actually have a big effect on the heart. So if you have really drastic we call them swings like very big jumps, that's often a sign early on of early onset like cardiovascular disease, because there's a lot of strain on the heart during those episodes. So we built the sense of three months.

Victoria 17:43

Again, we started from scratch and my role was I did a lot of the electrical stuff this time, cool yeah. So we built a PCB actually for the sensor. I helped design the circuitry, again working with some microcontrollers as well. How how our device worked was we were measuring the impedance. What we did was we, we actually put electrodes on the body, so kind of like around here. Yeah, we, we built a circuit. It would actually sit on a watch, um, on your wrist, yeah, and we'd have cords from the, from the watch, going up your arm to these electrodes.

Julian 18:25

So we'd shoot in For the audio listeners that you're talking about your chest. You'd put electrodes on your chest.

Victoria 18:31

Yeah, on our chest, okay yeah, and we'd shoot in roughly around 10 microamps of current into your body. So that's a very small amount, like given what is like currently out there. And what your body does is it actually brings back out a signal and that signal will go back down the cords, into our watch, into our circuitry and from there we can actually measure the impedance of your thoracic cavity. So, for those who don't know, impedance is like electrical resistance inside your cavity and that actually correlates well with thoracic pressure. So that's the research that we did, and so this is where the python came in, like, yeah, also, let me know if I, if I should explain things a bit.

Julian 19:17

No, no, that's okay, believe me, I will cut you up. No, I'm kidding, I I'll. I'll remind you if there's something that you know you need to explain. But no, please keep going.

Victoria 19:25

This is awesome yeah, so so where the python came in was so we had the hardware that was setting in a signal uh, sorry, setting in like electricity. You had an electrical signal coming back out of these electrodes, going back into our circuitry. What we did was Python would take that signal, it would signal process. We'd apply some sort of filter, usually, and um, and then maybe some thresholding to extract respiratory pressure information from this signal, and so from there we were able to track what the pressure is looking like inside the thoracic cavity. So we were able to yeah, like we were able to, we'd have the original impedance signal, so the one that came raw out from the chest, and then we'd have the filtered signal and then you can see where like a swing episode happened, and it was.

Victoria 20:23

It was really cool to see. We used patient data from online. First we wanted to test it, not on a person first, just in case there was. You know, we built, we designed it so that there would not be any electrocution possible, but just in case. Obviously, you know we don't want the uni to be mad at us, so yeah, yeah, fair enough, but it worked and we were able to capture these swings and yeah, it was.

Victoria 20:49

It was really fun to build that's really cool.

Julian 20:52

So what a cool project. First of all that, and it's solving a real need. That's what I love about it. You're solving a real world problem. Right, it's the first project that we talked about first is cool, but you know that stuff exists, this doesn't exist, so you, you could make some real impact and really help people who just can't go through the tube experience and try and sleep an entire night with that balloon in there and stuff. So that is very impressive, and so the question is from me then, as a follow up to that, is where did Python come into? I know you've already mentioned a little bit, but are you able to go into the depth a little bit more about the Python usage in there? Just for the people who criticize me for not talking Python on my own podcast.

Victoria 21:36

Ye ah, of course. What do you? have.

Implementing Python in Biomedical Engineering

Victoria 21:39

Yeah, so Python played a big role in this as well. So, again, so the signal processing sorry, the detection and the signal processing of that original signal we get from the body, that was all done in Python. And the actual again kind of like the last project, the Bluetooth module that was connecting, for example, our Python code to the microcontrollers that were in our circuit. That was all written in Python as well. So we used again a Bluetooth API to connect to our microcontroller, and that was a Python based one. So, yeah, again, we all the statistical algorithms, any data analysis, we did as well, we did in Python of our results, and we again we built another GUI, pretty similar to last time, and it had like a data log where you could record events like overnight, because the idea was it was used for an overnight sleep study, and so that whole user interface was done in Python as well that's cool, yeah, so look.

Julian 22:44

so here's the thing. This and I really appreciate that you're simplifying a lot of this, for me mostly, but also for the audience. Yeah, it's one thing to say something like we created a GUI, we interacted with the Bluetooth module, we used data visualization to visualize some of these signals and these reports, right, that to me is and I appreciate this, it's a high level almost like pseudocode way of saying it. Yeah, we'll visualize the data when it comes out.

Julian 23:20

But the actual implementation of doing that, you have to learn all sorts of different libraries. You have to import whatever the Bluetooth module is of a library, of a PyPI, and figure it out. I mean, you've got to read the docs, you've got to do that. So there were a lot of different libraries there that you probably had to interact with for the very first time, maybe pulling on some knowledge from the first project. But let's just say you were doing it for the first time, right, how did you learn those? Like you go okay, you have to Google, what library do I use to solve this problem? And then you go from there. What's your process?

Victoria 23:58

Yeah, good question. So yeah, I definitely worked with a lot of libraries I hadn't heard of before, sorry, and that was especially for things like actually the, the filtering of the data. So when I mean like filtering of data, often we apply like mathematical kind of filters and things like mathematical equations and translations to a signal. So those are actually. There are Python APIs that do that, and I didn't realize this because I used to do it manually in Python, write out the equations myself, and it would take a while.

Victoria 24:38

No, like, when I found out these APIs existed, I was like wow and cut down my work from no, it cut down my work from like four days to, or maybe like three days to like an hour, which is really a big difference. So I what I did was I would actually just look up on Google Python API to implement a Butterworth filter, for example. So they actually have like an API called Butter Filter, python or something, and you just run the API on your signal and then, bam, you get a result. Um, that's so Google. Google is a big help. Documentation. Um, yeah, like, I read a lot of Python API documentation and I didn't use all of it. So you know, sometimes you have to experiment. I downloaded a package if it just wasn't working or if it was just outdated, and sometimes, yeah, I got a hold of really outdated packages. You just move on.

Victoria 25:37

It takes a lot of work to, I guess, research and see what works best for you, but you know you do it once and then for the next project that you do, you kind of have that stored, yeah. So you know like, so now for the next project I'll do. I have those tools now. So in my opinion it's worth putting in the research and a bit of that extra work at the start so that you know pays off dividends in the future.

Julian 26:03

Nice, I love hearing that, because you know it's so easy to get stuck into just. We say this all the time and some of our listeners are probably groaning right now oh, here we go, julian. It's so easy to get stuck into just watching videos and reading books, and the real work and success and progress comes in when you actually implement right. So I love that you were just experimenting. Let me try this, let me try this, Let me try this library. Let me try that one. See what works.

Julian 26:31

That's another big problem as well, which we won't get into now, but you can find a lot of stuff that's out of date, that the maintainers have moved on from that package and gone. Yeah, no, I don't get paid to do this, so I've got no time, but yeah, that's one of the real challenges. I think that people get stuck in when they try to build stuff like what you're doing maybe not at that scale, but still when they're building, they get stuck in the consumption of videos on how to, how to and here's a tutorial and here's that instead of just trying it out and experimenting and seeing what works. So this is really cool and I love this idea, and so the question for me is what's the end result of all of this? So it worked. What happened after that?

Victoria 27:19

Yeah, so again, so for the university subject aspect of it. So I received a 90%, which was really great, wow, yeah, no, it was. I'm really glad that. Yeah, thank you A lot of the hard work. These projects, you know, they do take a good few months. So all the hard work paid off and yeah, again, we were collaborating with this external clinical school and they really liked the idea and yeah, so we had some good interactions with them after, um, after the university side of it was done in terms of how we could implement this into a sleep study, which is which is really good. Um, but yeah, things are still ongoing and in process, so that's right I want to be able to say oh yeah, this is this.

Julian 28:12

Things now used across all of the hospitals in sydney or whatever.

Victoria 28:16

Right, that would be super cool oh yeah, that takes a long time to implement, though. Like half of what we learn in our degree is like how hard it is to actually just take a prototype that you build and then actually scaling it to just, you know, nationwide use and all the regulation, and oh yeah, it's a whole different ballgame, yeah.

Julian 28:36

Yeah, the medical side of it. No, keep me out of it, I don't want to deal with that. No, but that's super cool. I mean, this is the kind of stuff that gets me excited about technology and Python in general. You can, you know, with an idea you can really just put stuff together in your own living room. Essentially, I mean, yeah, okay, you created a PCB that's now we're getting up there. You need a workshop or you need all that material, but now you can even print, get PCBs made up at different companies that will print on demand. Right, it's just mind blowing what you can get done, but it's super cool that that you're able to just do this in a couple of months through university that's. That's amazing. You should be really proud.

Victoria 29:18

I love it thank you all.

Julian 29:20

Right now we're gonna move to the last one. Now this one I I couldn't make heads or tails of, because we're getting into quantum something yeah quantum, yeah quantum physics quantum physics okay yeah, say quantum mechanics. I don't know if that's the same thing, but yeah look, thank you for humoring me.

Exploring Quantum Computing With Python

Julian 29:42

I I appreciate it. You made this old man feel intelligent. So this is the stuff I love but have no idea about, and to sit there and study it I can't dedicate the time. So this to me is very exciting and interesting, but I don't understand the why.

Victoria 29:58

So let's, let's start with the what what is it that you were doing in this space, and then we'll go into the why yeah, so last year, this time last year, I was working as a data and machine learning engineering intern, so at a, at a quantum-based lab, um, so they were all about performing quantum experiments for those um for those who I'm not fully sure, it was kind of based around quantum computing and looking at how we can use different small I'm going to call them microparticles, but they're much smaller in terms of, they're very small Basically how we can turn those into something called qubits. So I can go on a long. Have you heard of qubits before?

Julian 30:47

Yeah, I have heard the term yeah.

Victoria 30:49

Yes, so if we think about a classical computer, this is kind of like the way I like to introduce the whole idea of quantum computing, because I feel like it's a big hot topic yeah, please yeah, so you know a regular computer that we use, you've got bits, so bits hold information, um, and these classical bits which kind of make up the information process of a computer, they kind of have they're binary, right, they've got two states, you've got a one or a zero, and you put all those together and the different, I guess, combinations and stuff.

Victoria 31:24

That's what makes up information inside of a computer and the way at which information is is like transmitted, is based on these, on these bits. Um, in a quantum computer what makes I guess people know that quantum computers are much more faster, like their computations, that they do are significantly faster than a classical computer. So that's because instead of classical bits we're're using qubits. Qubits are really small particles that the best way I can explain it is, instead of having finite states of like one and zero, there's kind of like an infinite's, through a process called superposition. But basically information can be stored at kind of levels in between one and zero. So the transmission of information is significantly times faster. Superposition is essentially the transfer of information between these qubits or these particles, kind of instantly. So everything is working really fast.

Victoria 32:34

But there's many different ways to make qubits. So what my lab was doing is using ions. So there are a bunch of different ways You've got. There's superconducting qubits, there's photon qubits, all these different ways to make them. So in my lab we were were using a terbium, ions. A terbiums, it's an element you can see on the periodic table um, okay yep, uh, some fun fact.

Victoria 33:00

Um, these are terbium ions. What we so? The best way to explain this is imagine a newton's cradle.

Julian 33:09

Oh, in Newton's cradle, newton's cradle Like.

Victoria 33:11

I actually, you know, do you want me to bring it down, like I have it right here on my shelf, or do you want to?

Julian 33:15

Go on, go for it Can.

Victoria 33:16

I do it.

Victoria 33:18

Let me be one second, because it's a great way to visualize what's happening. Hold on, okay. So what we did at this lab was you put the ions in a chain like this and this chain's in a, a we're in a vacuum chamber at the moment and if you shoot a laser at this chain, so if I start like hold on, if I like move this, you'll see that they're kind of like oscillating, yeah, like this, but they will like they'll individually oscillate. They're kind of like oscillating, yeah, like this, but they were like they'll individually oscillate, like that, but the whole chain's actually moving as well. So if you move, like the whole chain, you kind of got all these interactions yep and you'll see that over time they kind of slow down.

Victoria 34:04

So that's what my work was. It was characterizing how these ions kind of change frequency in this chain, because you've got like this communal chain but then individual movements and yeah, just tracking the frequency change yeah so by what are you aiming to understand by tracking that change?

Victoria 34:29

yeah. So when these ions are, we call it drifting, their frequency is drifting off. That's not a good thing. So ideally in a quantum computer you'd have qubits where their frequencies and oscillations are very controlled, and that's how the transfer of information and such is in a very controlled environment. Um, the aim is for our oscillations and everything to be controlled. We're trying to understand why this drift is occurring, like what the cause of it is. So what my project was was to create a model that was able to predict and estimate firstly what the drift is and then seeing is there something causing that? Like, is there a machine in the lab that's causing it? Like, if there's, if it's drifting at a certain pattern or frequency.

Victoria 35:20

Is there a cause for that? Okay, um, yeah, if that makes sense.

Julian 35:22

So yeah, that does that. That's super cool. Thank you for breaking down quantum computing for me.

Victoria 35:29

I was not expecting that it was a very, very yeah, it was very high. That's a very high level approach and, like even myself, there's so much you know, there's so much that I don't know and that there's so much to talk about. With quantum computing, it's very complicated.

Julian 35:44

Oh yeah, I can't even imagine. But thank you for trying and for giving us that very simplistic overview For anyone who's listening and watching, who does understand this intrinsically. Don't judge us for breaking it down to the simple level. For my brain, okay, but no. So that's really cool. And so where did Python and I'm always going to come back to the Python here so where did Python come into this measurement and tracking?

Victoria 36:13

Yeah, so basically it was a bunch of data analysis and the building of a Kalman filter. That was all done in Python.

Julian 36:23

Okay, what is the Kalman filter, if you can define that?

Victoria 36:29

Yeah, so basically it's a mathematical estimation model. A lot of it's done a lot in engineering to estimate a signal or a quantity over time. Because we know that in real life whenever we have a series of data, sometimes you know there's a little bit of error in if you have like a series of data points. What the kalman filter does is you feed it your data and it will kind of apply like a series of data points. What the kalman filter does is you feed it your data and it will kind of apply like a corrective estimate of what's meant to happen. And yeah, it kind of it. It's a filter so it kind of cleans your data a little bit and kind of tries to filter out that error that we're seeing in real life cool, okay, all right, so okay.

Julian 37:12

So back to the python and the common filter. Where did, where does this all work together?

Victoria 37:18

yeah, so I basically I coded the common filter in python and then visualized what the um, what the, what the I guess the frequency drift data would look like if it was cleaned up, if that makes sense. So what the real? I guess what the cleaned behavior of of these qubits would do it? Um? And for those who've seen the kalman filter before, there's a lot of mathematics behind it. There's a lot of matrix multiplication and transposition and everything like that. You have to multiply all these matrices together to get an end result and that's like your predicted data point versus your actual data point that you give the the filter, and so all that mathematics I basically coded in python and all the algorithms and then you put it all together and you get a new series of data points and you plot it and you see, I guess, what the cleaned version is.

Julian 38:20

Yeah that's cool. Well, what kind of what Python libraries and modules did you use for this?

Victoria 38:29

yeah, so, um, I use NumPpy, of course. So, for all the matrices, applications always need numpy. Um, nothing else really, because I I built it, the filter, from scratch, but they, there are, I think one or two apis in python. That's a, it's like a common filter, one, which is cool.

Julian 38:50

Yeah, wow this is cool. So first, all this is so impressive. I almost lost for words at how and you know, everyone knows I don't stop talking. So I'm incredibly lost for words at just at your age not to put age on anything but just how incredible this stuff is right, thank you. Calling it stuff is an understatement, right and doing it a disservice.

Julian 39:20

But the thing that gets to me about all this is that it doesn't matter how advanced and technical and almost out of this world the stuff you're working on is, the fundamentals are the same. They're still there. You're still using NumPy, which is a library that people use to do all sorts of number crunching. You know financial data, whatever it is, and people use that at home. So it's when you hear that that same library that I might use on my computer at home to do some calculations, different things like that data gathering, manipulation, whatever that is also being used at your level of. I don't even know how to describe how you're using it, but in quantum physics it shows that. I hope people can be inspired as they hear this and go. I hope people can be inspired as they hear this and go.

Julian 40:13

Understanding the fundamentals can get me anywhere and investing time and learning and building and experimenting and just playing with the stuff and having, more than anything, the drive to get up and do it is what's going to get you there. You're using the same building blocks, the same pieces, the same Python concepts of. You mentioned OOP, object-oriented programming you mentioned. You know you're dividing stuff into functions and classes and you know you'll learn some best practices. You know one function one use these kinds of. You know says you know one function one use these kinds of. You know laws and rules of Python. They still apply to things of this scale and this intensity. So I guess, to wrap this up because, by the way, this is incredible, actually, no, before I wrap it up, I just want to ask you how long were you working at this quantum lab that you were doing this work in?

Victoria 41:13

So I was working there for seven or eight months, yeah.

Julian 41:17

Okay, but you were originally there for a short amount, right?

Victoria 41:21

Yeah, so it started off as an internship for six weeks over the summer and they really liked the direction of the work and so we kept working on this model for six, seven months, which was crazy, yeah such a good experience.

Julian 41:36

I love it. They appreciated the work you were doing. After just six weeks they said stay for almost a year yeah, great it was, crazy it was.

Victoria 41:45

I learned a lot, learned a lot from them.

Julian 41:47

So and is there any like now? Now, this was outside of university, right.

Victoria 41:53

Yeah.

Julian 41:54

So what's the outcome of the work? Does this get taken anywhere? Are they going to continue the work now that you're not there anymore?

Victoria 42:02

What's the go? Yeah, so I believe they're either still continuing the work or there may be talks to eventually using this as part of a publishing, like a published paper. I'm not sure that was mentioned before I left. Okay. But I think they're still working on it. It would be very cool if it was part of a paper, but we'll let you know if anything happens.

Julian 42:24

We'll get a petition out there. We'll go knock on their door and be Oi, put this in a. No, I'm just kidding. Yeah, we can make it happen. We'll make it happen. Send them this podcast episode. They'll get a.

Julian 42:41

Have a nice little shout out yeah, yeah, yeah, nice, no, no, that's really cool. So I'm what inspires me beyond the technical and, um, I hope I haven't done your technical background here any disservice. Right, because you've worked your butt off for this stuff, as I know personally over the years. Right, so you've worked so incredibly hard to get to this point of understanding, of knowledge, of capability, you know. And, that said, there are some key traits that you would be deploying or you'd be utilizing, you'd be working on, no matter what you tackle, right, such as just getting in there and doing the work, just making the choice to have your morning coffee and get to work, right, yeah, so what advice would you give to someone else who is interested in this stuff and wants to get into the building, wants to get into exploring these kinds of interests of theirs and getting into all this? You know how do you push through and just just do this stuff. You know it's not easy.

Victoria 43:47

Yeah, no, it's not, and I think I don't know if what I'm going to say is going to sound cheesy to people, but I think I think what got me through is that nothing is impossible. Um, like just saying that to myself, because, really, like you can do anything as long as you put the work and the effort into it. I don't believe that you know there's something that you know. If everyone says believe that, you know there's something that you know. If everyone says, oh, that can't be done, don't let that be the thing that stops you.

Victoria 44:17

Um, yeah, like, for example, again, like with the common filter that I built, I had people say to me, like other students, like other friends of mine who had been doing it in class oh, it took me like a month to build that, like it. It's really hard. You know you're not going to get it first try or second try and you'll. You'll come across people like that, and sometimes that person is yourself, but you can't let that demotivate you from trying. Um, that would be my biggest form of advice and I think as well, like, I feel like imposter syndrome is something that's quite big in the in the programming community. I know, know, I felt it I felt. I felt it at work whenever I work with a bunch of engineers or something. Don't let imposter syndrome as well be the thing that holds you back. That's cool.

Victoria 45:05

Yeah.

Julian 45:06

I'm so glad you said that we, we, that's one of our big, big four things that we always have to tell people. You know, drop the imposter syndrome and just do it, you know. So, out of curiosity, how long did it take you to make the Kalman filter?

Victoria 45:21

Yeah, so my original Kalman filter took me two days to build, which is which isn't like cause. When you're writing out the math and stuff, it doesn't actually take that long to build. What took me long? What took me seven? To be fair, it took seven months because I wasn't working full time, I was working very limited hours a week. But the mathematical model and all the research I did, which was kind of very much off script in the fact that a lot of the maths I kind of the models I made up on my own to pair with this filter, um, that's what took the time and even with that, like that's something I had never done before but I just because I hadn't done it before and it, I didn't let that stop me from doing it. I'm all about giving things a go. Even if you don't get it first try. Don't let that demotivate you.

Julian 46:12

Yeah, yeah, that's cool. No, I love it. I love hearing this, this motivation thing, because that's ultimately what it comes down to. Right, you can learn the tech, you can be as technical as you can, but if you don't have the motivation or the confidence in yourself and just the willpower to get up and start coding, um, you're going to struggle.

Julian 46:30

So last question from me before we cause I know this is a long episode for you and you're probably like well, just let me go, julia, I've been having fun. Yeah, good, excellent, my brain's melted by the end of this, but no. So the question I have for you, then, is and this is more of a modern day problem but there are so many distractions around us right now with, you know, your mobile phone, with social media, with TV, with and I won't say friends are a distraction. But when you're pushing hard on building something, sometimes too much socializing can actually slow you down, like I'm very I can. When we were building Pie Bites, starting from scratch, there was a good, solid period of time where we were like no, we have to sacrifice going out and doing things because we need to be here coding and building and creating. So how do you balance all of that and and have you found that to be a challenge at all?

Victoria 47:29

yeah, I think, uh, it's been a challenge at the start to balance, you know, uni with work and then like socializing and then sports.

Victoria 47:39

I think I work well with timetables, as if I will kind of timetable my time to to which is it's quite, um, I don't know, I guess I don't know if you call it robotic in a way, but like sometimes I just block out a day to work on this problem or this task and I feel like it's kind of a habit for me now to just kind of stick to it.

Victoria 48:02

It helps me a lot because I get distracted really easily. I go on my phone so often, honestly, I turn off my phone and kind of just put it out of arm's reach whenever I'm coding really hard on a problem or like I just have to really lock in with like the maths or something. But I think, yeah, I take breaks as well, breaks especially when I you know there's been times where I'm debugging a problem for over 12 hours or debugging a piece of code. It's oh, debugging is, you know it's. It can be a good thing, but sometimes it really just drains you. So taking good breaks when you're not getting anywhere sometimes the best thing yeah, walking away from the problem, yeah, yeah, it's cool.

Julian 48:42

We we talk about these kinds of these concepts and techniques a lot, um, through the podcast, through our writing. You know coaching, all that stuff, and it's really interesting to me to hear this from you, someone that I've watched grown up and who is now doing incredible things, but you still use the same techniques that we use for other things, right, and so you know. So what you're talking about. With having a timetable, we call that time blocking, right, so we get our calendar and we block out times for specific things. I did this in the corporate world, I do it now between, you know, 11am and midday every day. That's my workout time and you're encroaching in my workout time because it's 11 o'clock, um, no, but then there's also, you know, nine to 10am is deep work and the phone goes away. My computer is set to go on, do not disturb. And just like you, turning off the phone, putting it on the side. It's something we constantly tell people. Turn off your notifications, put your phone away, do not disturb. So this is really cool.

Julian 49:47

I love hearing that. And walking away, right, we promote walking away because your mind gets into diffuse mode when you go for a walk down the street, you go for a shower, you go do the dishes, whatever it might be that we do in our day to day. That's when the things connect in your head the ideas, your memories, the stuff you know, sleep on it. So I love that. That is so cool. And you didn't even do any of my coaching and you've learned all that stuff. That's incredible. Oh, don't love that. That is so cool.

Victoria 50:13

And you didn't even do any of my coaching and you've learned all that stuff. That's incredible. Oh, don't say that. Don't say that to your viewers.

Julian 50:17

That's right, we talk about this stuff all the time. So, victoria, that is so inspiring, the stuff that you're doing, and I can't wait till you're, you know, walking on the moon or Mars or wherever, right, yeah, stuff is crazy. So you're going to change the world, I know it.

Victoria 50:39

Oh, thanks Kaz, that means a lot, no that's all right.

Julian 50:44

So you read. We always ask a question as we end the podcast. What are you reading? So talk to me.

Victoria 50:55

Yeah, so I do a lot of reading, a lot of research papers actually just on like different medical topics. Um, I read a lot about neuroscience actually, like how you know, you know breakthrough developments in like mental health treatment and you know how genes affect how your brain grows and what you're capable of when you're an adult, how genes at your birth you know are capable for that like all that cool research stuff I love.

Julian 51:21

So that's awesome yeah, we should. We should do a regular sync. I want to learn all this stuff for me. I love hearing having these kinds of conversations.

Victoria 51:28

It's great yeah, no, it's good, it's. I'm, I'm all about, you know, just continuously learning, um, and I never really want to stop learning, so always have a paper nearby, yeah.

Julian 51:42

There's good scientific papers.

Victoria 51:44

Nah, I sound like a nerd, don't I?

Julian 51:48

I can just picture it. Now. I grabbed a morning coffee and we're all picking up novels and other technical books. I just grabbed a 5,000-page scientific paper on neuroscience. That's awesome. I love hearing it. That's very, very cool. Well, I'm reading and I did show you this before the podcast recording.

Julian 52:08

I've just picked up this book by Yuval Noah Harari, his new book from September last year called Nexus A Brief History of Information Networks from the Stone Age to AI. So it's highly recommended by Bob and one of our community members, craig. When I met with him the other night he was like you've got to read this, mike, in his strong UK-British accent. So British-New Zealander-K kiwi hybrid, british new zealand interesting, yeah, he's special, like that. It's cool. But, um, no, I love, I love you, craig. So I'm reading that. I I've got to open the first page because I only just bought it yesterday, but I'm very excited to dive into it because I loved his other books. So, um, you're gonna have to. You're gonna have to read it too, and we can read it together after you finish reading it, I'll I'll take it from you borrow mine.

Julian 53:02

Get your own, okay, all right, oh, excellent, all right. Well, look, victoria, it was. You tried to take it. It was such a pleasure having you on here and you're doing such cool stuff. And this is you haven't even finished uni yet. You know, I just I can't wait. You're going to do so many cool things, and is there anything else you want to add before we drop?

Victoria 53:28

I just want to say thank you so much for having me. You know, this is, you know, first podcast. This has been a great experience. Um, oh, that's cool.

Julian 53:35

Yeah, really really appreciate it oh, my pleasure, happy to do it. Sorry we didn't do it sooner. This has been. We should have done this years ago. This is really cool. Um, and if people, when people want to follow you and and say hello and whatever, where can they find you? I know you're on linkedin, and is that it or anywhere else you want to?

Victoria 53:55

yeah, I, I got instagram, facebook. Um, I can happy to put the handles up somewhere in the editing maybe not the personal ones oh, okay, then just linkedin, just linkedin to the professional stuff yeah, just my linkedin handle is there okay, perfect I'll put that in there yeah, yeah, victoria's doing great stuff.

Julian 54:21

Please, you know, follow her and, um, you know, give her, give her some motivation and and cheer her on as she goes through this, this stuff, this is great. So we're going to end it there, victoria. Thank you again. Pleasure having you on the podcast. Everyone, thank you for tuning in, for listening, for watching and make sure you share this out with people, because, again, very special episode for me personally and I definitely want to get people inspired and show them the cool stuff that Victoria is doing, that anything is possible and, yeah, we'll see you next week. Thanks for everyone for joining. Catch you later.

Victoria 54:56

Thanks, guys See ya.

Julian 54:58

Hey everyone. Thanks for tuning into the PyBytes podcast. I really hope you enjoyed it. A quick message from me and Bob before you go To get the most out of your experience with PyBytes including learning more Python, engaging with other developers, learning about our guests, discussing these podcast episodes, and much, much more please join our community at pybytescircleso. The link is on the screen if you're watching this on YouTube and it's in the show notes for everyone else. When you join, make sure you introduce yourself. Engage with myself and Bob and the many other developers in the community. Make sure you introduce yourself. Engage with myself and Bob and the many other developers in the community. It's one of the greatest things you can do to expand your knowledge and reach and network as a Python developer. We'll see you in the next episode and we will see you in the community. We will see you in the community.