We're committed to growing our content collection to provide examples and demonstrating the ways Deephaven can empower your use cases. Read on to learn more about our new blog articles, documentation updates, and videos.
This month we debuted our new blog design. The list of tags at the top help you find content on topics that interest you, as well as making it easier to follow ongoing series.
- We introduced our high-performance CSV reader with a write-up about our motivation and methodology, including some seriously impressive performance benchmarks.
- Data platforms usually provide built-in integrations with popular data formats. But what about your custom dynamic data sources? Our post on input tables shows you how to use our flexible and easy API to add data your own way.
AI / ML
Last month, we introduced our Deephaven Learn library. JJ Brosnan follows up with examples showing Learn in action:
- Real-time classification with Deephaven and SciKit-Learn: a machine learning model that can analyze diabetes risk based on health factors.
- Real-time sentiment analysis using an LSTM network in TensorFlow: a basic sentiment analysis within the Deephaven framework deployed on a (simulated) real-time feed of tokenized Twitter data.
We published several updates to ongoing blog series:
- Visualize your crypto assets with OLHC plots. Learn about Deephaven's dynamic plotting tools - as your data changes in real-time, your plot changes as well.
- Aggregate Podcast metadata. Use our customizable code to ingest and aggregate data from RSS feeds into tables.
- Solve the daily WORDLE by crowdsourcing real-time tweets. In our computer science approach to the game, we figure out how to use other people's anonymous guesses to predict the correct word without knowing any letters.
Our developers shared some useful tips and tricks this month.
- While working through some code, Chip Kent encountered unexpected results in Python default arguments and lamdbas. He documents how he modified his Python queries as a result.
- Amanda Martin needed to reduce her disk usage while working with massive, streaming datasets. She discusses how to minimize storage requirements without sacrificing speed by pairing Kafka and Parquet.
Noteworthy additions to our user guide include:
- A new concept guide on Deephaven's technical building blocks.
- How to create table listeners.
- How to add custom dependencies to Python packages.
- A performance tables cheat sheet.
Subscribe to our YouTube channel to view new videos weeklys as we continue to post our weekly developer demos and learning sessions. This month we also posted new additions to our Capabilities series.