11

Looking for a way of visualizing public power generation data. Typical data set has 1–5 years of hourly data, and anything up to 30 columns. Needing to pan and zoom, and perhaps select sections. Needs to run on Linux.

The sometimes recommended TimeSearcher 2 is very shaky under Mono, and has unusual data format requirements.

EDIT: I'm looking to compare years of hourly generation data against regional demand. Some of the data streams may not have corresponding frequencies. I'm not looking for much programming input, as most of the analysis will be visual, so zooming and panning are important. There's unlikely to be a pre-made importer for the XML format used by the particular ISO, so I'd likely be spending the programming time on that. Not sure if I need any stats capabilities, though rolling daily/weekly averages/minima/maxima could be helpful.

I'm potentially going up against a very polished but misguided speaker in a few days who is using an entrenched industry position to ensure that expensive, polluting energy remains the status quo.

  • If you could add more information, that would help narrow down the answers. For example: Are you only looking for desktop applications, or are web-based solutions fine? How much programming should be involved? Is your focus exploration or presentation? – Patrick Hoefler Apr 24 '15 at 9:47
  • @scruss, what did you end up using? – Tom LaGatta Jun 6 '15 at 6:38
  • None, yet. Disappeared down a bit of an R rabbit hole. It looks like it could handle and graph the data (after a lot of cleanup) but it's a steep learning curve if you've never used it before. – scruss Jun 7 '15 at 23:13
5

It's hard to suggest a good tool without knowing how deep into programming you want to go, or if the tool is for exploration or presentation.

But here is a sample of many good tools out there:

Javascript:

R:

Python:

Tableau

  • Thanks - dygraphs looks great. I was kind of looking for a pre-rolled tool, as I've already had to do a lot of fighting with the data just to get it in a place to analyse. It's Generation Output and Capability Data from Ontario. – scruss Apr 23 '15 at 21:37
  • For exploration and hooking up with data sets, Tableau (Public) is pretty good, but doesn't work on Linux as far as I know. – philshem Apr 24 '15 at 7:08
-3

Python and matplotlib is the way to go

We're looking for long answers that provide some explanation and context. Don't just give a one-line answer; explain why your answer is right, ideally with citations. Answers that don't include explanations may be removed.

  • 5
    care to elaborate? – scruss Apr 23 '15 at 11:43
  • 1
    Thanks for the downvote. Just look over the internet. Python is great to visualize time series but you have to dig a little bit – ℕʘʘḆḽḘ Apr 23 '15 at 12:11
  • 4
    @Noobie : you've given a clue as to where to look, but the answer on its own isn't really useful. You say 'just look over the internet', but that's pretty much passing the buck -- you could've instead suggesed matplotlib, and then offered up some links to documentation on using it in situations that they've asked about. Does matplotlib have a dedicated viewer that supports pan & zoom, or would that be some higher level tool? If you're in a rush and don't have time to flesh out a comprehensive answer, you can always leave a comment instead of an answer. – Joe Apr 23 '15 at 20:44
  • Then you have the criticise the other answer as well. It is also limited to simply enumerate some solutions ... – Maître Peseur Apr 24 '15 at 5:24
  • The answer above is a Community Wiki that you are welcome to add to, or expand on. – philshem Apr 24 '15 at 11:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.