PyGraphistry – A library to extract, transform, and visually explore big graphs | Hacker News Reader

PyGraphistry – A library to extract, transform, and visually explore big graphs

github.com

125 pointssndean27 comments

Fun to see a link here!

Always happy to describe what's happening underneath w/ connecting GPUs in the browser to GPUs in the datacenter. Likewise, the connection between event data & graph analytics is powerful as data scales, so happy to dig into that too.

Not shown there, we're piloting a 'visual playbook' investigation layer to help teams who investigate through a lot of event data. This has been especially relevant for security (SOC/IR/hunt) and anti-fraud as a team grows and needs to cover more ground. Playbooks let you finally record common multi-step multi-datasource workflows and get real visibility out of them. When your alerting flags something, running the playbook will gather & correlate that data for you, and unique to Graphistry, present it in a full visual (graph) analytics session. Think visually automating multi-step queries across Splunk + Spark + various APIs. And of course, for the advanced analysts, giant GPU-accelerated visualizations. We're actively piloting with interesting teams, so please ping info@graph....com if it may be your team's kind of thing.

Is it possible to use a local GPU from the machine that has the data? Kind of like Gephi?

Yep! And we're working towards multi-GPU boxes!

Nice! Thank you

Is it possible to visualize graphs entirely on the client side (without sending any data to your backend)? We have some very large graphs that we'd like to explore, but unfortunately it's not possible to send the data to the cloud, hence a local solution would be great. I have investigated Gephi but unfortunately the performance is quite disappointing for very large graphs.

Graphistry is impressive, but not free, and needs their server and an API key. You probably already know about graphviz sfdp (not interactive). There's been interesting recent work on distributed maxent optimization https://arxiv.org/pdf/1506.04383.pdf and sparse stress models https://arxiv.org/abs/1608.08909 (see its references as well, for example, Khoury's MARS is still available at https://github.com/marckhoury/mars). This problem requires a ton of algorithm engineering, a lot of messy user interface engineering, and integration into an HPC environment is a big issue, so progress in open source has not been that rapid.

Those projects are pretty exciting. We accelerated FA2 and are now adding acceleration to the surrounding analytics interactions. And yep, to support an eco-system here, we've been iterating on a sustainable & scalable model (... "enterprise").

To that end, we have a variety of non-profits & scientists using our cloud tier. We want to streamline that program in the coming months: If that's you today, we'd love to help, so feel free to reach out for early access. For example, we've been excited by data science and cybersecurity schools, journalists, and scientists.

Not sure how large very large is or the performance you need but you might want to check out the fairly recently added graph support in datashader (examples: https://anaconda.org/jbednar/edge_bundling/notebook)

Here's an example of the output which I then put into tiles & rendered out: https://proseandcode.co.uk/beta_gtr_viz/

I think there's 600k edges in this, all rendered.

I did some of the work on the edge bundling, the split out code is here: https://gitlab.com/ianjcalvert/edgehammer

All of this is more work than just loading it into a current program though, but it might be a useful component of what you need, hopefully a large useful part.

Get in touch if I can help out.

I'm a fan of the continuum team - we have been working together on the GPU GoAI project. We'll load 2M+ edges, and edge bundling helps make that visually useful.

Graphistry is a bit different where the result is a full interactive visual analytics session, not a zoomable png. So you get visual filtering, histogramming, search, etc. Our goal is to get from a lot of data to the answer, including whatever data pivots/cleaning/etc., as fast as possible, and that includes helping analysts skip a lot of the visualization/data coding and instead do direct visual interactions.

Worth stating: Edge bundling is beautiful! It was an early algorithm we implemented. A bit differently, we did it interactively, so you could do things like adjust sliders real-time to get the right physics settings. However, we found you'd want to hover over individual nodes/edges to see what's in them, and "zoomable image" style makes that hard. I've been wanting to bring it back now that we're getting close to supporting dynamic grouping interactions, so cool to see you call it out.

That looks impressive, will definitely check it out!

Indeed, most of our users run on-premise / air-gapped. I recommend trying our cloud API to make sure it's in the right ballpark, and then we can get you setup!

Personally from using it, it goes well beyond just visualization of large graphs. Its not readily useful to see so many nodes, but what they've done is made it easier for people to parse through large graphs without a lot of hand holding. I would point out the work of others in this field, like Marc Khoury (which @MurrayHill1980 has already kindly mentioned), but this points to making large graphs usable for a variety of folks in my opinion.

Is it possible to run it locally? (i.e. without an API key)

Yep! Feel free to give the cloud version a go, and if it looks useful, we can help get you going on-premise. That would include our investigation platform as well.

I realized this is a great time to say -- we're hiring!

If you're into data visualization / UI engineering, fullstack node for data/security, or enterprise security sales, would love to chat. Our team is mostly in the bay area. If you've worked remotely before, that works great too.

We're especially growing in the security market around incident response + hunt. Our engineering work is around establishing more scalable best practices for investigation teams, building out our fullstack app, and we're in the middle of our next GPU visual analytics initiatives (accelerating interactive visual analytics another 100X!). So a lot of good stuff happening.

Another plot.ly pushing their freemium product via Github? No pricing on the homepage yet. Will come later, after enough people integrated it in their projects.

Anyone looking for an opensource version of plot.ly, take a look at http://www.metabase.com/

We're more complementary to Plotly & D3 than competitive. As an analogy for our developer API layers, we're closer to Google Maps than say D3 geo. Data tool teams shouldn't have to sink so much time on building out all the little interactions, just some configuring & skinning, and then back to data engineering & analytics. The result is we're getting embedded side-by-side. E.g., drop us into your Splunk/ELK dashboard that already does the basic charts. Or pilot our visual investigator, and get a direct visual exploration interface through a variety of connectors.

In terms of paid vs. unpaid, we're a pretty transparent team. We're iterating on a sustainable pricing model for enterprise investigation teams, advanced analysts, and developers building internal investigation tools. I don't believe in charging source developers, research scientists, non-profits, etc., and we'll be making our ongoing outreach work with those kinds of folks into a more formal program.

We have been and continue to be serious about community-minded open source contributions. Our team has contributed research that helped shape the modern web and leads open source projects that power a lot of tools used by the HN community today. Just this week, we released some work as part of the GoAI / Apache Arrow project, and that is part of our ongoing efforts to bring real GPU compute to the web world.

Maybe a good time to write -- we're hiring!

Does anyone know a good tool to visually explore large ontology-type graphs?

You may want to look into Linkurious: https://linkurio.us/ We provide a graph visualization interface that can connect to Neo4j, DataStax (DSE Graph), Titan or Allegrograph.

Disclaimer: I'm a co-founder of the company

https://linkurio.us/try/ should go straight to a demo page, not some signup popup

It goes to a signup popup.

I'm interested to see if anyone's offering a large-scale graph visualization that's not the "pulsating rainbow hairball".

Our full investigation platform supports connectors like those, and to non-graph sources (Spark, APIs, ...). We found teams need more than just a graph db frontend, especially as we enable them to put in more data into a session than previous-gen technologies. Ex: we created visual playbooks for running repeat investigations and binding the result to an interactive visual analytics session.

We're actively piloting our investigation platform with interesting teams, so please reach out: info@gr...com.

Hi, I'm Gephi+Linkurious co-founder. I've found visualizing large graphs pretty useless beyond the "I see meatballs!" effect and my opinion, after a decade in the field, is that it's the wrong problem for data analytics.

Much more interesting information is discovered during the process of dynamically building a visualization that is focused on user questions. I see with Linkurious that investigators usually need to visualize less than 1000 edges of a 1M+ edges graph to get answers.

The ultimate answer is generally a small graph: Graphistry is a tool that helps you get there. Why that's hard is most Splunk, Spark, etc. queries will return a bunch of events, and each event has a bunch of metadata. A tool should help, not fall over.

I think you're referring to scenarios closer to why we created the visual playbook concept and our embedding APIs. Small visualizations are often a good starting point in investigative scenarios. Even better.. no visualization, just full automation. We find this thinking comes up when the investigative flow is more established and curated. With visual playbooks, teams can record & automate multistep flows, run them whenever an incident happens, take action, and share & document the results. If part of the incident involves a bunch of events, or the analysts wants to dig in, our stack won't fall over. Instead, it provides a full visual analytics session with multiple cross-linked data views.

And we're fans of Gephi. We GPU accelerated the core algorithm -- we may be coming from a different perspective and user base.

Yup, it's important that people understand the role of visualization in the complete data chain.

I'm not sure I understand. Is there a resource that explains the role of visualizations in data flows in the context explained here?

Reply on news.ycombinator.com