def cross_section(self,**kwargs):
field_array=kwargs['field']
''' calculate wind components'''
u_array=self.u_array
v_array=self.v_array
''' along cross section assuming self.azimuth within [0, 90] '''
wind_dir_section = self.azimuth -180.
wx = u_array*np.sin(wind_dir_section*np.pi/180.)
wy = v_array*np.cos(wind_dir_section*np.pi/180.)
wind_array = -(wx+wy)
''' perpendicular to cross section '''
orthogonal_dir_section =wind_dir_section + 90.
qx = u_array*np.sin(orthogonal_dir_section*np.pi/180.)
qy = v_array*np.cos(orthogonal_dir_section*np.pi/180.)
orth_array = (qx+qy)
self.slice_type='cross_section'
self.set_panel(option=self.slice_type,wind=False)
figsize=self.figure_size['vertical']
''' get indices of starting and ending coordinates '''
latix_0=cm.find_index_recursively(array=self.lats,value=self.slice[0][0],decimals=2)
lonix_0=cm.find_index_recursively(array=self.lons,value=self.slice[0][1],decimals=2)
latix_1=cm.find_index_recursively(array=self.lats,value=self.slice[1][0],decimals=2)
lonix_1=cm.find_index_recursively(array=self.lons,value=self.slice[1][1],decimals=2)
''' create grid for the entire domain '''
xx = np.arange(0,self.axesval['x'].size)
yy = np.arange(0,self.axesval['y'].size)
zz = np.arange(0,self.axesval['z'].size)
xm,ym,zm = np.meshgrid(xx,yy,zz)
''' convert grid and plotting field to vector columns'''
xd=np.reshape(xm,[1,xm.size]).tolist()
yd=np.reshape(ym,[1,ym.size]).tolist()
zd=np.reshape(zm,[1,zm.size]).tolist()
xd=xd[0]
yd=yd[0]
zd=zd[0]
''' specify grid for interpolated cross section '''
hres = 100
vres = 44
xi = np.linspace(lonix_0, lonix_1, hres)
yi = np.linspace(latix_0, latix_1, hres)
zi = np.linspace(0, 43, vres)
zi=np.array([zi,]*hres)
ki=np.ma.empty([vres,hres])
wi=np.ma.empty([vres,hres])
qi=np.ma.empty([vres,hres])
# fillv = field_array.fill_value
''' convert to standard numpy array (not masked)
and replace fill values for nans '''
kd=np.ma.filled(field_array, fill_value=np.nan)
wd=np.ma.filled(wind_array, fill_value=np.nan)
qd=np.ma.filled(orth_array, fill_value=np.nan)
''' convert to 1 column array '''
kd=np.reshape(kd,[kd.size,1])
wd=np.reshape(wd,[wd.size,1])
qd=np.reshape(qd,[qd.size,1])
''' create kdTree with entire domain'''
coords=zip(xd,yd,zd)
tree = cKDTree(coords)
''' interpolate using kdTree (nearest neighbor)
and averaging the neighborhood
'''
neigh = 8
for k in range(vres):
coords = zip(yi, xi, zi[:,k])
dist, idx = tree.query( coords, k=neigh, eps=0, p=1, distance_upper_bound=10)
kd_mean = np.nanmean(kd[idx],axis=1)
wd_mean = np.nanmean(wd[idx],axis=1)
qd_mean = np.nanmean(qd[idx],axis=1)
ki[k,:]=kd_mean.T
wi[k,:]=wd_mean.T
qi[k,:]=qd_mean.T
component = [wi, qi]
comptitle = ['Along-section wind speed [m s-1] (contours)\n',
'Cross-section wind speed [m s-1] (contours)\n']
for n in range(2):
"""make plot with wind speed along cross section """
with sns.axes_style("white"):
fig,ax = plt.subplots(figsize=(8,11*0.5))
''' add field as image '''
zsynth = self.axesval['z']
gate_hgt=0.25 #[km]
extent = [0, self.distance, zsynth[0]-gate_hgt/2., zsynth[-1]+gate_hgt/2.]
# im, cmap, norm = self.add_field(ax,array=ki,field=self.var, extent=extent)
im, cmap, norm = self.add_field2(ax,array=ki,field=self.var, extent=extent)
''' add terrain profiel '''
prof = Terrain.get_altitude_profile(self)
prof = np.asarray(prof)
self.add_terrain_profile(ax, prof ,None)
self.terrain.array['profile']=prof
''' add contour of wind section '''
X,Y = np.meshgrid(np.linspace(0, self.distance, hres), zsynth)
sigma=0.5
section = gaussian_filter(component[n], sigma,mode='nearest')
cs = ax.contour(X,Y,section,colors='k',
linewidths=1.5, levels=range(-4,26,2))
# cs = ax.contour(X,Y,component[n],colors='k',linewidths=0.5, levels=range(-4,26,2))
ax.clabel(cs, fontsize=12, fmt='%1.0f',)
ax.set_yticks(zsynth[1::2])
ytlabels = ["{:3.1f}".format(z) for z in zsynth[1::2]]
ax.set_yticklabels(ytlabels)
ax.set_ylim(0. , 7.5)
ax.set_xlim(0. , self.distance)
legname = os.path.basename(self.file)
ta=ax.transAxes
ax.text(0.05,0.9, legname[:3].upper() + " " + legname[3:5], transform = ta,weight='bold')
y,x = zip(*self.slice)
stlat='{:3.2f}'.format(y[0])
stlon='{:3.2f}'.format(x[0])
ax.text(0.05,0.85, 'start: ('+stlat+','+stlon+')', transform = ta)
enlat='{:3.2f}'.format(y[-1])
enlon='{:3.2f}'.format(x[-1])
ax.text(0.05,0.8, 'end: ('+enlat+','+enlon+')', transform = ta)
staz='{:3.1f}'.format(self.azimuth)
ax.text(0.05, 0.75, 'az: '+staz, transform=ta)
ax.set_xlabel('Distance along cross section [km]')
ax.set_ylabel('Altitude [km]')
if self.verticalGridMajorOn:
ax.grid(True, which = 'major',linewidth=1)
if self.verticalGridMinorOn:
ax.grid(True, which = 'minor',alpha=0.5)
ax.minorticks_on()
''' add color bar '''
fig.colorbar(im,cmap=cmap, norm=norm)
''' add title '''
titext='Dual-Doppler Synthesis: '+ self.get_var_title(self.var)+' (color coded)\n'
titext=titext+comptitle[n]
line_start='Start time: '+self.synth_start.strftime('%Y-%m-%d %H:%M')+' UTC\n'
line_end='End time: '+self.synth_end.strftime('%Y-%m-%d %H:%M')+' UTC'
fig.suptitle(titext+line_start+line_end)
plt.subplots_adjust(top=0.85,right=1.0)
plt.draw()
return ki,component
def vertical_plane(self,**kwargs):
field_array=None
for key,value in kwargs.iteritems():
if key == 'field':
isWind=False
field_array=value
if key == 'spd':
isWind=True
windname=value
elif key == 'sliceo':
self.sliceo=value
u_array=self.u_array
v_array=self.v_array
w_array=self.w_array
self.slice_type='vertical'
self.set_panel(option=self.slice_type,wind=isWind)
figsize=self.figure_size['vertical']
fig = plt.figure(figsize=figsize)
plot_grids=ImageGrid( fig,111,
nrows_ncols = self.rows_cols,
axes_pad = 0.0,
add_all = True,
share_all=False,
label_mode = "L",
cbar_location = "top",
cbar_mode="single",
aspect=True)
""" get list with slices """
uComp = self.get_slices(u_array)
vComp = self.get_slices(v_array)
wComp = self.get_slices(w_array)
profiles = Terrain.get_altitude_profile(self)
if isWind:
if windname == 'u':
field_group = uComp
colorName='U'
varName=colorName
elif windname == 'v':
field_group = vComp
colorName='V'
varName=colorName
elif windname == 'w':
field_group = wComp
colorName='WVA'
varName=colorName
else:
field_group = self.get_slices(field_array)
varName=self.var
''' field extent '''
extent1=self.get_extent()
''' if zoomOpt is false then extent1=extent2 '''
if self.zoomOpt:
opt=self.zoomOpt[0]
extent2=cm.zoom_in(self,extent1,self.zoomCenter[opt])
else:
extent2=extent1
''' scale for horizontal axis'''
self.scale=20
''' adjust vertical extent '''
if self.sliceo=='meridional':
extent3=cm.adjust_extent(self,extent1,'meridional','data')
extent4=cm.adjust_extent(self,extent2,'meridional','detail')
horizontalComp=vComp
geo_axis='Lon: '
elif self.sliceo=='zonal':
extent3=cm.adjust_extent(self,extent1,'zonal','data')
extent4=cm.adjust_extent(self,extent2,'zonal','detail')
horizontalComp=uComp
geo_axis='Lat: '
"""creates iterator group """
group=zip(plot_grids,
field_group,
horizontalComp,
wComp,
profiles['altitude'],profiles['axis'])
"""make gridded plot """
p=0
for g,field,h_comp,w_comp,prof,profax in group:
if isWind:
im, cmap, norm = self.add_field(g,
array=field.T,
field=varName,
name=colorName,
ext=extent3)
else:
im, cmap, norm = self.add_field(g,
array=field.T,
field=varName,
name=self.var,
ext=extent3)
self.add_terrain_profile(g,prof,profax)
if self.wind and not isWind:
self.add_windvector(g,h_comp.T,w_comp.T)
self.add_slice_line(g)
g.set_xlim(extent4[0], extent4[1])
g.set_ylim(extent4[2], extent4[3])
if p == 0:
self.match_horizontal_grid(g)
self.adjust_ticklabels(g)
if self.verticalGridMajorOn:
g.grid(True, which = 'major',linewidth=1)
if self.verticalGridMinorOn:
g.grid(True, which = 'minor',alpha=0.5)
g.minorticks_on()
if self.sliceo=='meridional':
geotext=geo_axis+str(self.slicem[p])
elif self.sliceo=='zonal':
geotext=geo_axis+str(self.slicez[p])
g.text( 0.03, 0.9,
geotext,
fontsize=self.zlevel_textsize,
horizontalalignment='left',
verticalalignment='center',
transform=g.transAxes)
p+=1
# add color bar
plot_grids.cbar_axes[0].colorbar(im,cmap=cmap, norm=norm)
# add title
titext='Dual-Doppler Synthesis: '+ self.get_var_title(varName)+'\n'
line_start='\nStart time: '+self.synth_start.strftime('%Y-%m-%d %H:%M')+' UTC'
line_end='\nEnd time: '+self.synth_end.strftime('%Y-%m-%d %H:%M')+' UTC'
fig.suptitle(titext+self.file+line_start+line_end)
# show figure
plt.draw()
