Skip to content

Latest commit

 

History

History
42 lines (39 loc) · 2.26 KB

ideas.md

File metadata and controls

42 lines (39 loc) · 2.26 KB
Raw

Random ideas and notes about direction

  1. Events with manual triggering

    • Stabilization guage: display the slope on an indicator dial with a range of +/- 10X noise level on first point (5 s moving average updated every second?).
    • Digital display of last measurement updated at a minimum of 1 Hz.
    • Set delay time after manual collection click (0 seconds or more in steps of 1 second).
    • Data added to standard self-updating plot.
    • User can choose X - axis.
    • User can choose to collect one or more channels (will have to pick a channel for the slope and value display).
    • User can have more than one column of manual data input.

  2. Copy settings from a previous run.

    • Drop down defaulting to 'NONE'.
    • On selection sets parameters to match (deep copy, so that changes made to them will not impact previous run).

  3. Consider storing list of all associated data and external files in the notebook metadata. This could then be used to zip up an experiment that can be transferred as one file to any other jupyter server.

    • May also be able to implement a way to reload the notebook with the initial raw datasets loaded in.
    • Could possibly save things in a way that the state of dataframes could be recovered. However, I am not sure we really want to do that. In a research lab situation it is probalby best to make a new copy of the notebook reload the data and reprocess it, rather than making changes to an existing notebook. This maintains a better record of what was done. Maybe a way to clearly continue, but force saving of a checkpoint?
    • This could save state as a bundle that could rebuild dataframes and DAQinstances, but I do not see a way to cleanly reproduce the state of any other variables.
    • From a usability standpoint this needs to work. Maybe initialization reloads the most recent versions of the data sets.
    • Save bundle should be a .gz with data files an html representation and the .ipynb. Maybe a .jpsl.gz file?

  4. .jpsl.gz definition: file containing .ipynb, html representation of it, .csv of any plain dataframes and .html of any DAQinstances.

    • Should running initialization load everything in this or should they just be available?