Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Show HN: XY – A Fast, composable, GPU-accelerated interactive plotting library

github.com

Post reply on HN
Print view View original

41–50 of 50 posts

Re: Show HN: XY – A Fast, composable, GPU-accelerated interactive plotting library

#41

Post by mamcx »

kasts wrote: post #6

I’m not convinced GPU acceleration is a meaningful advantage for most charting use cases. Most dashboards don’t render enough data for it to matter. Once a chart is dense enough for rendering to become the bottleneck, it normally is already be too crowded to be meaningful. Zooming can justify supporting larger datasets, but sampling/viewport culling and level of detail often avoid drawing unnecessary points...

The good reason for worry about it is the same for data grid, list, scrolls and any other UI component that loads arbitrary data.

All UI, honestly, is only meaningful in what the screen size and our vision permit. END.

UNFORTUNATELY, you can't avoid that a user is writing "a___" and the source data has millions of things that start with `a` and all the others are dozens.

So, you can end with a massive influx of data, and sure the user see that big mess and wanna dial in, but in the meantime is nice if the UI not die in the process.

Re: Show HN: XY – A Fast, composable, GPU-accelerated interactive plotting library

#42

Post by rossant »

HoneySpoons wrote: post #22

This is awesome, can easily see this becoming a standard library. Can't wait for the 3D and volume visualizations.

See also Datoviz for 3D support in addition to fast, scalable, GPU-based 2D visualization: https://datoviz.org/ (I'm the main developer)

Re: Show HN: XY – A Fast, composable, GPU-accelerated interactive plotting library

#43

Post by all2 »

apetuskey wrote: post #20
Evidlo wrote: post #17

Earlier quoted context omitted.

I constantly have to work around the slowness of matplotlib when creating animated sequences for my scientific work (even with the Agg back end)

How many points are you working with usually?

Not GP, but my datasets weigh in at around 2-20GB depending.

Re: Show HN: XY – A Fast, composable, GPU-accelerated interactive plotting library

#45

Post by NortySpock »

kasts wrote: post #6

I’m not convinced GPU acceleration is a meaningful advantage for most charting use cases. Most dashboards don’t render enough data for it to matter. Once a chart is dense enough for rendering to become the bottleneck, it normally is already be too crowded to be meaningful. Zooming can justify supporting larger datasets, but sampling/viewport culling and level of detail often avoid drawing unnecessary points...

Why settle for sampling when you can have the whole dataset?

The spiral pattern is an excellent example. "Sure it looks like this when you zoom out, but when you zoom in, you can see the finer structure of the points..."

Re: Show HN: XY – A Fast, composable, GPU-accelerated interactive plotting library

#46

Post by FuckButtons »

Evidlo wrote: post #17
kasts wrote: post #6

I’m not convinced GPU acceleration is a meaningful advantage for most charting use cases. Most dashboards don’t render enough data for it to matter. Once a chart is dense enough for rendering to become the bottleneck, it normally is already be too crowded to be meaningful. Zooming can justify supporting larger datasets, but sampling/viewport culling and level of detail often avoid drawing unnecessary points...

I constantly have to work around the slowness of matplotlib when creating animated sequences for my scientific work (even with the Agg back end)

I too have felt that pain. Funcanimation is passible for a few datapoints but it really struggles with anything meaningful in real time.

Re: Show HN: XY – A Fast, composable, GPU-accelerated interactive plotting library

#47

Post by Eridrus »

kasts wrote: post #6

I’m not convinced GPU acceleration is a meaningful advantage for most charting use cases. Most dashboards don’t render enough data for it to matter. Once a chart is dense enough for rendering to become the bottleneck, it normally is already be too crowded to be meaningful. Zooming can justify supporting larger datasets, but sampling/viewport culling and level of detail often avoid drawing unnecessary points...

Managing sampling is itself not totally trivial. Much easier from a DevX perspective to just have a library that can render all the datapoints.

Re: Show HN: XY – A Fast, composable, GPU-accelerated interactive plotting library

#48

Post by cozzyd »

kasts wrote: post #6

I’m not convinced GPU acceleration is a meaningful advantage for most charting use cases. Most dashboards don’t render enough data for it to matter. Once a chart is dense enough for rendering to become the bottleneck, it normally is already be too crowded to be meaningful. Zooming can justify supporting larger datasets, but sampling/viewport culling and level of detail often avoid drawing unnecessary points...

don't you just love having a point cloud so dense that it's completely unreadable?

it seems a lot of people don't know about histograms...

Re: Show HN: XY – A Fast, composable, GPU-accelerated interactive plotting library

#49

Post by cycomanic »

kasts wrote: post #6

I’m not convinced GPU acceleration is a meaningful advantage for most charting use cases. Most dashboards don’t render enough data for it to matter. Once a chart is dense enough for rendering to become the bottleneck, it normally is already be too crowded to be meaningful. Zooming can justify supporting larger datasets, but sampling/viewport culling and level of detail often avoid drawing unnecessary points...

I can give you a prime example where you want to render all the data and quickly, oscilloscopes. It's quite common that you fetch traces with millions of samples, but then want to zoom into specific regions. There are lots of similar applications in experimental signal processing, where you want to have large data sets, but sampling easily will hide details that you want to see (unless you already know what exactly your data looks like).
Post reply on HN