My friend and I just finished what I would call the first usable version of my app, StudyBits.
TLDR; it’s a Studying Tool, and you should download the app by using the link below!
I wanted to create this because of my freshman year of high school; AP HuG had taught me the error in my studying habits, so I decided to try and design a solution. Unfortunately, once sophomore year started, I didn’t have time to finish the app, which I was developing in Ionic at the time. Luckily, #C073DTGENJ2| reminded me of this idea, it also convinced me to switch to React Native, which was much better than Ionic for my purposes.
Now the app: StudyBits helps you avoid cramming for your exams by allowing you to study on your phone. You can simply subscribe to a course, which will add a set of questions to your question pool. Then, you can answer questions from your question pool by going to the “Answer” tab in the app which will randomly select a question for you to answer. In the future, we plan to implement an algorithm which selectively chooses questions which will maximize your learning in combination with spacial repetition. You might be wondering who creates the questions, well you do! To create questions, you simply create a channel and a course with whatever units you would like. Then, you create questions in the create tab to your heart’s content. WE NEED YOU TO CREATE USEFUL COURSES! Right now, all we have is a filler brainrot course, so we encourage you to make usable courses (or courses on how to mew). In the future, we plan to use artificial intelligence to inflate our question sets. Finally, you can use our app to motivate you by competing for a spot on the leaderboard. By answering questions correctly, you gain points which rank you higher on our leaderboard, we plan to make this algorithm more complex in the future. Try to beat “The Sigma Rizzler” who is at 10,000,003 points 😉.
Testflight Link: testflight.apple.com/join/t0SncOyx
Demo video attached below!
I know right now the UX is a little inconsistent, and has room for improvement. This app is still a work in progress :loading:…
I think this is the first time I’ve “shipped” something here that’s not for Hack Club (don’t quote me on that), but I really enjoyed this personal project so I thought it would be worth sharing.
What started as a side-project for my internship became more of a personal project that I have worked on more after (since I didn’t get it done during the internship). Anyway… I’ve been working on a genetic algorithm for the stock market. It works by training on historical data downloaded from Yahoo finance using the yfinance library. Then, it extracts indicators from the data, such as RSI, MACD, and DOSC (of course, you could add more). Next, the actual model subtracts an offset from each of these indicators, and calculates a weighted sum from the offset indicators. It then puts this result into a function called a sigmoid, which returns a value from -1 to 1 and buys or sells a percentage of the stock depending on if it is positive or negative. This is a little bit of a simplification, but that’s the general gist of it.
It turns out that this makes a lot of money (see the image below which starts with $100,000). The list of tickers I used, is in the repository under tickers.txt.
Now, I really want to take this to the next level, so feel free to reach out to me or make a PR if you have any ideas. There are some problems with the simulation/model right now. For instance, it is difficult to predict how the market will behave if the bot takes an influential stake in a business, and I’m not sure how to combat that, especially since I’m using historical data. Also, the amount of money that it is making right now seems insane to me, and it’s possible that there’s a bug (even though I’ve searched the code and validated purchase a gazillion times). So please let me know if you think something is wrong. I’d also really love help on how to make the simulation faster.
Thanks, and I look forward to hearing some feedback! - Urjith Mishra
git repo: codeberg.org/TheScientist101/genetic-millionaire
github mirror: github.com/TheScientist101/genetic-millionaire
Thank you to @zrl for sending me stickers! The fire one is so cool, it’s like reflective and a different material. I humbly accept your curse to visit hw within the next 3 years. Again, thanks so much, your act of kindness made my week!
I’m filling out my application for making a Hack Club at my school and I think my Arch configuration is most meaningful to me, so I thought I’d log it :).
My friend and I just finished the UI “refresh” for most of the pages on our app, StudyBits. This update not only looks a lot better, it allows you to customize your own channel. You can download and test our app on Testflight by using the link: testflight.apple.com/join/t0SncOyx. Hopefully our build gets approved for public testing, it hasn’t yet, but it takes a couple hours. If it doesn’t DM me your email so I can add you to the testing group.
I just finished Easel for #building-programming-languages. I learned a LOT. Most importantly, how much simpler it is to break down code into functions. Idk how to do spoiler images, but this was the most intricate Rickroll I have ever participated in. I changed a very few things, so if you’re interested in looking at the code, it’s at: github.com/TheScientist101/easel.
I just made a PR for #arcade to fix the truncation on the scrapbook banking messages. Idk if it works, and it was literally 1 line of code (so I don’t expect any tickets from this), but it’s the first PR I’ve made in a while so hey!
I can’t show any pictures of it in action (since it hasn’t been accepted yet), but here’s a thread of the issue.
hackclub.slack.com/archives/C0266FRGV/p1718717652692109
My #hardware-party reflection:
This project was tons of fun and taught me a lot. I learned to work with Bluetooth, how to solder, and how to work with motors. Although I didn’t completely finish, I wouldn’t call my project a failure. As you can see in the photo, the camera and robot are separate. The camera and robot worked separately, but the chassis was smaller than I anticipated and I didn’t have room for the raspberry pi. If I were to do this again, I’d get rid of the microcontroller and use the raspberry pi to control the whole thing. I would also try to ensure all my parts are compatible. For example, my battery bank has a way lower amperage than the raspberry pi needs. This forces the raspberry pi to work very slowly and was almost unusable unless it was plugged into the wall. This also may mean I need to charge the battery pack which I actually haven’t tried and I just realized while writing this. Overall, this project was tons of fun and I’d love to do something similar in the future.
Day 10 of #hardware-party :
Today I worked on streaming the video from the raspberry pi to the phone. I ended up creating a flask server to stream the video; however, for some reason the processed video wasn’t showing in the app, but it’s probably a silly mistake. It is late now so here’s what I have. If I were to do this again, I’d probably try to get rid of the adafruit feather and control the motor with the raspberry pi for east autonomous control and less power management issues. I have attached a screenshot of the app because it’s late, I’ll try to send a video tomorrow!
Day 9 of #hardware-party:
I came home late from school and I told myself to do just one thing which was to roughly mount the raspberry pi on the car. I tried to 3D print a case but I haven’t gotten access to a 3D printer to print it. If y’all have any suggestions on how to protect the camera’s exposed ribbon cable please let me know. I also have ran into the problem of powering the Raspberry Pi. I have a power bank that is USB A; however I don’t think I have a USB A to USB C cable. I hope to find an adapter to finish this in time. I will go all in tomorrow to try to finish this in time.
Day 8 of #hardware-party :
Today I didn’t make as much progress as I hoped in terms of the project. I tried to work with ffmpeg for video streaming; however, I don’t yet fully understand the concept. I switched the camera from a webcam to the raspberry pi camera.
Day 7 of #hardware-party :
I worked on getting tensorflow to work headless today. I am using a virtual x11 display and streaming it using vnc. It was a lot of just learning how to do object detection with tensorflow, and the virtual display was pretty simple. It is a startup script now so that the object detection starts as soon as the pi gets power.Day 6 of #hardware-party :
I worked on the parsing of the input today. It led to me fixing the way I send data in Flutter from a random cursed array to a byte buffer. I also messed up the firmware of my adafruit feather twice today, but it finally drives. I am proud of myself. I use a joystick with arcade drive and send it to the feather. BTW, Adafruit’s documentation on the Bluetooth seemed nonexistent, but ChatGPT helped me work with Bluetooth on the Feather. If anyone does find documentation, please send it to me. I spent about 3 hours finding a typo today, luckily my brother wanted to see what I was working on and found it. I am really surprised with my progress today, considering the last 2 days were pretty slow. Here’s a video of the robot driving.Day 5 of #hardware-party :
Today I finally got the connectivity and data transfer between the robot and the phone for controls done. I am using Flutter for the app because of its ease of use and cross compatibility. Here’s an image of the joystick below. I need to learn how to parse the data on the robot. I will hopefully work on that today.
Day 4 of #hardware-party:
I have made some progress on flutter today; however, it still can’t connect to the device. I have a long weekend and I will work on the project as much as possible in the following days. Hopefully, I will have an app. Here is it just displaying the Bluetooth devices.
Day 3 of #hardware-party :
Today, I didn’t get as much done as I wanted to. I tried to make an app, but I was having difficulty working with bluetooth in Xcode, so I decided to that I will try to switch to Flutter tomorrow. I don’t have any photos of the project today so here’s Android Studio, so hopefully tomorrow goes better.
Day 2 of #hardware-party:
I did some soldering today (don’t mind the chicken table). It was the first time in a while. They were headers so it was tedious and took more time than I anticipated. I also burnt my pinky. The robot is drivable with Adafruit’s code, but I don’t want to send that because it’s not mine. Tomorrow I will make it driveable with my own code. I might use MIT App Inventor or XCode or Android studio, but probably XCode. I’ve never played around with Bluetooth before so it will be a new experience. In terms of hardware, I just need to mount the Raspberry Pi and its camera.
Day 1 of #hardware-party:
Today I worked on the chassis of the robot. The chassis is updated from Adafruit’s instructions so I had to get a little creative with the roller ball bearing. It’s not drivable yet, but hopefully tomorrow I’ll work with electronics and code. I have a lot of homework today, but I’m glad I could squeeze this in.
