def _default_plot_boundaries(basemap) : """Default function for 'plot_boundaries' output of plot_params()""" # ---------- Draw other stuff on the map # draw line around map projection limb. # color background of map projection region. # missing values over land will show up this color. basemap.drawmapboundary(fill_color='0.5') basemap.drawcoastlines() # draw parallels and meridians, but don't bother labelling them. basemap.drawparallels(np.arange(-90.,120.,30.)) basemap.drawmeridians(np.arange(0.,420.,60.))
def _default_plot_boundaries(basemap):
"""Default function for 'plot_boundaries' output of plot_params()"""
# ---------- Draw other stuff on the map
# draw line around map projection limb.
# color background of map projection region.
# missing values over land will show up this color.
basemap.drawmapboundary(fill_color='0.5')
basemap.drawcoastlines()
# draw parallels and meridians, but don't bother labelling them.
basemap.drawparallels(np.arange(-90., 120., 30.))
basemap.drawmeridians(np.arange(0., 420., 60.))
def plot_file(ifname, ofname):
"""region.center_ll:
(lat, lon) of center of region to plot"""
print('Plotting {}'.format(ifname))
with netCDF4.Dataset(ifname) as nc:
nlon = nc.variables['lon'].shape[0]
nlat = nc.variables['lat'].shape[0]
screenres = (1800, 2880) # y,x
dotpitch = 1. / 220 # 220 dpi monitor
# Read lons and lats (cell centers)
lon = nc.variables['lon'][:]
lat = nc.variables['lat'][:]
# Get data variable name
lonlat = {'lon', 'lat'}
for vname, ncvar in nc.variables.items():
if vname not in lonlat:
break
val = ncvar[:]
val[val == 0] = np.nan
plotter = giss.plot.LonLatPlotter(lon, lat, boundaries=False)
basemap = giss.basemap.global_map()
# One plot per page (figsize must be in inches)
figure = matplotlib.pyplot.figure(figsize=(screenres[1] * dotpitch,
screenres[0] * dotpitch))
ax = figure.add_subplot(111)
cb_args = dict()
if ifname.endswith('_lc_lr.nc'):
plot_args = dict(vmin=0., vmax=1.0)
elif ifname.endswith('_lai_lr.nc'):
plot_args = dict(vmin=0., vmax=7.0)
plot_args['cmap'] = giss.plot.read_cpt('NYT_drought14b.cpt').cmap
else:
raise ValueError('Unknown plot type (lc vs lai)')
_title = nc.title if hasattr(nc, 'title') else os.path.split(ifname)[1]
plt = giss.plot.plot_var(ax=ax,
basemap=basemap,
show=False,
plotter=plotter,
val=val,
title='{}\n{}'.format(_title,
ncvar.long_name),
plot_args=plot_args,
cb_args=cb_args)
basemap.drawcoastlines(linewidth=.1)
# draw parallels and meridians.
# label parallels on right and top
# meridians on bottom and left
# labels = [left,right,top,bottom]
meridians = np.arange(-180., 180., 30.)
basemap.drawmeridians(meridians, labels=[False, False, False, True])
parallels = np.arange(-90., 90., 30.)
basemap.drawparallels(parallels, labels=[True, False, False, False])
# Save to a file as png
figure.tight_layout(
pad=2.0
) # https://stackoverflow.com/questions/4042192/reduce-left-and-right-margins-in-matplotlib-plot
print(' Writing {}'.format(ofname))
figure.savefig('_tmp.png', dpi=1. / dotpitch, transparent=False)
os.rename('_tmp.png', ofname) # Write atomically
def plot_region(ps):
#ifname, ofname, region):
"""region.center_ll:
(lat, lon) of center of region to plot"""
region = regions[ps.algo]
if ps.ifname.endswith('_err.nc'):
return
print('Plotting {} ({})'.format(ps.ifname, region.name))
with netCDF4.Dataset(ps.ifname) as nc:
nlon = nc.variables['lon'].shape[0]
nlat = nc.variables['lat'].shape[0]
screenres = (1800, 2880) # y,x
dotpitch = 1. / 220 # 220 dpi monitor
# Convert region center from geographic lon/lat to i/j
# region.center_ll = (39., -98) # (lat,lon)
# region.center_ll = (-180+dxy[0]*1.5001,90.) # (lon, lat)
center_ji = ( # Zero-based indexing (lat,lon)
int(round((90. + region.center_ll[0]) * (nlat - 1) * (1. / 180.))),
int(round(
(180. + region.center_ll[1]) * (nlon - 1) * (1. / 360.))))
# Determine region bounds
# region_size = (int(screenres[0]*.8),int(screenres[1]*.95)) # Size of region, in gridcells (y,x)
region_size = (1320, 2600
) # Size of region to plot, in gridcells (y,x)
_base = (center_ji[0] - region_size[0] // 2,
center_ji[1] - region_size[1] // 2)
_top = (_base[0] + region_size[0]), _base[1] + region_size[1]
region_ji = (_base, _top)
if (_base[0] < 0 or _base[1] < 0 or _top[0] > nlat or _top[1] > nlon):
raise ValueError('Coordinates out of bounds', _base, _top,
region.center_ll)
# Read lons and lats (cell centers)
lon = nc.variables['lon'][_base[1]:_top[1]]
lat = nc.variables['lat'][_base[0]:_top[0]]
# Convert to boundaries
lonb = np.zeros(len(lon) + 1)
hdlon = .5 * 360. / nlon
lonb[0:-1] = lon - hdlon
lonb[-1] = lon[-1] + hdlon
latb = np.zeros(len(lat) + 1)
hdlat = .5 * 180. / nlat
latb[0:-1] = lat - hdlat
latb[-1] = lat[-1] + hdlat
# Get data variable name
ncvar = nc.variables[ps.vname]
if ps.layer < 0:
val = ncvar[_base[0]:_top[0], _base[1]:_top[1]]
else:
val = ncvar[ps.layer, _base[0]:_top[0], _base[1]:_top[1]]
val[val == 0] = np.nan
plotter = giss.plot.LonLatPlotter(lonb, latb, boundaries=True)
print(' Region ({:.2f}, {:.2f}) -- ({:.2f}, {:.2f})'.format(
lonb[0], latb[0], lonb[-1], latb[-1]))
# Use a custom basemap
#resolution='l',
basemap = mpl_toolkits.basemap.Basemap(
projection='laea',\
lat_0=region.center_ll[0], lon_0=region.center_ll[1],
llcrnrlon=lonb[0], llcrnrlat=latb[0],
urcrnrlon=lonb[-1], urcrnrlat=latb[-1])
# One plot per page (figsize must be in inches)
figure = matplotlib.pyplot.figure(figsize=(screenres[1] * dotpitch,
screenres[0] * dotpitch))
ax = figure.add_subplot(111)
cb_args = dict()
if ps.ifname.endswith('_lc.nc') or ps.ifname.endswith(
'brightratio.nc'):
plot_args = dict(vmin=0., vmax=1.0)
elif ps.ifname.endswith('_lai.nc'):
plot_args = dict(vmin=0., vmax=7.0)
plot_args['cmap'] = giss.plot.read_cpt('NYT_drought14b.cpt').cmap
else:
raise ValueError('Unknown plot type (lc vs lai)')
_title = nc.title if hasattr(nc, 'title') else os.path.split(
ps.ifname)[1]
plt = giss.plot.plot_var(ax=ax,
basemap=basemap,
show=False,
plotter=plotter,
val=val,
title='{}\n{}: ({})'.format(
_title, ncvar.long_name, region.name),
plot_args=plot_args,
cb_args=cb_args)
# draw parallels and meridians.
# label parallels on right and top
# meridians on bottom and left
# labels = [left,right,top,bottom]
meridians = np.arange(int(lonb[0]) - 1., int(lonb[-1]) + 1., 4.)
basemap.drawmeridians(meridians, labels=[False, False, False, True])
parallels = np.arange(int(latb[0]) - 1., int(latb[-1]) + 1., 4.)
basemap.drawparallels(parallels, labels=[True, False, False, False])
# Save to a file as png
figure.tight_layout(
pad=2.0
) # https://stackoverflow.com/questions/4042192/reduce-left-and-right-margins-in-matplotlib-plot
print(' Writing {}'.format(ps.ofname))
odir = os.path.split(ps.ofname)[0]
os.makedirs(odir, exist_ok=True)
tmp_fname = os.path.join(odir, '_tmp.png')
figure.savefig(tmp_fname, dpi=1. / dotpitch, transparent=False)
os.rename(tmp_fname, ps.ofname) # Write atomically