Skip to content

Conversation

@ahijevyc
Copy link

Substituted cartopy for basemap. Removed HRRR and WRF stuff for now. regrid multiple levels with groupby('level').apply(griddata). plot and save new maps automatically. Eliminate driver script make_webplot.py.

There already is a ptype field, which can be called with 'ptype_prob' in web_plot.py.
testing with mpas hwt2017 ensemble.
…ic. netCDF4 MFDataset can't do that. More debugging messages
Mask speed_kts beyond 300 nm
Only warn about anti-cyclonic vmax if vmax > 17 m/s.
Chop out lat-lon box of MPAS for speed.

Deal with 1D and 2D arrays for spatial dimensions and refer to them as
*spatial_dimensions (i.e. use the * splat operator to expand the tuple
into comma-separated arguments).
Fixed tab, pickle read, no overplot

Replaced tabs with spaces.

Added binary flag to pickle read ('r' to 'rb')--for python3.

Removed reference to 'over' option (for overplot).
48-hr precip table

specifiy diagnostic file source for t2depart and t2-0c

thetapv

vertical velocity field name changed

reflectivity field name changed

specify diagnostic file source for srh

No vertical dimension for pressure-level height, speed and rh.

Top and bottom level for shear derivation.

Many other wind barb levels.

Every 50th instead of every 40th wind barb.

North America domain
Trim global mesh to cells inside lat/lon box.

Interpolate vorticity from verticies to cells using mpas_vort_cell
fortran program.

Added toServer method to webplot class.

Create output directory if needed.

Save and load a lot more in the pickle file, like triangulation,
weights, min_grid_spacing of mpas mesh.

Subtract accumulated precipitation at different lead times for
accumulated precipitation.

Interpolate to lat-lon grid before spatially smoothing.

In plotReflectivityUH, wrote a hack to plot zero contour if there are no other valid contours.

Rotate vectors so the point the right way on the map projection.

In some cases, convert xarray to numpy array. xarray is missing some
methods that numpy arrays have, like flatten method.

Unit conversions tied to new field names in hwt2017 output.

When deriving max/min/variance/prob over threshold, etc. use a tuple
with dimensions and use the splat operator *spatial_dimensions so it can
deal with 2 dimensional fields (like WRF) or 1 dimensional fields (like MPAS).
use logging

list of forecast hours, not start and end hour

TODO: get different domains to work
pk_file an object method, not a stand-alone method

Use fieldinfo.domains

Don't griddata the neighborhood data in readEnsemble. just return 3-d mesh
Preserve Dataset metadata when gridding. long_name and units with
colorbar.

Use --nbarbs to space wind barbs instead of --barbskip

Delimit fill/contour/barb arguents with slash, not underscore.

Remove makeEnsembleList variations.

Open fhr x member Dataset with "Time" and "mem" dimensions.

f2py directives for vert2cell.f90

Remove "filename" from fieldinfo. It was always "diag"

Add loadMap, readEnsemble, plotFields, and plotTitleTimes to webPlot
class initialization.

Remove webPlot methods plotDepartures, plotReports, plotInterp, plotWarnings,
plotOutlook, plotMRMS, plotFill_ptype, plotStreamlines.

Added webPlot.get_mpas_mesh method to return xarray Dataset of mesh lat/lon/area.

Use .startswith() and .endswith() a lot.

Don't bother inserting np.nan layers for missing ens members.

Simplify pickle file.
keep figure width constant. replicate a lot of basemap functions like
minlon, aspect...

call set_extent

Simplify calculation of nlons an nlats by doing it in lcc projecion space (meters).

place colorbar and logo correctly again with cartopy.

Use plt.subplots for postage stamp axes.

Rename fieldtype -> ensprod.

get neprob working again.
only use meshstr for prefix of pickle file and output file.

Delaunay
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant