Online monitor, simple plots of database data.

[PaoloGB]

• Talking with Edgar today, it is clear that the user interface plays a massive role in how appealing the kit is: we produce comprehensive raw data, but they mostly care about at-a-glance visualization of what is going on.
• A granularity of a few minutes is sufficient.
• He mentioned it would be useful to have a plot "show up" when you hover the cursor over a roi, with the graph showing you a speed distribution, light/dark indication, maybe a temperature overlay, etc.
• A company (Tecniplast) was showing their setup where a mouse was tracked by means of capacitive plates. Their software showed instant speed, total displacemente, average speed, etc
• The DAM system seems to be very common in labs: it is the one where the "movement" is detected with a IR led: when the fly crosses the threshold, it counts +1.
• We chatted with Kaya, from James' lab: she studies bees in tubes and tags bumblebees using QR codes stuck on their backs.
• Chatted with bethan.shaw @ emr (Southampton) they also use the DAM system. She seemed interested in the Ethoscope. I sent her an email with links and documentation.
• I asked how much a DAM system is:
  • 400 £ for the software license, it lasts 2-3 years
  • 100 £ for the hardware

[mark-grimes]

The more I think of it, the more I think some form of DQM is a perfectly reasonable expectation. We can't cover every requirement because a lot of it will be experiment specific, but we can cover some basics and we can allow it to be easily customised. I was thinking of using javascript because it can be modified in the browser - we provide code that runs in the browser which gets (probably binned) data from the database and a function signature that receives the data. The user could then modify the body of that function to control what actually gets plotted. It then struck me that I'm sure there are also python interpreters running in the browser, which is how I think Jupyter notebooks work. We could still offer live editing and not have to split off into another language. Whatever happens it should be something that can run live or in post analysis. So the same plot can be made for live DQM or for offline analysis.

[PaoloGB]

I think this weekend could be a good time to try and set up a basic tool for this. The hardware I want to develop will not take long (last famous words). You and Ryan could start setting up a skeleton tool for the DQM, then we can all have a tasks/functions do write: with some guidance I think I can write some bits of it as well.

I think we need to define the specifications before we start though.

[ryanfpage]

DQM does seem like a reasonable request. Do we have a list of what they might expect from the DQM?

[ryanfpage]

Do we access to a DAM system to have a look at?