Options to generate only a subset of graphs of interest

[jteresco]

This is mainly in relation to the highway data check mode of the site update program. If only a subset of the data is used, the graph processing will likely crash. Do we want to be able to generate only a small subset of graphs?

[yakra]

Do we want to be able to generate only a small subset of graphs?

I think it would be nice to have. @michihdeu?

If only a subset of the data is used, the graph processing will likely crash.

As demonstrated here.
Would my solution there work for the system data graphs' routine?
Could a similar solution be used for the remaining "special interest" graphs?

And it all depends on implementation of course. If we only want one specific graph, we could probably skip over a lot of those sections of code accordingly.

[jteresco]

The thought that comes to mind for both this and #79, though, is how much time is worth putting in to get code to generate only a subset of graphs to save the few minutes it takes to generate all of them. So these are likely to be low priority.

[yakra]

Few minutes? Graph generation, for me, takes about 26 minutes, roughly 65% of the total processing time. Cutting that down = a substantial time savings in my book.

Nitpicking aside though, I understand about low priority. Writing & modifying the code takes labor, effort, time, all finite resources. If I want just a few graphs I'm content to edit systems.csv and go from there.

[michihdeu]

If I got it right, this is a bug report because graph generation crashes if the number of systems/regions is too low and #79 is a feature to detect and select the required graphs?

I agree that it's not high priority because Jim's server is a little bit faster than yakra's offline version. Assuming 65% it would about 20 minutes in total. Well, still annoying...

[yakra]

I just tested siteupdate using only two systems, one active & one preview. It ran to completion; graph generation was fine. 2c9d199 seems to have done the trick.

That, and graph generation has become an order of magnitude faster since this issue was opened.

Can this be closed?

[jteresco]

Deeming no longer a significant issue. @yakra or @michihdeu or others, feel free to reopen if you disagree.