Vito Botta’s Post

Your data science notebook just became an RCE vector. Marimo — an open-source Python notebook used by data teams everywhere — had a pre-authentication remote code execution vulnerability

The "Problem-Solver" (Best for engagement) Isolation is great for tasks, but terrible for data flow. 🧱 One of the biggest hurdles for beginners in Apache Airflow is understanding how to pass information between tasks without breaking the "atomic" rule of workflow design. XComs are the bridge. Whether you’re passing a GCS URI, a row count, or a status flag, XComs keep your DAGs dynamic and intelligent. 🚀 Check out the infographic below for a quick refresher on Pushing, Pulling, and the modern TaskFlow API approach. #DataEngineering #ApacheAirflow #Python #ETL #DataPipelines #CloudComputing

Refactored my Piotroski F-Score module into a fully standalone Python component today. I removed QuantConnect dependencies and redesigned the file so it can score stocks from either: 1. Local CSV fundamentals. 2. Online financial statements (via yfinance). What’s inside: - A clean PiotroskiFactors dataclass for standardized inputs. - Core PiotroskiScore logic across all 9 F-Score signals. - Input adapters: - compute_piotroski_from_csv(...) - compute_piotroski_from_online(...) Why this matters: - Portability: run it in any Python environment. - Reusability: drop it into screening pipelines, notebooks, or APIs. - Transparency: explicit factor construction and scoring logic. - Extensibility: easy to plug into broader quant workflows. This refactor is part of a broader effort to make my quant stack platform-agnostic, testable, and production-friendly. Next step: add a simple CLI and batch scoring across a universe of tickers. If you’re working on fundamental factor models, I’d love to compare approaches for handling missing/dirty statement data across providers. #QuantFinance #AlgorithmicTrading #Python #MachineLearning #TradingSystems #DataScience #RiskManagement #TimeSeries #SoftwareEngineering

Today Development Seed published v0.1.0 of lazycogs: a flexible tool for loading STAC collections as lazy mosaicked arrays in xarray. Inspired by stackstac and odc-stac, lazycogs uses stac-geoparquet as a metadata backend and async-geotiff + obstore for all data i/o. It allows you to define the target coordinate reference system, extent, and resolution then it uses rustac to query a stac-geoparquet dataset (local or remote) to locate the assets that are relevant for spatial/temporal subsets. When you actually want to load some pixels it fetches only the bytes you need from the cloud-optimized geotiff assets. https://lnkd.in/gZZsWJpa If you work with STAC and COGs in Python, give it a try and let me know how it goes! The image below is a low-cloud 300 meter Sentinel 2 L2A mosaic from May 2025 for the Southwest US that combines assets from hundreds of STAC items.

Really excited about Henry Rodman's work on lazycogs! It's an efficient way to lazily create an xarray DataArray of a big stack of COGs Now let's see about getting this integrated with lonboard 🙂↕️

Vector databases are great, but they aren't always the right tool for complex document intelligence. 🧠📉 If you are tired of context fragmentation and untraceable LLM hallucinations, it is time to look at Vectorless RAG with Page Index. By swapping out mathematical embeddings for a reasoning-based, hierarchical document tree, you can achieve upwards of 98% accuracy on complex Q&A tasks with perfect citation traceability. I wrote a complete guide on how this architecture works, including a full Python code implementation. Read it here: https://lnkd.in/gRuXiSxK #ArtificialIntelligence #RAG #PythonDeveloper #MachineLearning #AIEngineering

QK's power comes from its best-in-class data API — designed by people who raised the bar of how financial data is consumed around the world. A single line item gets augmented with 39 additional dimensions allowing for more consistent interpretation and depth of analysis. Oh, and the data is delivered directly into R or Python with an incredibly intuitive call. Fundamentals (fully auditable) Ownership (Beneficial Owners, Institutions, Insiders) #R #python #fundamentals #api

Built a Python-based Directory Sync Tool to compare and synchronize files between two directories with reliability and control. Instead of relying only on file names or timestamps, the tool uses a combination of metadata and SHA-256 hashing to accurately detect new, modified, and missing files. Key highlights: • Recursive directory scanning with structured metadata (name, extensions, size, hash) • Efficient change detection using size-first filtering followed by hash comparison • Memory-efficient hashing using chunk-based file reading (handles large files) • Synchronization support with metadata preservation using shutil.copy2 • Safe cleanup by optionally removing extra files from the destination While building this, I focused on moving beyond a basic script and treating it like a real tool, structuring the code into clear components, improving output readability, and adding validation and error handling to make it more reliable in real use. GitHub:https://lnkd.in/gt-Ec3rF #Python #CLI #GitHubProjects #SoftwareDevelopment #LearningByBuilding #SystemsThinking

🐍 Working with data? Save this. Honest truth — I keep coming back to these commands more than I'd like to admit. In most data projects, cleaning takes up more time than the actual analysis, and having the right commands at hand makes a real difference. This Python Data Cleaning cheat sheet covers the 5 essentials I rely on constantly: ✅ Handling nulls and duplicates ✅ Quickly inspecting your dataset ✅ Renaming, converting & cleaning columns ✅ Filtering and slicing rows efficiently ✅ Merging and grouping data If you work with pandas regularly, this should always be within reach. Which of these do you use the most? 👇 #Python #DataScience #DataCleaning #Pandas #DataAnalytics