Gabor Szarnyas

@breandan PR #75 was merged and there is a new issue #82 for using semirings. What should we do with this issue?

Hello @achalagarwal, apologies, I missed the notification for this question. The JSON format in Graphalytics is legacy stuff, and I'm personally not a big fan of it. In fact, I...

@achalagarwal Graphalytics v1.3.0 has minimal changes: one fixed configuration file and improved logging in case of validation errors: https://github.com/ldbc/ldbc_graphalytics/releases Making the switch should be fairly easy but not necessary so...

@achalagarwal how's your progress going with Gunrock × Graphalytics? Can you give an estimate which algorithms you'll be able to support - all six directed/undirected - and when an initial...

@achalagarwal Great. Is the code available somewhere? > (which is a preprocessing step and would thus go in loadtime I guess) Yes. > Are you okay with incremental updates to...

@achalagarwal - it's not super urgent, feel free to take your time.

@achalagarwal thanks! I'm using the CSV and it's almost confirming my script with two caveats: - I don't need the first column (no big deal, I'm just throwing it away)...

> did you mean that I provide the makespan value in the "load" column? If possible, you should put the preprocessing time into `load` column. According to the spec: "During...

> Yes, this is possible but then I would be overriding the load time set by graphalytics' v1.2.0 here Sure, that's not a problem as long as it has clear...

@achalagarwal yes, this restriction still holds in the current version (v1.0) of the benchmark. So there's a fixed number of iterations and in each iteration, vertices will be assigned the...