cseaton / crane

Pre-/post processing tools for ocean modeling results.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Crane

Pre-/post processing tools for ocean modeling products.

Roadmap

  • fix all imports
    • should use absolute: import crane.data.timeArray
    • or explicit relatime: import .timeArray
    • avoid from x import * to keep namespaces intact
  • clean unused functions
  • pep8 all source
  • data module is unfortunately named... change
  • fix setup.py so that scripts get copied to correct bin dir
  • revisit docstrings (use numpy convetion?)
  • treat tappy better
    • fork tappy, morph to a generic installable package
    • or find alternative
  • revisit plotting routines:
    • inherit from plot base classes
    • remove obsolete stackPlot routines
  • re-implement extraction routines
    • process all variables simultaneously
    • save each day to disk immediately; i.e. do not store whole time period in memory
  • remove all old station files, use only csv implementation
  • use a database in stationCollection (in-memory): PyDbLite?
  • move definitions to separate file
    • plot variable names/units
    • tracer names/file name conventions
  • once relatively stable, progressively change naming conventions
  • revisit dataContainer
    • transect data shape should be intuitive [nx, nz, nComp, nTime]
    • requires generization of data array?
    • use existing data model implementation? hdf5 object?
    • could meshContainer and dataContainer be merged?
      • dataContainer is a meshContainer with degenerate mesh (e.g. a point)

Wishlist

  • make quick plot routine for any data: plotDataContainer
  • use matplotlib styles
  • support interactive plots in ipython (via bokeh?)
    • time series and maps etc
  • add support for different time zones
    • epoch time stamps are OK
    • all datetime objs are currently assumed to be in PST(!)
    • same for simulation time/matlab datenums/plot time

About

Pre-/post processing tools for ocean modeling results.


Languages

Language:Python 100.0%Language:Shell 0.0%