Simon Willison’s Weblog

GPT-3 token encoder and decoder. I built an Observable notebook with an interface to encode, decode and search through GPT-3 tokens, building on top of a notebook by EJ Fox and Ian Johnson. # 27th April 2023, 11:48 pm

Introducing sqlite-vss: A SQLite Extension for Vector Search (via) This latest SQLite extension from Alex Garcia is possibly his best yet: it adds FAISS-powered vector similarity search directly to SQLite, enabling fast KNN similarity lookups against a virtual table that feels a lot like SQLite’s own built-in full text search feature. This write-up includes interactive demos using Datasette called from an Observable notebook, running similarity searches against an index of 200,000 news headlines and summaries in less than 50ms. # 10th February 2023, 10:53 pm

Spevktator: OSINT analysis tool for VK. This is a really cool project that came out of a recent Bellingcat hackathon. Spevktator takes 67,000 posts from five popular Russian news channels on VK (a popular Russian social media platform) and makes them available in Datasette, along with automated translations to English, post sharing metrics and sentiment analysis scores. This README includes some detailed analysis of the data, plus a link to an Observable notebook that implements custom visualizations against queries run directly against the Datasette instance. # 5th September 2022, 8:48 pm

Open every CSV file in a GitHub repository in Datasette Lite (via) I built an Observable notebook that accepts a GitHub repository as input, scans it for CSV files and generates a link to open all of those CSV files in Datasette Lite. # 1st September 2022, 7:24 pm

Introducing sqlite-lines—a SQLite extension for reading files line-by-line (via) Alex Garcia wrote a brilliant C module for SQLIte which adds functions (and a table-valued function) for efficiently reading newline-delimited text into SQLite. When combined with SQLite’s built-in JSON features this means you can read a huge newline-delimited JSON file into SQLite in a streaming fashion so it doesn’t exhaust memory for a large file. Alex also compiled the extension to WebAssembly, and his post here is an Observable notebook post that lets you exercise the code directly. # 30th July 2022, 7:18 pm

Datasette table diagram using Mermaid (via) Mermaid is a DSL for generating diagrams from plain text, designed to be embedded in Markdown. GitHub just added support for Mermaid to their Markdown pipeline, which inspired me to try it out. Here’s an Observable Notebook I built which uses Mermaid to visualize the relationships between Datasette tables based on their foreign keys. # 14th February 2022, 7:43 pm

GitHub Burndown (via) Neat Observable notebook by Tom MacWright—give it a GitHub access token and the name of a repo and it pulls the details of every issue and plots a burndown chart over time, showing how long issues stay open for. The code is worth spending some time with—the way it fetches data from the paginated JSON API is a really great example of using generators with Observable, and the chart itself is a lovely clear example of Observable Plot. # 10th February 2022, 4:29 pm

Observable Plot Cheatsheets (via) Beautiful new set of cheatsheets by Mike Freeman for the Observable Plot charting library. This is really top notch documentation—the cheatsheets are available as printable PDFs but the real value here is in the interactive versions of them, which include Observable-powered sliders to tweak the different examples and copy out the resulting generated code. # 25th January 2022, 10:12 pm

Datasette downloads per day (with Observable Plot) (via) I built an Observable notebook that imports PyPI package download data from datasette.io (itself scraped from pypistats.org using a scheduled GitHub Action) and plots it using Observable Plot. Datasette downloads from PyPI apparently jumped from ~800/day in May to ~4,000/day in July—would love to know why! # 17th July 2021, 5:01 pm

Plot & Vega-Lite. Useful documentation comparing the brand new Observable Plot to Vega-Lite, complete with examples of how to achieve the same thing in both libraries. # 4th May 2021, 4:32 pm

Observable Plot (via) This is huge: a brand new high-level JavaScript visualization library from Mike Bostock, the author of D3—partially inspired by Vega-Lite which I’ve used enthusiastically in the past. First impressions are that this is a big step forward for quickly building high-quality visualizations. It’s released under the ISC license which is “functionally equivalent to the BSD 2-Clause and MIT licenses”. # 4th May 2021, 4:28 pm

Render single selected county on a map (via) Another experiment at the intersection of Datasette and Observable notebooks. This one imports a full Datasette table (3,200 US counties) using streaming CSV and loads that into Observable’s new Search and Table filter widgets. Once you select a single county a second Datasette SQL query (this time retuning JSON) fetches a GeoJSON representation of that county which is then rendered as SVG using D3. # 5th April 2021, 4:48 am

Datasette Client for Observable (via) Really elegant piece of code design from Alex Garcia: DatasetteClient is a client library he built designed to work in Observable notebooks, which uses JavaScript tagged template literals to allow SQL query results to be executed against a Datasette instance and displayed as inline tables in a notebook, or used to return JSON data for further processing. His example notebook includes a neat d3 stacked area chart example built against a Datasette of congresspeople, plus examples using interactive widgets to update the Notebook. # 24th November 2020, 6:53 pm

Datasette table diagram, now with a DOT graph (via) Thomas Ballinger shared a huge improvement to my Observable notebook for rendering a diagram of a collection of Datasette tables. He showed how to use the DOT language to render a full schema digram with arrows joining together the different tables. I’ve applied his changes to my notebook. # 8th May 2020, 3:23 am

D3 Projection Comparison (via) Fun Observable notebook that lets you compare any two out of D3’s 96 (!) geographical projections of the world. # 10th March 2019, 10:58 pm

Develop Your Naturalist Superpowers with Observable Notebooks and iNaturalist (via) Natalie’s article for this year’s 24 ways advent calendar shows how you can use Observable notebooks to quickly build interactive visualizations against web APIs. She uses the iNaturalist API to show species of Nudibranchs that you might see in a given month, plus a Vega-powered graph of sightings over the course of the year. This really inspired me to think harder about how I can use Observable to solve some of my API debugging needs, and I’ve already spun up a couple of private Notebooks to exercise new APIs that I’m building at work. It’s a huge productivity boost. # 18th December 2018, 10:39 pm

I don’t like Jupyter Notebooks—a presentation by Joel Grus (via) Fascinating talk by Joel Grus at the Jupyter conference in New York. He highlights some of the drawbacks of he Jupyter way of working, including the huge confusion that can come from the ability to execute cells out of order (something Observable notebooks solve brilliantly using spreadsheet-style reactive cell associations). He also makes strong arguments that notebooks encourage a way of working that discourages people from producing stable, repeatable and well tested code. # 25th August 2018, 3:04 am

Observable Tutorial 2: Dog pictures (via) Observable have a neat new set of tutorials on how to get started with their reactive notebooks. You don’t even need to sign up for the service: they have a “Scratchpad” link in their navigation bar now which lets you spin up a test notebook with one click. # 18th August 2018, 7:55 pm

Notebook: How to build a Teachable Machine with TensorFlow.js (via) This is a really cool Observable notebook. It explains how to build image classification that runs in the browser on top of Tensorflow.js, and includes interactive demos that hook into your webcam and let you hold up items and use them to train a classifier. Since it’s built on Observable every single underlying line of source code is available to browse as part of the essay. # 20th June 2018, 9:10 pm

Changelog 2018-06-12 / Observable. The ability to download an Observable notebook as a stand-alone ES module and run it anywhere using their open source runtime is fascinating, but it’s also worth reading the changelog for some of the new clever tricks they are pulling using await—“await visibility();” in a notebook cell will cause execution to pause until the cell scrolls into view for example. # 13th June 2018, 3:50 pm

Observable: Downloading and Embedding Notebooks (via) Big news from the Observable team: firstly, they’ve released the open source runtime for their notebooks which means you can now execute the code from a notebook independently of their hosted service. On top of that they’ve constructed an elegant way of exporting and executing notebooks (or specific notebook cells) as ES6 modules and as installable npm package tarballs. # 22nd May 2018, 12:14 pm

Iodide Notebook: Project Examples (via) Iodide is a very promising looking open source JavaScript notebook project, and these examples do a great job of showing what it can do. It’s not as slick (yet) as Observable but it does run completely independently using just a browser. # 3rd May 2018, 6:42 pm

What do you mean “average”? (via) Lovely example of an interactive explorable demonstrating mode/mean/median, built as an Observable notebook using D3. # 12th April 2018, 4:41 pm

Observable notebook: San Francisco trees from Datasette. I used an Observable notebook to rebuild my San Francisco tree search demo against a Datasette API of a CSV of trees published by the SF Department of Public Works. The map updates live as you type a query, and every cell can be toggled to view the underlying source code. # 1st February 2018, 12:37 am

USGS World Earthquake Map (observable notebook). Here’s an extended version of the notebook constructed by Jeremy Ashkenas in that Observable YouTube demo. You really need to check this thing out—the notebook itself has sliders in that you can manipulate (even on a mobile browser) or you can click to edit the code and see your changes reflected in real-time. If you sign in with GitHub you can fork the project to your own account and save your changes. # 31st January 2018, 7:07 pm