tlon = tlon - 360.
pl.figure(figsize=(8.5, 11))
pl.subplot(2, 1, 1)
# subplot 1 just shows POP grid cells.
map = Basemap(projection='merc', lat_ts=20, llcrnrlon=-180, \
urcrnrlon=180, llcrnrlat=-84, urcrnrlat=84, resolution='c')
map.drawcoastlines()
map.fillcontinents(color='white')
x, y = map(tlon, tlat)
im = map.pcolor(x,y,MA.masked_array(N.zeros(temp.shape,'f'), temp.mask),\
shading='faceted',cmap=pl.cm.cool,vmin=0,vmax=0)
# disclaimer: these are not really the grid cells because of the
# way pcolor interprets the x and y args.
pl.title('(A) CCSM POP Grid Cells')
# subplot 2 is a contour plot of surface temperature from the
# CCSM ocean model.
pl.subplot(2, 1, 2)
map.drawcoastlines()
map.fillcontinents(color='white')
CS1 = map.contourf(x, y, temp, 15)
CS2 = map.contour(x, y, temp, 15, colors='black', linewidths=0.5)
pl.title('(B) Surface Temp contours on POP Grid')
pl.show()
#pl.savefig('ccsm_popgrid.ps')
# read in data on lat/lon grid.
hgt = load('500hgtdata.gz')
lons = load('500hgtlons.gz')
lats = load('500hgtlats.gz')
# shift data so lons go from -180 to 180 instead of 0 to 360.
hgt, lons = shiftgrid(180., hgt, lons, start=False)
lons, lats = meshgrid(lons, lats)
# setup of sinusoidal basemap
m = Basemap(resolution='c', projection='sinu', lon_0=0)
fig = figure()
ax = fig.add_axes([0.1, 0.1, 0.7, 0.7])
# make a filled contour plot.
x, y = m(lons, lats)
CS = m.contour(x, y, hgt, 15, linewidths=0.5, colors='k')
CS = m.contourf(x, y, hgt, 15, cmap=cm.jet)
l, b, w, h = ax.get_position()
cax = axes([l + w + 0.075, b, 0.05, h]) # setup colorbar axes
colorbar(drawedges=True, cax=cax) # draw colorbar
axes(ax) # make the original axes current again
# draw coastlines and political boundaries.
m.drawcoastlines()
m.drawmapboundary()
m.fillcontinents()
# draw parallels and meridians.
parallels = arange(-60., 90, 30.)
m.drawparallels(parallels, labels=[1, 0, 0, 0])
meridians = arange(-360., 360., 30.)
m.drawmeridians(meridians)
title('Sinusoidal Filled Contour Demo')
lons, lats = P.meshgrid(lons, lats)
# Example to show how to make multi-panel plots.
# 2-panel plot, oriented vertically, colorbar on bottom.
rcParams["figure.subplot.hspace"] = 0.4 # more height between subplots
rcParams["figure.subplot.wspace"] = 0.5 # more width between subplots
# create new figure
fig = P.figure()
# panel 1
mnh = Basemap(lon_0=-105, boundinglat=20.0, resolution="c", area_thresh=10000.0, projection="nplaea")
xnh, ynh = mnh(lons, lats)
ax = fig.add_subplot(211)
CS = mnh.contour(xnh, ynh, hgt, 15, linewidths=0.5, colors="k")
CS = mnh.contourf(xnh, ynh, hgt, 15, cmap=P.cm.Spectral)
# colorbar on bottom.
l, b, w, h = ax.get_position()
cax = P.axes([l, b - 0.05, w, 0.025]) # setup colorbar axes
P.colorbar(cax=cax, orientation="horizontal", ticks=CS.levels[0::4]) # draw colorbar
P.axes(ax) # make the original axes current again
mnh.drawcoastlines(linewidth=0.5)
delat = 30.0
circles = P.arange(0.0, 90.0, delat).tolist() + P.arange(-delat, -90, -delat).tolist()
mnh.drawparallels(circles, labels=[1, 0, 0, 0])
delon = 45.0
meridians = P.arange(0, 360, delon)
mnh.drawmeridians(meridians, labels=[1, 0, 0, 1])
P.title("NH 500 hPa Height (cm.Spectral)")
lat_150=[]
for xy in verts:
lon_150.append(xy[0])
lat_150.append(xy[1])
close(f)
dd_box=[array(lon_150).min(), array(lon_150).max(), array(lat_150).min(), array(lat_150).max()]
fat=0.05
box=[dd_box[0]-fat, dd_box[1]+fat, dd_box[2]-fat, dd_box[3]+fat]
f=figure()
mapobj= Basemap(projection='merc',lat_0=(box[2]+box[3])/2.0, lon_0=(box[0]+box[1])/2.0,llcrnrlat=box[2], llcrnrlon=box[0], urcrnrlat=box[3] , urcrnrlon=box[1], resolution='l',area_thresh=1., lat_ts=(box[2]+box[3])/2.0)
longr, latgr=meshgrid(lons,lats)
xx, yy = mapobj(longr, latgr)
mapobj.drawmapboundary()
mapobj.readshapefile('/flurry/home/scollis/shapes/cstntcd_r','coast',drawbounds=True,linewidth=0.5,color='k',antialiased=1,ax=None)
mapobj.contour(xx,yy,angs, levels=[150,30], colors=['r'])
mapobj.drawmeridians(array([130.2, 130.4, 130.6, 130.8,131.0,131.2, 131.4]), labels=[1,0,0,1])
mapobj.drawparallels(array([--12.8, -12.6, -12.4, -12.2, -12.0, -11.8, -11.6, -11.4]), labels=[1,0,0,1])
dd_box_xx, dd_box_yy=mapobj(array([dd_box[0], dd_box[1]]),array([dd_box[2], dd_box[3]]))
xy=dd_box_xx[0], dd_box_yy[0]
width, height= dd_box_xx[1]-dd_box_xx[0], dd_box_yy[1]-dd_box_yy[0]
my_patch=matplotlib.patches.Rectangle(xy, width, height, edgecolor='blue', facecolor='white')
ax=gca()
ax.add_patch(my_patch)
radar_xx, radar_yy=mapobj(array([ber_loc[1], gp_loc[1]]), array([ber_loc[0], gp_loc[0]]))
mapobj.plot(radar_xx, radar_yy, 'bo')
ax.text(radar_xx[0]+1000.0, radar_yy[0]-3000.0, 'Berrimah')
ax.text(radar_xx[1]+1000.0, radar_yy[1]-3000.0, 'Gunn Point')
savefig('/flurry/home/scollis/results/area_vis.png')
close(f)
def makeFigure(self):
"""
:todo: Uses pylab so may not be threadsafe.
"""
lon = self.grid.lon0 + N.arrayrange(self.grid.nlon) * self.grid.dlon
lat = self.grid.lat0 + N.arrayrange(self.grid.nlat) * self.grid.dlat
ll_lon, ll_lat = lon[0], lat[0]
ur_lon, ur_lat = lon[-1], lat[-1]
# Account for variable lat/lon axis ordering
#!TODO: Could we move this (along with the equivilent in render_imp.py into grid.py?
if self.grid.ilat < self.grid.ilon:
latLonOrdering = True
else:
latLonOrdering = False
if latLonOrdering:
var = self.grid.value
else:
var = MA.transpose(self.grid.value)
fig = p.figure()
map = Basemap(projection='cyl',
llcrnrlon=ll_lon,
llcrnrlat=ll_lat,
urcrnrlon=ur_lon,
urcrnrlat=ur_lat,
resolution='l')
## if self.grid.units:
## p.title("%s\n(%s)" % (self.grid.long_name, self.grid.units))
## else:
## p.title(self.grid.long_name)
p.title(self.grid.long_name)
if self.type == 'colour':
# transform_scalar doesn't support masked arrays so we must fill then replace the mask.
var_dat = map.transform_scalar(var.filled(1.0e20), lon, lat,
len(lon), len(lat))
var = MA.masked_values(var_dat, 1.0e20)
map.imshow(var,
cmap=self.cmap,
vmin=self.vmin,
vmax=self.vmax,
interpolation='nearest')
cbar = p.colorbar(orientation='horizontal', format='%.2g')
if self.grid.units:
cbar.ax.set_xlabel(self.grid.units)
else:
x, y = map(*p.meshgrid(lon, lat))
c = map.contour(x, y, var, 12, colors='black')
c.clabel(fontsize=8)
map.fillcontinents(color='#e0e0e0')
map.drawcoastlines(color='gray')
map.drawmeridians(p.arange(-180, 180, 30),
labels=[1, 0, 0, 1],
color='gray')
map.drawparallels(p.arange(-90, 90, 15),
labels=[1, 0, 0, 1],
color='gray')
# Wrap the caption
caption = word_wrap(self.caption, 80)
fig.text(0.1,
0.08,
caption,
fontsize=10,
horizontalalignment='left',
verticalalignment='top',
transform=fig.transFigure)
return fig
meridians = N.arange(-120., -60.0, 10.0)
fields = (cress1, cress2, cress3, cress_187, barnes_analyses[0],
barnes_analyses[1], barnes_analyses[2], barnes_3pass)
names = ('Cressman 1st Pass', 'Cressman 2nd Pass', 'Cressman 3rd Pass',
'Cressman R=1.87d', r'$\rm{Barnes} \gamma=1.0$',r'$\rm{Barnes} \gamma=0.4$',
r'$\rm{Barnes} \gamma=0.2$', r'$\rm{Barnes 3-pass} \gamma=1.0$')
filenames = ('cress1', 'cress2', 'cress3', 'cress187', 'barnes1', 'barnes2',
'barnes3', 'barnes3pass')
for field,name,filename in zip(fields, names, filenames):
f = P.figure()
bm_ps.drawstates()
bm_ps.drawcountries()
bm_ps.drawcoastlines()
bm_ps.drawparallels(parallels, labels = [1,1,0,0])
bm_ps.drawmeridians(meridians, labels = [0,0,1,1])
cp = bm_ps.contour(x_bm, y_bm, field, contours, cmap=colormap)
P.clabel(cp, fontsize=fontsize, fmt='%.1f')
f.text(0.5,0.95,name,horizontalalignment='center',fontsize=16)
if save_work:
P.savefig(filename + '.png', dpi = 300)
else:
P.show()
filename = 'cress12diff'
f = P.figure()
bm_ps.drawstates()
bm_ps.drawcountries()
bm_ps.drawcoastlines()
bm_ps.drawparallels(parallels, labels = [1,1,0,0])
bm_ps.drawmeridians(meridians, labels = [0,0,1,1])
cp = bm_ps.contour(x_bm, y_bm, cress1 - cress2, cmap=colormap)
contours = N.arange(5300., 6000., 60.0)
parallels = N.arange(25., 60.0, 10.0)
meridians = N.arange(-120., -60.0, 10.0)
fields = (heights_uniform, heights_cress, heights_uniforbm_ps, heights_cress2)
names = ('Uniform Weights', 'Cressman Weights',\
'Uniform weights with wind correction',\
'Cressman weights with wind correction')
filenames = ('uniform1','cress1','uniform2','cress2')
for field,name,filename in zip(fields, names, filenames):
f = P.figure()
bm_ps.drawstates()
bm_ps.drawcountries()
bm_ps.drawcoastlines()
bm_ps.drawparallels(parallels, labels = [1,1,0,0])
bm_ps.drawmeridians(meridians, labels = [0,0,1,1])
cp = bm_ps.contour(x_bm, y_bm, field, contours, cmap=colormap)
P.clabel(cp, fontsize=fontsize, fmt='%.1f')
f.text(0.5,0.95,name,horizontalalignment='center',fontsize=16)
if save_work:
P.savefig(filename + '.png', dpi = 300)
else:
P.show()
height_diff = list()
if save_work:
sys.stdout = open(filename + '.txt', 'w')
print name
print "\nAnalysis Array:"
for n in N.arange(y.size):
print "\t%d" % (n + 1),
print
for (num, row) in enumerate(field.T):
figure(figsize=(8, 8))
subplot(ncol, nrow, nslice)
for i in xrange(nslice):
lon = composites.getLongitude()
lat = composites.getLatitude()
xx, yy = meshgrid(lon, lat)
m = Basemap(resolution='l',
lat_0=N.average(lat),
lon_0=N.average(lon),
llcrnrlon=min(lon),
llcrnrlat=min(lat),
urcrnrlon=max(lon),
urcrnrlat=max(lat))
levels = vcs.mkscale(genutil.minmax(composites))
m.contourf(lon, lat, composites[i], levels=levels, cmap=bwr)
clabel(m.contour(lon, lat, composites[i], levels=levels))
title("Phase %i/%i" (i + 1, 6))
m.drawcoastlines()
m.drawcountries()
m.fillcontinents(color='coral')
m.drawparallels(vcs.mkscale(genutil.minmax(lat)), labels=[1, 0, 0, 0])
m.drawmeridians(vcs.mkscale(genutil.minmax(lon)), labels=[0, 0, 0, 1])
figtext(
.5, 1.,
"\n%s [%s]" % ("El Nino phase composites\nSea surface temperature",
composites.units))
savefig(sys.argv[0].replace(".py", ".png"))
show()
# Fall back to vcs because we have it!
except:
plots = ['contour','pcolor']
#plots = ['contour','imshow']
for np,plot in enumerate(plots):
fig.add_subplot(1,2,np+1)
ax = gca()
# plot data.
print plot+' plot ...'
if plot == 'pcolor':
m.pcolor(x,y,hgt,shading='flat')
elif plot == 'imshow':
im = m.imshow(hgt)
elif plot == 'contour':
levels, colls = m.contour(x,y,hgt,15,linewidths=0.5,colors='k')
levels, colls = m.contourf(x,y,hgt,15,cmap=cm.jet,colors=None)
# set size of plot to match aspect ratio of map.
l,b,w,h = ax.get_position()
b = 0.5 - 0.5*w*m.aspect; h = w*m.aspect
ax.set_position([l,b,w,h])
# draw map.
m.drawcoastlines()
# draw parallels
delat = 30.
delon = 90.
circles = arange(10.,90.+delat,delat).tolist()
m.drawparallels(circles,labels=[0,0,1,1])
'green': ((0.,)*3,(.5,1.,1.),(1.,0.,0.)),
'blue': ((0.,1.,1.),(.5,1.,1.),(1.,0.,0.))}
return matplotlib.colors.LinearSegmentedColormap('bwr',cdict,256)
figure(figsize=(8,8))
subplot(ncol,nrow,nslice)
for i in xrange(nslice):
lon = composites.getLongitude()
lat = composites.getLatitude()
xx,yy = meshgrid(lon,lat)
m = Basemap(resolution='l',lat_0=N.average(lat),lon_0=N.average(lon),
llcrnrlon=min(lon),llcrnrlat=min(lat),
urcrnrlon=max(lon),urcrnrlat=max(lat))
levels = vcs.mkscale(genutil.minmax(composites))
m.contourf(lon,lat,composites[i],levels=levels,cmap=bwr)
clabel(m.contour(lon,lat,composites[i],levels=levels))
title("Phase %i/%i" (i+1,6))
m.drawcoastlines()
m.drawcountries()
m.fillcontinents(color='coral')
m.drawparallels(vcs.mkscale(genutil.minmax(lat)),labels=[1,0,0,0])
m.drawmeridians(vcs.mkscale(genutil.minmax(lon)),labels=[0,0,0,1])
figtext(.5,1.,"\n%s [%s]" % ("El Nino phase composites\nSea surface temperature", composites.units))
savefig(sys.argv[0].replace(".py",".png"))
show()
# Fall back to vcs because we have it!
except:
# Plot 1 phase over two, then a time series
# TODO: we must do something nicer!!
x_bm,y_bm = ps2.to_basemap_xy(m2, x_grid / cm_per_m, y_grid / cm_per_m) #Radius factor used in all of the analyses R = 4.32 heights_cress = oban.analyze_grid(data.heights, x_grid, y_grid, data.x, data.y, oban.cressman_weights, R) contours = N.arange(5300., 6000., 60.0) parallels = N.arange(25., 60.0, 10.0) meridians = N.arange(-120., -60.0, 10.0) ##m2.drawstates() ##m2.drawcountries() ##m2.drawcoastlines() ##m2.drawparallels(parallels, labels = [1,1,0,0]) ##m2.drawmeridians(meridians, labels = [0,0,1,1]) ob_x, ob_y = ps2.to_basemap_xy(m2, data.x / cm_per_m, data.y / cm_per_m) #m2.plot(ob_x, ob_y, 'bx') #m2.plot(x_bm, y_bm, 'g.') for name in M.cm.cmapnames: f = P.figure() m2.drawstates() m2.drawcountries() m2.drawcoastlines() m2.drawparallels(parallels, labels = [1,1,0,0]) m2.drawmeridians(meridians, labels = [0,0,1,1]) c = m2.contour(x_bm, y_bm, heights_cress, contours, cmap=M.cm.get_cmap(name)) P.title(name) #P.clabel(c) P.show()
# Example to show how to make multi-panel plots.
# 2-panel plot, oriented vertically, colorbar on bottom.
rcParams['figure.subplot.hspace'] = 0.4 # more height between subplots
rcParams['figure.subplot.wspace'] = 0.5 # more width between subplots
# create new figure
fig=P.figure()
# panel 1
mnh = Basemap(lon_0=-105,boundinglat=20.,
resolution='c',area_thresh=10000.,projection='nplaea')
xnh,ynh = mnh(lons,lats)
ax = fig.add_subplot(211)
CS = mnh.contour(xnh,ynh,hgt,15,linewidths=0.5,colors='k')
CS = mnh.contourf(xnh,ynh,hgt,15,cmap=P.cm.Spectral)
# colorbar on bottom.
l,b,w,h = ax.get_position()
cax = P.axes([l, b-0.05, w, 0.025]) # setup colorbar axes
P.colorbar(cax=cax, orientation='horizontal',ticks=CS.levels[0::4]) # draw colorbar
P.axes(ax) # make the original axes current again
mnh.drawcoastlines(linewidth=0.5)
delat = 30.
circles = P.arange(0.,90.,delat).tolist()+\
P.arange(-delat,-90,-delat).tolist()
mnh.drawparallels(circles,labels=[1,0,0,0])
delon = 45.
meridians = P.arange(0,360,delon)
mnh.drawmeridians(meridians,labels=[1,0,0,1])
P.title('NH 500 hPa Height (cm.Spectral)')
plots = ['contour', 'pcolor']
#plots = ['contour','imshow']
for np, plot in enumerate(plots):
fig.add_subplot(1, 2, np + 1)
ax = gca()
# plot data.
print plot + ' plot ...'
if plot == 'pcolor':
m.pcolor(x, y, hgt, shading='flat')
elif plot == 'imshow':
im = m.imshow(hgt)
elif plot == 'contour':
levels, colls = m.contour(x, y, hgt, 15, linewidths=0.5, colors='k')
levels, colls = m.contourf(x, y, hgt, 15, cmap=cm.jet, colors=None)
# set size of plot to match aspect ratio of map.
l, b, w, h = ax.get_position()
b = 0.5 - 0.5 * w * m.aspect
h = w * m.aspect
ax.set_position([l, b, w, h])
# draw map.
m.drawcoastlines()
# draw parallels
delat = 30.
delon = 90.
circles = arange(10., 90. + delat, delat).tolist()
# compute native x,y coordinates of grid.
x, y = m(lons, lats)
# define parallels and meridians to draw.
parallels = p.arange(-80., 90, 20.)
meridians = p.arange(0., 360., 20.)
# number of repeated frames at beginning and end is n1.
nframe = 0
n1 = 10
l, b, w, h = ax.get_position()
# loop over times, make contour plots, draw coastlines,
# parallels, meridians and title.
for nt, date in enumerate(datelabels[1:]):
CS = m.contour(x,
y,
slp[nt, :, :],
clevs,
linewidths=0.5,
colors='k',
animated=True)
CS = m.contourf(x,
y,
slp[nt, :, :],
clevs,
cmap=p.cm.RdBu_r,
animated=True)
# plot wind vectors on lat/lon grid.
# rotate wind vectors to map projection coordinates.
#urot,vrot = m.rotate_vector(u[nt,:,:],v[nt,:,:],lons,lats)
# plot wind vectors over map.
#Q = m.quiver(x,y,urot,vrot,scale=500)
# plot wind vectors on projection grid (looks better).
def plot_ncdf_output(id, datafn, **kwds):
"""Plot model field id from the data in netCDF file datafn.
Positional Input Parameter:
* id: Name of the id of the field to plot. String.
* datafn: Filename containing the output data to plot. String.
Input keyword parameter descriptions are found in the docstring
for Qtcm methods ploti, plotm, and other methods that call this
private method. In general, those methods take the keyword
parameters they receive and pass it along unchanged as keyword
parameters to this function. In that sense, this function is
seldom used as a stand-alone function, but rather is usually
used coupled with a Qtcm instance.
The data fields read in from the netCDF output file are dimensioned
(time, lat, lon). This is different than how the data is stored
in the compiled QTCM model fields (lon, lat, time), and at the
Python level (lon, lat). The reason this is the case is that
f2py automatically makes the arrays passed between the Python
and Fortran levels match.
For a lat vs. lon plot, the contour plot is superimposed onto
a cylindrical projection map of the Earth with continents drawn
and labeled meridians and parallels. The title also includes
the model time, and x- and y-axis labels are not drawn.
All numerical data used for plotting come from the netCDF output
file for consistency (e.g., the dimensions of u1). Currently
this method only works for 3-D data arrays (two in space, one
in time).
"""
#- Accomodate other ids. The id that this routine will use
# (i.e., iduse corresponds to the name in the netCDF output file)
# is called iduse. Set this to id, except for the case where
# some aliases of ids are entered in which iduse is the alias:
if id == 'Qc': iduse = 'Prec'
elif id == 'FLWut': iduse = 'OLR'
elif id == 'STYPE': iduse = 'stype'
else: iduse = id
#- Set defined keyword defaults. All are set to None except for
# nlatlon which gets an integer:
plotkwds_ids = ['lat', 'lon', 'time', 'fn', 'levels', 'title',
'xlabel', 'ylabel',
'filled', 'nlatlon', 'tmppreview']
plotkwds = {}
for ikey in plotkwds_ids:
if kwds.has_key(ikey):
plotkwds[ikey] = copy.copy(kwds[ikey])
else:
plotkwds[ikey] = None
if not kwds.has_key('nlatlon'):
plotkwds['nlatlon'] = 8
#- Get data and dimensions of iduse to plot:
fileobj = S.NetCDFFile(datafn, mode='r')
data = N.array(fileobj.variables[iduse].getValue())
data_name = fileobj.variables[iduse].long_name
data_units = fileobj.variables[iduse].units
dim = {}
dimname = {}
dimunits = {}
dim['lat'] = N.array(fileobj.variables['lat'].getValue())
dimname['lat'] = fileobj.variables['lat'].long_name
dimunits['lat'] = fileobj.variables['lat'].units
dim['lon'] = N.array(fileobj.variables['lon'].getValue())
dimname['lon'] = fileobj.variables['lon'].long_name
dimunits['lon'] = fileobj.variables['lon'].units
dim['time'] = N.array(fileobj.variables['time'].getValue())
dimname['time'] = fileobj.variables['time'].long_name
dimunits['time'] = fileobj.variables['time'].units
fileobj.close()
#- Alter data long name to remove any units. The definition
# of units as the substring within the [] is the same as used in
# defVar in output.F90 of the compiled QTCM model. Remove
# underscores and extra whitespace in data_name and data_units,
# replacing with a single whitespace character between words:
idx1 = data_name.find('[')
idx2 = data_name.find(']')
if idx1 != -1 and idx2 != -1:
data_name = data_name[:idx1] + data_name[idx2+1:]
data_name = data_name.strip()
data_name = ' '.join(data_name.replace('_',' ').split())
data_units = ' '.join(data_units.replace('_',' ').split())
#- Alter dimension long name to remove any units. The definition
# of units as the substring within the [] is the same as used in
# defVar in output.F90 of the compiled QTCM model. Remove
# underscores and extra whitespace in name and units, replacing
# with a single whitespace character between words, and
# capitalizing like a title:
for idimkey in dim.keys():
idimname = dimname[idimkey]
idx1 = idimname.find('[')
idx2 = idimname.find(']')
if idx1 != -1 and idx2 != -1:
idimname = idimname[:idx1] + idimname[idx2+1:]
dimname[idimkey] = idimname.strip()
dimname[idimkey] = \
' '.join(dimname[idimkey].replace('_',' ').split()).title()
dimunits[idimkey] = \
' '.join(dimunits[idimkey].replace('_',' ').split()).title()
#- Some data checks:
if N.rank(data) != 3:
raise ValueError, '_plot: can only plot lat, lon, time fields'
if not N.allclose(dim['time'], N.sort(dim['time'])):
raise ValueError, '_plot: time not monotonically ascending'
if not N.allclose(dim['lat'], N.sort(dim['lat'])):
raise ValueError, '_plot: lat not monotonically ascending'
if not N.allclose(dim['lon'], N.sort(dim['lon'])):
raise ValueError, '_plot: lon not monotonically ascending'
if N.shape(data)[0] != N.size(dim['time']):
raise ValueError, '_plot: data time dim mismatch'
if N.shape(data)[1] != N.size(dim['lat']):
raise ValueError, '_plot: data lat dim mismatch'
if N.shape(data)[2] != N.size(dim['lon']):
raise ValueError, '_plot: data lon dim mismatch'
#- Choose and describe ranges for lat, lon, and time. The
# section cycles through the dictionary of dimensions. idim is
# the 1-D array of the values of that dimension. rngs is a
# dictionary where each entry corresponds to a dimension, and the
# value of the entry is the values of that dimension that are to
# be plotted. rngs_idxs are the indices in the original
# dimensions array corresponding to the values in rngs.
# keys_rngs_sizes_gt_1 is a list of the keys of ranges that have
# sizes greater than 1:
rngs = {}
rngs_idxs = {}
keys_rngs_sizes_gt_1 = []
for idimkey in dim.keys():
idim = dim[idimkey]
if plotkwds[idimkey] == None:
dim_mask = N.ones( N.size(idim), dtype=int )
elif N.isscalar(plotkwds[idimkey]):
dim_mask = where_close( idim, plotkwds[idimkey] )
if N.sum(dim_mask) != 1:
raise ValueError, 'no point chosen'
elif (not N.isscalar(plotkwds[idimkey])) and \
N.size(plotkwds[idimkey]) == 1:
dim_mask = where_close( idim, plotkwds[idimkey][0] )
if N.sum(dim_mask) != 1:
raise ValueError, 'no point chosen'
elif N.size(plotkwds[idimkey]) == 2:
dim_mask = N.logical_and( idim >= plotkwds[idimkey][0],
idim <= plotkwds[idimkey][-1] )
else:
raise ValueError, 'bad dimension range keyword entry'
rngs[idimkey] = N.compress( dim_mask, idim )
rngs_idxs[idimkey] = N.compress( dim_mask, N.arange(N.size(idim)) )
if N.size(rngs[idimkey]) > 1:
keys_rngs_sizes_gt_1.append(idimkey)
#- Set plot types (line or contour):
if len(keys_rngs_sizes_gt_1) == 0:
raise ValueError, 'cannot plot without any fixed dimension'
elif len(keys_rngs_sizes_gt_1) == 1:
plottype = 'line'
elif len(keys_rngs_sizes_gt_1) == 2:
plottype = 'contour'
else:
raise ValueError, 'cannot plot with > 2 varying dimensions'
#- Set plot axis fields and axis names, depending on what sort
# of dimensions will be plotted. If lon is to be plotted, it is
# always the x-axis. If lat is to be plotted, it is always the
# y-axis. In this section and later on in a few places, I
# rely on the count method for a list as a Boolean test: If it
# returns 0, consider that False; > 0 is True. The text for the
# title and axis labels to be passed to the xlabel, etc. methods,
# are called titlename, xname, and yname:
if plottype == 'line': #+ Choose x-axis vector and
x = rngs[keys_rngs_sizes_gt_1[0]] # x/y names for line plot
xname = dimname[keys_rngs_sizes_gt_1[0]] + ' [' \
+ dimunits[keys_rngs_sizes_gt_1[0]] + ']'
yname = data_name + ' [' + data_units + ']'
elif plottype == 'contour': #+ Choose axis vectors and
if keys_rngs_sizes_gt_1.count('lon'): # names for contour plot
x = rngs['lon']
xname = dimname['lon'] + ' [' + dimunits['lon'] + ']'
if keys_rngs_sizes_gt_1.count('lat'):
y = rngs['lat']
yname = dimname['lat'] + ' [' + dimunits['lat'] + ']'
if keys_rngs_sizes_gt_1.count('time'):
if keys_rngs_sizes_gt_1.count('lon'):
y = rngs['time']
yname = dimname['time'] + ' [' + dimunits['time'] + ']'
elif keys_rngs_sizes_gt_1.count('lat'):
x = rngs['time']
xname = dimname['time'] + ' [' + dimunits['time'] + ']'
else:
raise ValueError, 'bad treatment of time'
else:
raise ValueError, 'unrecognized plottype'
#- Override xname, yname, and titlename with keywords, if they
# are not None. titlename receives data_name and data_units
# by default:
if plotkwds['xlabel'] != None:
xname = plotkwds['xlabel']
if plotkwds['ylabel'] != None:
yname = plotkwds['ylabel']
if plotkwds['title'] != None:
titlename = plotkwds['title']
else:
titlename = data_name + ' [' + data_units + ']'
#- Pick data to be plotted and plot:
pylab.clf() #+ Clear any previous figures
pylab.figure(1) #+ Open a pylab figure
if plottype == 'line': #+ Select data for a line plot
y = data[rngs_idxs['time'], # and plot
rngs_idxs['lat'],
rngs_idxs['lon']]
pylab.plot(x, y)
elif plottype == 'contour': #+ Select data for a contour
ritim = rngs_idxs['time'] # plot and plot
rilat = rngs_idxs['lat']
rilon = rngs_idxs['lon']
#* Extract subarrays depending on which two dimensions are
# chosen:
if N.size(rngs_idxs['time']) == 1:
zgrid = num.MLab.squeeze(data[ ritim[0],
rilat[0]:rilat[-1]+1,
rilon[0]:rilon[-1]+1 ])
elif N.size(rngs_idxs['lat']) == 1:
zgrid = num.MLab.squeeze(data[ ritim[0]:ritim[-1]+1,
rilat[0],
rilon[0]:rilon[-1]+1 ])
elif N.size(rngs_idxs['lon']) == 1:
zgrid = num.MLab.squeeze(data[ ritim[0]:ritim[-1]+1,
rilat[0]:rilat[-1]+1,
rilon[0] ])
else:
raise ValueError, 'unrecognized configuration'
#* Change zgrid for special case of a lat. vs. time contour
# plot. Calculate xgrid and ygrid:
if keys_rngs_sizes_gt_1.count('time') and \
keys_rngs_sizes_gt_1.count('lat'):
zgrid = N.transpose(zgrid)
xgrid, ygrid = pylab.meshgrid(x, y)
#* Set contour levels:
if plotkwds['levels'] == None:
levels = nice_levels(zgrid)
else:
levels = plotkwds['levels']
#- Plot (creating continents first if is a lat vs. lon plot)
# and write contour levels/color bar as appropriate:
if keys_rngs_sizes_gt_1.count('lon') and \
keys_rngs_sizes_gt_1.count('lat'):
mapplot = Basemap(projection='cyl', resolution='l',
llcrnrlon=N.min(xgrid), llcrnrlat=N.min(ygrid),
urcrnrlon=N.max(xgrid), urcrnrlat=N.max(ygrid))
mapplot.drawcoastlines()
mapplot.drawmeridians(nice_levels(rngs['lon'],
approx_nlev=plotkwds['nlatlon']),
labels=[1,0,0,1])
mapplot.drawparallels(nice_levels(rngs['lat'],
approx_nlev=plotkwds['nlatlon']),
labels=[1,0,0,1])
if plotkwds['filled']:
plot = mapplot.contourf(xgrid, ygrid, zgrid, levels)
pylab.colorbar(plot, orientation='horizontal', format='%g')
else:
plot = mapplot.contour(xgrid, ygrid, zgrid, levels)
pylab.clabel(plot, inline=1, fontsize=10, fmt='%g')
else:
if plotkwds['filled']:
plot = pylab.contourf(xgrid, ygrid, zgrid, levels)
pylab.colorbar(plot, orientation='horizontal', format='%g')
else:
plot = pylab.contour(xgrid, ygrid, zgrid, levels)
pylab.clabel(plot, inline=1, fontsize=10, fmt='%g')
else:
raise ValueError, 'unrecognized plottype'
#- Add titling. Lat vs. lon plots do not have axis labels because
# the map labels already make it clear, and for those plots the
# title also includes the time value:
if keys_rngs_sizes_gt_1.count('lon') and \
keys_rngs_sizes_gt_1.count('lat'):
titlename = titlename + ' at ' \
+ dimname['time'] + ' ' \
+ str(rngs['time'][0]) + ' ' \
+ dimunits['time']
titlename = mpl_latex_script1(titlename)
pylab.title(titlename)
else:
titlename = mpl_latex_script1(titlename)
xname = mpl_latex_script1(xname)
yname = mpl_latex_script1(yname)
pylab.xlabel(xname)
pylab.ylabel(yname)
pylab.title(titlename)
#- Output plot to PNG file or screen. The show command seems to
# have a problem on my Mac OS X, so save to a temporary file
# and use preview to view for fn == None and tmppreview set to
# True. Note that the temporary file is not deleted by this
# method:
if plotkwds['fn'] == None: #+ Screen display
if plotkwds['tmppreview'] and sys.platform == 'darwin':
outputfn = tempfile.mkstemp('.png','qtcm_')
pylab.savefig(outputfn[-1])
os.system('open -a /Applications/Preview.app '+outputfn[-1])
else:
pylab.show()
elif type(plotkwds['fn']) == type('a'): #+ Write to file
pylab.savefig(plotkwds['fn'])
pylab.close(1)
else:
raise ValueError, 'cannot write to this type of file'
map.drawmeridians(p.arange(0, 360, 30))
map.drawparallels(p.arange(-90, 90, 30))
# lat/lon coordinates of five cities.
lats = [40.02, 32.73, 38.55, 48.25, 17.29]
lons = [-105.16, -117.16, -77.00, -114.21, -88.10]
cities = [
'Boulder, CO', 'San Diego, CA', 'Washington, DC', 'Whitefish, MT',
'Belize City, Belize'
]
# compute the native map projection coordinates for cities.
x, y = map(lons, lats)
# plot filled circles at the locations of the cities.
map.plot(x, y, 'bo')
# plot the names of those five cities.
for name, xpt, ypt in zip(cities, x, y):
p.text(xpt + 50000, ypt + 50000, name, fontsize=9)
# make up some data on a regular lat/lon grid.
nlats = 73
nlons = 145
delta = 2. * p.pi / (nlons - 1)
lats = (0.5 * p.pi - delta * p.indices((nlats, nlons))[0, :, :])
lons = (delta * p.indices((nlats, nlons))[1, :, :])
wave = 0.75 * (p.sin(2. * lats)**8 * p.cos(4. * lons))
mean = 0.5 * p.cos(2. * lats) * ((p.sin(2. * lats))**2 + 2.)
# compute native map projection coordinates of lat/lon grid.
x, y = map(lons * 180. / p.pi, lats * 180. / p.pi)
# contour data over the map.
cs = map.contour(x, y, wave + mean, 15, linewidths=1.5)
p.show()
#p.savefig('wiki_example.ps')
# Transform the grid to basemap space for plotting
cm_per_m = 100.0
x_bm,y_bm = to_basemap_xy(bm_ps, x_grid / cm_per_m, y_grid / cm_per_m,
-110.0, 90.0, 1.0/15000000.0)
#Specify contour intervals, as well as what parallels and meridians to show
contours = N.arange(5300., 6000., 60.0)
parallels = N.arange(25., 60.0, 10.0)
meridians = N.arange(-120., -60.0, 10.0)
#These draw the various maps (self explanatory)
bm_ps.drawstates()
bm_ps.drawcountries()
bm_ps.drawcoastlines()
#Draw grid lines, labels specify where to label the grid lines
#[left, right, top, bottom]
bm_ps.drawparallels(parallels, labels = [1,1,0,0])
bm_ps.drawmeridians(meridians, labels = [0,0,1,1])
#Remove the cmap argument below to get the normal, colored colortable
cp = bm_ps.contour(x_bm, y_bm, data, contours, cmap=M.cm.get_cmap('gist_gray'))
#Label contours
P.clabel(cp)
#Normally just P.title('My plot'), but the longitude labels get overlapped
P.gcf().text(0.5,0.95,'My Plot',horizontalalignment='center',fontsize=16)
#P.savefig('plot.png', dpi = 300) # This saves to a file
P.show()
pl.figure(figsize=(8.5,11))
pl.subplot(2,1,1)
# subplot 1 just shows POP grid cells.
map = Basemap(projection='merc', lat_ts=20, llcrnrlon=-180, \
urcrnrlon=180, llcrnrlat=-84, urcrnrlat=84, resolution='c')
map.drawcoastlines()
map.fillcontinents(color='white')
x, y = map(tlon,tlat)
im = map.pcolor(x,y,MA.masked_array(N.zeros(temp.shape,'f'), temp.mask),\
shading='faceted',cmap=pl.cm.cool,vmin=0,vmax=0)
# disclaimer: these are not really the grid cells because of the
# way pcolor interprets the x and y args.
pl.title('(A) CCSM POP Grid Cells')
# subplot 2 is a contour plot of surface temperature from the
# CCSM ocean model.
pl.subplot(2,1,2)
map.drawcoastlines()
map.fillcontinents(color='white')
CS1 = map.contourf(x,y,temp,15)
CS2 = map.contour(x,y,temp,15,colors='black',linewidths=0.5)
pl.title('(B) Surface Temp contours on POP Grid')
pl.show()
#pl.savefig('ccsm_popgrid.ps')
rsphere=6371200.,
resolution='l',
area_thresh=5000.,
projection='lcc',
lat_1=standardpar,
lon_0=centerlon)
x, y = m(lons, lats)
# setup figure size so each panel will have same aspect ratio as map.
xsize = rcParams['figure.figsize'][0]
ysize = (3. / 2.) * m.aspect * xsize
fig = figure(figsize=(xsize, ysize))
yoffset = (m.urcrnry - m.llcrnry) / 30.
for npanel, fcsthr in enumerate(arange(0, 72, 12)):
nt = fcsthrs.index(fcsthr)
ax = fig.add_subplot(320 + npanel + 1)
levels, colls = m.contour(x, y, t2m[nt, :, :], clevs, colors='k')
levels, colls = m.contourf(x,
y,
t2m[nt, :, :],
clevs,
cmap=cm.jet,
colors=None)
m.drawcoastlines()
m.drawstates()
m.drawcountries()
m.drawparallels(arange(25, 75, 20),
labels=[1, 0, 0, 0],
fontsize=8,
fontstyle='oblique')
m.drawmeridians(arange(-140, 0, 20),
labels=[0, 0, 0, 1],
fig = p.figure()
ax = fig.add_axes([0.1,0.1,0.7,0.7])
# set desired contour levels.
clevs = p.arange(960,1061,5)
# compute native x,y coordinates of grid.
x, y = m(lons, lats)
# define parallels and meridians to draw.
parallels = p.arange(-80.,90,20.)
meridians = p.arange(0.,360.,20.)
# number of repeated frames at beginning and end is n1.
nframe = 0; n1 = 10
l,b,w,h=ax.get_position()
# loop over times, make contour plots, draw coastlines,
# parallels, meridians and title.
for nt,date in enumerate(datelabels[1:]):
CS = m.contour(x,y,slp[nt,:,:],clevs,linewidths=0.5,colors='k',animated=True)
CS = m.contourf(x,y,slp[nt,:,:],clevs,cmap=p.cm.RdBu_r,animated=True)
# plot wind vectors on lat/lon grid.
# rotate wind vectors to map projection coordinates.
#urot,vrot = m.rotate_vector(u[nt,:,:],v[nt,:,:],lons,lats)
# plot wind vectors over map.
#Q = m.quiver(x,y,urot,vrot,scale=500)
# plot wind vectors on projection grid (looks better).
# first, shift grid so it goes from -180 to 180 (instead of 0 to 360
# in longitude). Otherwise, interpolation is messed up.
ugrid,newlons = shiftgrid(180.,u[nt,:,:],longitudes,start=False)
vgrid,newlons = shiftgrid(180.,v[nt,:,:],longitudes,start=False)
# transform vectors to projection grid.
urot,vrot,xx,yy = m.transform_vector(ugrid,vgrid,newlons,latitudes,51,51,returnxy=True,masked=True)
# plot wind vectors over map.
Q = m.quiver(xx,yy,urot,vrot,scale=500)
resolution="l",
area_thresh=5000.0,
projection="lcc",
lat_1=standardpar,
lon_0=centerlon,
)
x, y = m(lons, lats)
# setup figure size so each panel will have same aspect ratio as map.
xsize = rcParams["figure.figsize"][0]
ysize = (3.0 / 2.0) * m.aspect * xsize
fig = figure(figsize=(xsize, ysize))
yoffset = (m.urcrnry - m.llcrnry) / 30.0
for npanel, fcsthr in enumerate(arange(0, 72, 12)):
nt = fcsthrs.index(fcsthr)
ax = fig.add_subplot(320 + npanel + 1)
levels, colls = m.contour(x, y, t2m[nt, :, :], clevs, colors="k")
levels, colls = m.contourf(x, y, t2m[nt, :, :], clevs, cmap=cm.jet, colors=None)
m.drawcoastlines()
m.drawstates()
m.drawcountries()
m.drawparallels(arange(25, 75, 20), labels=[1, 0, 0, 0], fontsize=8, fontstyle="oblique")
m.drawmeridians(arange(-140, 0, 20), labels=[0, 0, 0, 1], fontsize=8, yoffset=yoffset, fontstyle="oblique")
# panel title
title(repr(fcsthr) + "-h forecast valid " + verifdates[nt], fontsize=12)
clim(clevs[1], clevs[-2])
# figure title
figtext(
0.5,
0.95,
u"2-m temp (\N{DEGREE SIGN}K) forecasts from %s" % verifdates[0],
horizontalalignment="center",
# draw the edge of the map projection region (the projection limb)
map.drawmapboundary()
# draw lat/lon grid lines every 30 degrees.
map.drawmeridians(p.arange(0,360,30))
map.drawparallels(p.arange(-90,90,30))
# lat/lon coordinates of five cities.
lats=[40.02,32.73,38.55,48.25,17.29]
lons=[-105.16,-117.16,-77.00,-114.21,-88.10]
cities=['Boulder, CO','San Diego, CA',
'Washington, DC','Whitefish, MT','Belize City, Belize']
# compute the native map projection coordinates for cities.
x,y = map(lons,lats)
# plot filled circles at the locations of the cities.
map.plot(x,y,'bo')
# plot the names of those five cities.
for name,xpt,ypt in zip(cities,x,y):
p.text(xpt+50000,ypt+50000,name,fontsize=9)
# make up some data on a regular lat/lon grid.
nlats = 73; nlons = 145; delta = 2.*p.pi/(nlons-1)
lats = (0.5*p.pi-delta*p.indices((nlats,nlons))[0,:,:])
lons = (delta*p.indices((nlats,nlons))[1,:,:])
wave = 0.75*(p.sin(2.*lats)**8*p.cos(4.*lons))
mean = 0.5*p.cos(2.*lats)*((p.sin(2.*lats))**2 + 2.)
# compute native map projection coordinates of lat/lon grid.
x, y = map(lons*180./p.pi, lats*180./p.pi)
# contour data over the map.
cs = map.contour(x,y,wave+mean,15,linewidths=1.5)
p.show()
#p.savefig('wiki_example.ps')
def plot_ncdf_output(id, datafn, **kwds):
"""Plot model field id from the data in netCDF file datafn.
Positional Input Parameter:
* id: Name of the id of the field to plot. String.
* datafn: Filename containing the output data to plot. String.
Input keyword parameter descriptions are found in the docstring
for Qtcm methods ploti, plotm, and other methods that call this
private method. In general, those methods take the keyword
parameters they receive and pass it along unchanged as keyword
parameters to this function. In that sense, this function is
seldom used as a stand-alone function, but rather is usually
used coupled with a Qtcm instance.
The data fields read in from the netCDF output file are dimensioned
(time, lat, lon). This is different than how the data is stored
in the compiled QTCM model fields (lon, lat, time), and at the
Python level (lon, lat). The reason this is the case is that
f2py automatically makes the arrays passed between the Python
and Fortran levels match.
For a lat vs. lon plot, the contour plot is superimposed onto
a cylindrical projection map of the Earth with continents drawn
and labeled meridians and parallels. The title also includes
the model time, and x- and y-axis labels are not drawn.
All numerical data used for plotting come from the netCDF output
file for consistency (e.g., the dimensions of u1). Currently
this method only works for 3-D data arrays (two in space, one
in time).
"""
#- Accomodate other ids. The id that this routine will use
# (i.e., iduse corresponds to the name in the netCDF output file)
# is called iduse. Set this to id, except for the case where
# some aliases of ids are entered in which iduse is the alias:
if id == 'Qc': iduse = 'Prec'
elif id == 'FLWut': iduse = 'OLR'
elif id == 'STYPE': iduse = 'stype'
else: iduse = id
#- Set defined keyword defaults. All are set to None except for
# nlatlon which gets an integer:
plotkwds_ids = [
'lat', 'lon', 'time', 'fn', 'levels', 'title', 'xlabel', 'ylabel',
'filled', 'nlatlon', 'tmppreview'
]
plotkwds = {}
for ikey in plotkwds_ids:
if kwds.has_key(ikey):
plotkwds[ikey] = copy.copy(kwds[ikey])
else:
plotkwds[ikey] = None
if not kwds.has_key('nlatlon'):
plotkwds['nlatlon'] = 8
#- Get data and dimensions of iduse to plot:
fileobj = S.NetCDFFile(datafn, mode='r')
data = N.array(fileobj.variables[iduse].getValue())
data_name = fileobj.variables[iduse].long_name
data_units = fileobj.variables[iduse].units
dim = {}
dimname = {}
dimunits = {}
dim['lat'] = N.array(fileobj.variables['lat'].getValue())
dimname['lat'] = fileobj.variables['lat'].long_name
dimunits['lat'] = fileobj.variables['lat'].units
dim['lon'] = N.array(fileobj.variables['lon'].getValue())
dimname['lon'] = fileobj.variables['lon'].long_name
dimunits['lon'] = fileobj.variables['lon'].units
dim['time'] = N.array(fileobj.variables['time'].getValue())
dimname['time'] = fileobj.variables['time'].long_name
dimunits['time'] = fileobj.variables['time'].units
fileobj.close()
#- Alter data long name to remove any units. The definition
# of units as the substring within the [] is the same as used in
# defVar in output.F90 of the compiled QTCM model. Remove
# underscores and extra whitespace in data_name and data_units,
# replacing with a single whitespace character between words:
idx1 = data_name.find('[')
idx2 = data_name.find(']')
if idx1 != -1 and idx2 != -1:
data_name = data_name[:idx1] + data_name[idx2 + 1:]
data_name = data_name.strip()
data_name = ' '.join(data_name.replace('_', ' ').split())
data_units = ' '.join(data_units.replace('_', ' ').split())
#- Alter dimension long name to remove any units. The definition
# of units as the substring within the [] is the same as used in
# defVar in output.F90 of the compiled QTCM model. Remove
# underscores and extra whitespace in name and units, replacing
# with a single whitespace character between words, and
# capitalizing like a title:
for idimkey in dim.keys():
idimname = dimname[idimkey]
idx1 = idimname.find('[')
idx2 = idimname.find(']')
if idx1 != -1 and idx2 != -1:
idimname = idimname[:idx1] + idimname[idx2 + 1:]
dimname[idimkey] = idimname.strip()
dimname[idimkey] = \
' '.join(dimname[idimkey].replace('_',' ').split()).title()
dimunits[idimkey] = \
' '.join(dimunits[idimkey].replace('_',' ').split()).title()
#- Some data checks:
if N.rank(data) != 3:
raise ValueError, '_plot: can only plot lat, lon, time fields'
if not N.allclose(dim['time'], N.sort(dim['time'])):
raise ValueError, '_plot: time not monotonically ascending'
if not N.allclose(dim['lat'], N.sort(dim['lat'])):
raise ValueError, '_plot: lat not monotonically ascending'
if not N.allclose(dim['lon'], N.sort(dim['lon'])):
raise ValueError, '_plot: lon not monotonically ascending'
if N.shape(data)[0] != N.size(dim['time']):
raise ValueError, '_plot: data time dim mismatch'
if N.shape(data)[1] != N.size(dim['lat']):
raise ValueError, '_plot: data lat dim mismatch'
if N.shape(data)[2] != N.size(dim['lon']):
raise ValueError, '_plot: data lon dim mismatch'
#- Choose and describe ranges for lat, lon, and time. The
# section cycles through the dictionary of dimensions. idim is
# the 1-D array of the values of that dimension. rngs is a
# dictionary where each entry corresponds to a dimension, and the
# value of the entry is the values of that dimension that are to
# be plotted. rngs_idxs are the indices in the original
# dimensions array corresponding to the values in rngs.
# keys_rngs_sizes_gt_1 is a list of the keys of ranges that have
# sizes greater than 1:
rngs = {}
rngs_idxs = {}
keys_rngs_sizes_gt_1 = []
for idimkey in dim.keys():
idim = dim[idimkey]
if plotkwds[idimkey] == None:
dim_mask = N.ones(N.size(idim), dtype=int)
elif N.isscalar(plotkwds[idimkey]):
dim_mask = where_close(idim, plotkwds[idimkey])
if N.sum(dim_mask) != 1:
raise ValueError, 'no point chosen'
elif (not N.isscalar(plotkwds[idimkey])) and \
N.size(plotkwds[idimkey]) == 1:
dim_mask = where_close(idim, plotkwds[idimkey][0])
if N.sum(dim_mask) != 1:
raise ValueError, 'no point chosen'
elif N.size(plotkwds[idimkey]) == 2:
dim_mask = N.logical_and(idim >= plotkwds[idimkey][0],
idim <= plotkwds[idimkey][-1])
else:
raise ValueError, 'bad dimension range keyword entry'
rngs[idimkey] = N.compress(dim_mask, idim)
rngs_idxs[idimkey] = N.compress(dim_mask, N.arange(N.size(idim)))
if N.size(rngs[idimkey]) > 1:
keys_rngs_sizes_gt_1.append(idimkey)
#- Set plot types (line or contour):
if len(keys_rngs_sizes_gt_1) == 0:
raise ValueError, 'cannot plot without any fixed dimension'
elif len(keys_rngs_sizes_gt_1) == 1:
plottype = 'line'
elif len(keys_rngs_sizes_gt_1) == 2:
plottype = 'contour'
else:
raise ValueError, 'cannot plot with > 2 varying dimensions'
#- Set plot axis fields and axis names, depending on what sort
# of dimensions will be plotted. If lon is to be plotted, it is
# always the x-axis. If lat is to be plotted, it is always the
# y-axis. In this section and later on in a few places, I
# rely on the count method for a list as a Boolean test: If it
# returns 0, consider that False; > 0 is True. The text for the
# title and axis labels to be passed to the xlabel, etc. methods,
# are called titlename, xname, and yname:
if plottype == 'line': #+ Choose x-axis vector and
x = rngs[keys_rngs_sizes_gt_1[0]] # x/y names for line plot
xname = dimname[keys_rngs_sizes_gt_1[0]] + ' [' \
+ dimunits[keys_rngs_sizes_gt_1[0]] + ']'
yname = data_name + ' [' + data_units + ']'
elif plottype == 'contour': #+ Choose axis vectors and
if keys_rngs_sizes_gt_1.count('lon'): # names for contour plot
x = rngs['lon']
xname = dimname['lon'] + ' [' + dimunits['lon'] + ']'
if keys_rngs_sizes_gt_1.count('lat'):
y = rngs['lat']
yname = dimname['lat'] + ' [' + dimunits['lat'] + ']'
if keys_rngs_sizes_gt_1.count('time'):
if keys_rngs_sizes_gt_1.count('lon'):
y = rngs['time']
yname = dimname['time'] + ' [' + dimunits['time'] + ']'
elif keys_rngs_sizes_gt_1.count('lat'):
x = rngs['time']
xname = dimname['time'] + ' [' + dimunits['time'] + ']'
else:
raise ValueError, 'bad treatment of time'
else:
raise ValueError, 'unrecognized plottype'
#- Override xname, yname, and titlename with keywords, if they
# are not None. titlename receives data_name and data_units
# by default:
if plotkwds['xlabel'] != None:
xname = plotkwds['xlabel']
if plotkwds['ylabel'] != None:
yname = plotkwds['ylabel']
if plotkwds['title'] != None:
titlename = plotkwds['title']
else:
titlename = data_name + ' [' + data_units + ']'
#- Pick data to be plotted and plot:
pylab.clf() #+ Clear any previous figures
pylab.figure(1) #+ Open a pylab figure
if plottype == 'line': #+ Select data for a line plot
y = data[rngs_idxs['time'], # and plot
rngs_idxs['lat'], rngs_idxs['lon']]
pylab.plot(x, y)
elif plottype == 'contour': #+ Select data for a contour
ritim = rngs_idxs['time'] # plot and plot
rilat = rngs_idxs['lat']
rilon = rngs_idxs['lon']
#* Extract subarrays depending on which two dimensions are
# chosen:
if N.size(rngs_idxs['time']) == 1:
zgrid = num.MLab.squeeze(data[ritim[0], rilat[0]:rilat[-1] + 1,
rilon[0]:rilon[-1] + 1])
elif N.size(rngs_idxs['lat']) == 1:
zgrid = num.MLab.squeeze(data[ritim[0]:ritim[-1] + 1, rilat[0],
rilon[0]:rilon[-1] + 1])
elif N.size(rngs_idxs['lon']) == 1:
zgrid = num.MLab.squeeze(data[ritim[0]:ritim[-1] + 1,
rilat[0]:rilat[-1] + 1, rilon[0]])
else:
raise ValueError, 'unrecognized configuration'
#* Change zgrid for special case of a lat. vs. time contour
# plot. Calculate xgrid and ygrid:
if keys_rngs_sizes_gt_1.count('time') and \
keys_rngs_sizes_gt_1.count('lat'):
zgrid = N.transpose(zgrid)
xgrid, ygrid = pylab.meshgrid(x, y)
#* Set contour levels:
if plotkwds['levels'] == None:
levels = nice_levels(zgrid)
else:
levels = plotkwds['levels']
#- Plot (creating continents first if is a lat vs. lon plot)
# and write contour levels/color bar as appropriate:
if keys_rngs_sizes_gt_1.count('lon') and \
keys_rngs_sizes_gt_1.count('lat'):
mapplot = Basemap(projection='cyl',
resolution='l',
llcrnrlon=N.min(xgrid),
llcrnrlat=N.min(ygrid),
urcrnrlon=N.max(xgrid),
urcrnrlat=N.max(ygrid))
mapplot.drawcoastlines()
mapplot.drawmeridians(nice_levels(rngs['lon'],
approx_nlev=plotkwds['nlatlon']),
labels=[1, 0, 0, 1])
mapplot.drawparallels(nice_levels(rngs['lat'],
approx_nlev=plotkwds['nlatlon']),
labels=[1, 0, 0, 1])
if plotkwds['filled']:
plot = mapplot.contourf(xgrid, ygrid, zgrid, levels)
pylab.colorbar(plot, orientation='horizontal', format='%g')
else:
plot = mapplot.contour(xgrid, ygrid, zgrid, levels)
pylab.clabel(plot, inline=1, fontsize=10, fmt='%g')
else:
if plotkwds['filled']:
plot = pylab.contourf(xgrid, ygrid, zgrid, levels)
pylab.colorbar(plot, orientation='horizontal', format='%g')
else:
plot = pylab.contour(xgrid, ygrid, zgrid, levels)
pylab.clabel(plot, inline=1, fontsize=10, fmt='%g')
else:
raise ValueError, 'unrecognized plottype'
#- Add titling. Lat vs. lon plots do not have axis labels because
# the map labels already make it clear, and for those plots the
# title also includes the time value:
if keys_rngs_sizes_gt_1.count('lon') and \
keys_rngs_sizes_gt_1.count('lat'):
titlename = titlename + ' at ' \
+ dimname['time'] + ' ' \
+ str(rngs['time'][0]) + ' ' \
+ dimunits['time']
titlename = mpl_latex_script1(titlename)
pylab.title(titlename)
else:
titlename = mpl_latex_script1(titlename)
xname = mpl_latex_script1(xname)
yname = mpl_latex_script1(yname)
pylab.xlabel(xname)
pylab.ylabel(yname)
pylab.title(titlename)
#- Output plot to PNG file or screen. The show command seems to
# have a problem on my Mac OS X, so save to a temporary file
# and use preview to view for fn == None and tmppreview set to
# True. Note that the temporary file is not deleted by this
# method:
if plotkwds['fn'] == None: #+ Screen display
if plotkwds['tmppreview'] and sys.platform == 'darwin':
outputfn = tempfile.mkstemp('.png', 'qtcm_')
pylab.savefig(outputfn[-1])
os.system('open -a /Applications/Preview.app ' + outputfn[-1])
else:
pylab.show()
elif type(plotkwds['fn']) == type('a'): #+ Write to file
pylab.savefig(plotkwds['fn'])
pylab.close(1)
else:
raise ValueError, 'cannot write to this type of file'
# read in data on lat/lon grid.
hgt = load('500hgtdata.gz')
lons = load('500hgtlons.gz')
lats = load('500hgtlats.gz')
# shift data so lons go from -180 to 180 instead of 0 to 360.
hgt,lons = shiftgrid(180.,hgt,lons,start=False)
lons, lats = meshgrid(lons, lats)
# create new figure
fig=figure()
# setup of sinusoidal basemap
m = Basemap(resolution='c',projection='sinu',lon_0=0)
ax = fig.add_axes([0.1,0.1,0.7,0.7])
# make a filled contour plot.
x, y = m(lons, lats)
CS = m.contour(x,y,hgt,15,linewidths=0.5,colors='k')
CS = m.contourf(x,y,hgt,15,cmap=cm.jet)
l,b,w,h=ax.get_position()
cax = axes([l+w+0.075, b, 0.05, h]) # setup colorbar axes
colorbar(drawedges=True, cax=cax) # draw colorbar
axes(ax) # make the original axes current again
# draw coastlines and political boundaries.
m.drawcoastlines()
m.drawmapboundary()
m.fillcontinents()
# draw parallels and meridians.
parallels = arange(-60.,90,30.)
m.drawparallels(parallels,labels=[1,0,0,0])
meridians = arange(-360.,360.,30.)
m.drawmeridians(meridians)
title('Sinusoidal Filled Contour Demo')
# create Basemap instance for Lambert Conformal Conic projection.
m = Basemap(llcrnrlon=llcrnrlon,llcrnrlat=llcrnrlat,
urcrnrlon=urcrnrlon,urcrnrlat=urcrnrlat,
rsphere=6371200.,
resolution='l',area_thresh=5000.,projection='lcc',
lat_1=standardpar,lon_0=centerlon)
x, y = m(lons, lats)
# create figure.
# mult ysize by 3/2 to take into account there will
# be three rows and two columns.
fig=m.createfigure(figsize=(8,(3./2.)*8))
yoffset = (m.urcrnry-m.llcrnry)/30.
for npanel,fcsthr in enumerate(arange(0,72,12)):
nt = fcsthrs.index(fcsthr)
ax = fig.add_subplot(320+npanel+1)
cs = m.contour(x,y,t2m[nt,:,:],clevs,colors='k')
cs = m.contourf(x,y,t2m[nt,:,:],clevs,cmap=cm.jet)
m.drawcoastlines()
m.drawstates()
m.drawcountries()
m.drawparallels(arange(25,75,20),labels=[1,0,0,0],fontsize=8,fontstyle='oblique')
m.drawmeridians(arange(-140,0,20),labels=[0,0,0,1],fontsize=8,yoffset=yoffset,fontstyle='oblique')
# panel title
title(repr(fcsthr)+'-h forecast valid '+verifdates[nt],fontsize=12)
# figure title
figtext(0.5,0.95,u"2-m temp (\N{DEGREE SIGN}K) forecasts from %s"%verifdates[0],
horizontalalignment='center',fontsize=14)
# a single colorbar.
cax = axes([0.1, 0.03, 0.8, 0.025])
colorbar(tickfmt='%d', cax=cax, orientation='horizontal',clabels=clevs[1:-1:2])
show()
