How to analyse 100s of GBs of data on your laptop with Python vaex.io/blog/how-to-an… #Python #vaex #datascience via vaex

Together with the #Rich library, Vaex can show detailed information about how it executes a computation. Next to the time spent on each step, we also see how many passes over the data it takes (in square brackets). Great for understanding and optimizing your processing pipelines!

Wow. Neato. Seeing a lot of clever uses of Rich today!

Wordify 2.0 is out! 🚀 Now faster, more interactive, and has more languages. These improvements have been possible thanks to great libraries in the ecosystem, respectively: vaex, Streamlit, and spaCy. Thanks to Federico Bianchi, @dirk_hovy, and the MilaNLP lab!

See our latest talk on Vaex with Maarten A. Breddels and Jovan

Did you know that Vaex can do multiple computations with a single pass over the data? Set “delay=True” to schedule operations to be executed in parallel. The Rich progress bars nicely indicate this with [1] and [2], showing how we aggregate the 1.1 billion rows on a single node!

Turn caching on to speed up computations in Vaex! Especially useful for dashboard backends and interactive EDA. In this example we see that the groupby operation of 1 billion rows is sped up significantly by reusing a previous result from the unique calculation.

It would be interesting to add support for the SIMD instruction sets of various POWER architectures in #xsimd, which underlies SIMD acceleration in Scipy (through Pythran), Apache Arrow, Xtensor, and more. Would folks at @ibm be interested in funding such work?

Vaex 4.8.0 is out! It includes various performance improvements, improved asyncio support, fancy progressbars and more! See the full changelog at github.com/vaexio/vaex/bl…

Is your data in the Cloud? No problem: Vaex can open and stream your data directly from your favorite cloud storage provider. Even better - it will only download the parts of the data that you view or use! Learn more about reading data from the cloud at: vaex.io/docs/example_i…

Jovan and Maarten will showcase vaex, an open-source DataFrame library in Python, tailor-made to allow fast, interactive workflows with datasets that are too large to fit in RAM on a single node. Join us live here go.mlops.community/register on Wednesday at 9am PST/ 5pm UK!

RADIS 0.12.1 was released today ; - more accurate for very thin lineshapes. - vaex is the new default HDF5 library for all platforms (faster!) - use arbitrary Astropy units when rescaling spectra - and many docs improvements But mainly : we welcome 5 new contributors !

Awesome writeup :)

Wondering how to make simple and fast Streamlit apps using large datasets? In our latest article we use combine vaex and Streamlit to do just that. We cover: - caching for efficient computations - optimized delayed evaluations - progress bars vaex.io/blog/vaex-plus…

College completely failed to teach me data analysis. So I spent over 10,000 hours learning Python. Then, I picked the 13 best libraries for machine learning and data analysis. But unlike college, these won't cost you $120,000. Here they are for free:

Vaex: "One of the 13 best libraries for machine learning and data analysis."

Did you know that vaex 's documentation contains a section with guides? Guides can help one better understand certain functionalities of Vaex, especially those that are not common in other DataFrame libraries. vaex.io/docs/guides.ht… 1/4

Nick Cerny DuckDB vaex I know duckdb but haven't seen Vaex before. Looks pretty interesting!

If you are in a hurry, read this thread, which summarizes Jovan 's latest Vaex article. If you want to read the full article: medium.com/@jovan.veljano…

Introducing Reacton 🥳 - a pure Python port of React for ipywidgets! With the proven success of ReactJS, we've implemented a similar API in Python to help create larger applications using ipywidgets. Check out my article at: maartenbreddels.medium.com/advance-your-i… #React #Python #Jupyter