The Data Science Lab is a live weekly call. Register at pos.it/dslab! Discord invites go out each week on lives calls. We'd love to have you! The Lab is an open, messy space for learning and asking questions. Think of it like pair coding with a friend or two. Learn something new, and share what you know to help others grow. On this call, Libby Heeren is joined by Joey Marshall, who walks through how he uses Claude Code for data science workflows and building interactive data applications. Joey shows us how he uses Claude Code, Wispr Flow, cmux, and the 1Password CLI to do data science in his everyday. He uses a TidyTuesday dataset about frog call observations in Australia to build an interactive web app called "Frog Finder" - complete with a cottagecore aesthetic, radius-based species search, and deployment to Vercel - all in about 30 minutes total (we ask Joey a lot of questions the rest of the time)! The dataset is publicly available, so you can try this at home, too! Hosting crew from Posit: Libby Heeren, Isabella Velasquez Joey Marshall's LinkedIn: http://linkedin.com/in/joeymarshall Resources mentioned in the video and chat: 🔗 Frog Finder App (built live during the session): https://frog-finder-iota.vercel.app/ 🔗 Australian Frogs TidyTuesday Dataset: https://github.com/rfordatascience/tidytuesday/tree/master/data/2024/2024-08-27 🔗 Claude Code: https://www.anthropic.com/claude/code 🔗 wispr flow for dictation: https://wisprflow.ai/ 🔗 cmux Terminal: https://github.com/aca/cmux 🔗 1Password CLI: https://developer.1password.com/docs/cli 🔗 Vercel: https://vercel.com/ 🔗 Positron IDE: https://positron.posit.co/ 🔗 Simon Couch's LLM Evaluation Blog: https://simonpcouch.com/blog/ 🔗 Anthropic Course Marketplace: https://www.anthropic.com/courses 🔗 Posit Data Science Hangout every Thursday: https://pos.it/dsh ► Subscribe to Our Channel Here: https://bit.ly/2TzgcOu Follow Us Here: Website: https://www.posit.co Hangout: https://pos.it/dsh The Lab: https://pos.it/dslab LinkedIn: https://www.linkedin.com/company/posit-software Bluesky: https://bsky.app/profile/posit.co Thanks for hanging out with us! 💛 Timestamps of Questions / Topics: 00:00 Introduction 02:30 "How long have you been using LLMs for your data workflows?" 04:00 "For someone who's bearish on AI, what benefit does using something like Claude Code bring?" 07:00 Starting the Claude Code demo with the frogs dataset 08:00 Setting up terminal environment and cmux 10:00 Using 1Password CLI for secrets management 12:17 Asking Claude to explore the Australian frogs dataset 13:00 Managing secrets with 1Password vs plain text environment files 17:22 "Have you done any analysis of how different generative AIs perform for the kind of work you're doing?" 18:00 Using Opus vs Sonnet vs Haiku models in Claude Code 19:20 "With your Claude prompts, what components are you intentionally saying versus what did you feel free to be more liberal about?" 21:00 "How does the computer know you're talking to the AI versus just talking?" 22:00 Reviewing Claude's three project proposals for the frog data 24:30 Going into planning mode and setting up deployment to GitHub and Vercel 26:00 "Do voice-to-text things require a lot of hardware to work fine?" 28:00 Choosing a cottagecore aesthetic and color scheme 28:30 "Do you have advice on setting up agents that are working on the same folder?" 34:00 "Is there an equivalent to the CLI 1Password functionality in Apple passwords on Mac?" 35:00 Reviewing the implementation plan from Claude's planning mode 36:30 "What does the four percent context mean?" 41:00 Skills in Claude Code and system prompts 41:30 Frog Finder app is live and deployed 42:30 Viewing the deployed Frog Finder app in the browser 45:30 "Can you switch back and forth between models during the same session?" 46:30 "What resources would you recommend to learn all of this stuff?" 48:30 Verifying where Claude got the frog images (Wikipedia) 49:00 Asking Claude to use frog emojis instead of dots on the map 51:00 "How accurately did Claude pull the images from Wikipedia?" 52:30 Frog emojis are live on the map
The information provided is not trading advice. kdj.com does not assume any responsibility for any investments made based on the information provided in this article. Cryptocurrencies are highly volatile and it is highly recommended that you invest with caution after thorough research!
If you believe that the content used on this website infringes your copyright, please contact us immediately (info@kdj.com) and we will delete it promptly.