def plot_correction_models(calibration, correction_factor, cubic_coeff, quadratic_coeff, gp_x, gp_y):
plt.figure(figsize=(9,4))
plt.clf()
proj = pyproj.Proj("+proj=robin")
plot_reference_grid(proj, labels=False)
plot_proj = proj_wrapper(proj)
for ix,row in calibration.iterrows():
def plot_corrected(corrector_fn, *args, **kwargs):
lon, lat = corrector_fn(row["touch_lon"], row["touch_lat"])
tx, ty = plot_proj(row["target_lon"], row["target_lat"])
x, y = plot_proj(lon, lat)
thresh = 1e6
if abs(x-tx)<thresh and abs(y-ty)<thresh:
plt.plot([tx,x], [ty,y], *args, **kwargs)
plot_corrected(lambda x,y:polar_adjust(x,y,s=correction_factor), 'k')
plot_corrected(lambda x,y:polar_adjust(x,y,1), 'k', alpha=0.2)
plot_corrected(lambda x,y:quadratic_polar_adjust(x,y,quadratic_coeff), 'r')
plot_corrected(lambda x,y:cubic_polar_adjust(x,y,cubic_coeff), 'm')
plot_corrected(lambda x,y:gp_adjust(x,y,gp_x,gp_y), 'g')
plt.legend(handles=[mlines.Line2D([], [], color='k', alpha=0.2, linestyle="-",label='Original offset'),
mlines.Line2D([], [], color='k', linestyle="-",label='Constant corrected'),
mlines.Line2D([], [], color='r', linestyle="-",label='Quadratic corrected'),
mlines.Line2D([], [], color='m', linestyle="-",label='Cubic corrected'),
mlines.Line2D([], [], color='g', linestyle="-",label='GP corrected'),
], frameon=False, prop={'size':5})
def error_shading(calibration, fn):
"""Plot the GP offset model for x and y, after removing the constant correction"""
proj = proj_wrapper(pyproj.Proj("+proj=robin"))
plot_reference_grid(pyproj.Proj("+proj=robin"))
xs, ys, zs = [], [], []
for ix, row in calibration.iterrows():
tlon, tlat = row["target_lon"], row["target_lat"]
lon, lat = row["touch_lon"], row["touch_lat"]
# compute offset from constant corrected
px,py = proj(lon, lat)
lonc, latc = fn(lon, lat)
d = np.degrees(sphere.spherical_distance((tlon, tlat), (lonc, latc)))
xs.append(px)
ys.append(py)
zs.append(d)
xi = np.linspace(np.min(xs), np.max(xs), 500)
yi = np.linspace(np.min(ys), np.max(ys), 500)
zi = matplotlib.mlab.griddata(xs, ys, zs, xi, yi, interp='linear')
#xr, yr = np.meshgrid(xi,yi)
#iv = scipy.interpolate.interp2d(xs,ys,zs,kind='cubic')
#zi = iv(xi, yi)
#zi = scipy.interpolate.griddata((np.array(xs).flatten(),np.array(ys).flatten()),np.array(zs).flatten(),(xr,yr), method='cubic')
levels = np.linspace(np.min(zi), np.max(zi), 30)
plt.contourf(xi, yi, zi, cmap="bone", levels=levels, vmin=0, vmax=9)
plt.colorbar()
def gp_offset_plot(gp_x, gp_y):
plt.figure(figsize=(9,4))
plt.clf()
proj = pyproj.Proj("+proj=robin")
plot_reference_grid(proj, labels=True)
plot_proj = proj_wrapper(proj)
for ix,row in calibration.iterrows():
target = [row["touch_lon"], row["touch_lat"]]
corr_touch_lon, corr_touch_lat = gp_predict(gp_x, gp_y, row["touch_az_x"], row["touch_az_y"])
xp, yp = plot_proj(row["target_lon"],row["target_lat"])
xr, yr = plot_proj(row["touch_lon"],row["touch_lat"])
xq, yq = plot_proj(corr_touch_lon,corr_touch_lat)
plt.plot([xp,xq],[yp,yq],'k-')
plt.plot([xp,xr],[yp,yr],'k', alpha=0.1)
def gp_offset_shading(calibration, gp_x, gp_y, best_s):
"""Plot the GP offset model for x and y, after removing the constant correction"""
proj = proj_wrapper(pyproj.Proj("+proj=robin"))
plt.figure(figsize=(9,4))
plot_reference_grid(pyproj.Proj("+proj=robin"), labels=False)
xs, ys, zs = [], [], []
us, vs = [], []
test_pts = spiral_layout(2000)
for lon, lat in test_pts:
# gp corrected lon, lat
az_x,az_y = sphere.polar_to_az(lon, lat)
tinput = [az_x, az_y]
xc = gp_x.predict(tinput)
yc = gp_y.predict(tinput)
lonc, latc = sphere.az_to_polar(az_x+xc, az_y+yc)
# constant corrected lon, lat
lons, lats = polar_adjust(lon, lat, best_s)
# compute offset from constant corrected
px,py = proj(lon, lat)
px_c,py_c = proj(lonc, latc)
px_s,py_s = proj(lons, lats)
d = (px_s-px_c)**2 + (py_s-py_c)**2
xs.append(px)
ys.append(py)
if np.sqrt(d)<6e7:
us.append(px_s-px_c)
vs.append(py_s-py_c)
else:
us.append(0)
vs.append(0)
zs.append(np.sqrt(d))
xi = np.linspace(np.min(xs), np.max(xs), 500)
yi = np.linspace(np.min(ys), np.max(ys), 500)
#zi = matplotlib.mlab.griddata(xs, ys, zs, xi, yi, interp='linear')
#levels = np.linspace(np.min(zi), np.max(zi), 30)
#plt.contourf(xi, yi, zi, cmap="bone", levels=levels)
plt.quiver(xs,ys,us,vs,scale=3e7,width=5e-4)