def _plot_data(data, parameters):
storm_dir, storm_spd = parameters['storm_motion']
bl_dir, bl_spd = parameters['bunkers_left']
br_dir, br_spd = parameters['bunkers_right']
u, v = vec2comp(data['wind_dir'], data['wind_spd'])
alt = data['altitude']
storm_u, storm_v = vec2comp(storm_dir, storm_spd)
bl_u, bl_v = vec2comp(bl_dir, bl_spd)
br_u, br_v = vec2comp(br_dir, br_spd)
seg_idxs = np.searchsorted(alt, _seg_hghts)
seg_u = np.interp(_seg_hghts, alt, u, left=np.nan, right=np.nan)
seg_v = np.interp(_seg_hghts, alt, v, left=np.nan, right=np.nan)
ca_u = np.interp([0.5], alt, u, left=np.nan, right=np.nan)
ca_v = np.interp([0.5], alt, v, left=np.nan, right=np.nan)
for idx in xrange(len(_seg_hghts) - 1):
idx_start = seg_idxs[idx]
idx_end = seg_idxs[idx + 1]
if not np.isnan(seg_u[idx]):
pylab.plot([seg_u[idx], u[idx_start]], [seg_v[idx], v[idx_start]], '-', color=_seg_colors[idx], linewidth=1.5)
if idx_start < len(data['rms_error']) and data['rms_error'][idx_start] == 0.:
# The first segment is to the surface wind, draw it in a dashed line
pylab.plot(u[idx_start:(idx_start + 2)], v[idx_start:(idx_start + 2)], '--', color=_seg_colors[idx], linewidth=1.5)
pylab.plot(u[(idx_start + 1):idx_end], v[(idx_start + 1):idx_end], '-', color=_seg_colors[idx], linewidth=1.5)
else:
pylab.plot(u[idx_start:idx_end], v[idx_start:idx_end], '-', color=_seg_colors[idx], linewidth=1.5)
if not np.isnan(seg_u[idx + 1]):
pylab.plot([u[idx_end - 1], seg_u[idx + 1]], [v[idx_end - 1], seg_v[idx + 1]], '-', color=_seg_colors[idx], linewidth=1.5)
for upt, vpt, rms in zip(u, v, data['rms_error'])[idx_start:idx_end]:
rad = np.sqrt(2) * rms
circ = Circle((upt, vpt), rad, color=_seg_colors[idx], alpha=0.05)
pylab.gca().add_patch(circ)
pylab.plot([storm_u, u[0], ca_u], [storm_v, v[0], ca_v], 'c-', linewidth=0.75)
if not (np.isnan(bl_u) or np.isnan(bl_v)):
pylab.plot(bl_u, bl_v, 'ko', markersize=5, mfc='none')
pylab.text(bl_u + 0.5, bl_v - 0.5, "LM", ha='left', va='top', color='k', fontsize=10)
if not (np.isnan(br_u) or np.isnan(br_v)):
pylab.plot(br_u, br_v, 'ko', markersize=5, mfc='none')
pylab.text(br_u + 0.5, br_v - 0.5, "RM", ha='left', va='top', color='k', fontsize=10)
if not (np.isnan(storm_u) or np.isnan(storm_v)) and storm_u != bl_u and storm_v != bl_v and storm_u != br_u and storm_v != br_v:
pylab.plot(storm_u, storm_v, 'k+', markersize=6)
pylab.text(storm_u + 0.5, storm_v - 0.5, "SM", ha='left', va='top', color='k', fontsize=10)
def _plot_vwp_data(data):
ivals = [x * ((1 - x_start) / (len(data))) for x in range(0, len(data))]
knt = 0
for iline in ivals:
u, v = vec2comp(data[knt]['wind_dir'], data[knt]['wind_spd'])
alt = data[knt]['altitude']
x = np.empty_like(alt)
x.fill(iline)
mpl.pyplot.barbs(x + x_start, (alt / max_alt) + 0.03,
u,
v,
data[knt]['wind_spd'],
length=6,
cmap=_vwp_cols,
clim=(_vwp_levs[0], _vwp_levs[-1]),
transform=pylab.gca().transAxes,
clip_on=True,
zorder=4,
linewidth=1.)
# If the RMS exceeds _bad_rms, highlight with a red circle. Also, plot
# the wind speeds as color-coded text next to the barbs
for klev in range(0, len(data[knt]['wind_spd'])):
rms = data[knt]['rms_error'][klev]
x_loc = x[klev] + (x_start + 0.002)
y_loc = (alt[klev] / max_alt) + 0.02
if rms >= _bad_rms:
ring = Circle((x_loc, y_loc + 0.01),
0.01,
linestyle='solid',
fc='none',
ec='red',
linewidth=2)
pylab.gca().add_patch(ring)
spd = data[knt]['wind_spd'][klev]
text_spd = roundup(spd)
spd_idx = np.where(text_spd > _vwp_levs)[0]
if len(spd_idx) == 0:
spd_idx = 0
else:
spd_idx = spd_idx[-1]
spd_idx = np.clip(spd_idx, 0, len(_vwp_colors) - 1)
pylab.text(x_loc,
y_loc,
str(int(spd)),
transform=pylab.gca().transAxes,
fontsize=9,
color=_vwp_colors[spd_idx],
va='bottom')
knt += 1
def plot_hodograph(data,
parameters,
fname=None,
web=False,
fixed=False,
archive=False):
img_title = "%s VWP valid %s" % (
data.rid, data['time'].strftime("%d %b %Y %H%M UTC"))
if fname is not None:
img_file_name = fname
else:
img_file_name = "%s_vad.png" % data.rid
u, v = vec2comp(data['wind_dir'], data['wind_spd'])
sat_age = 6 * 3600
if fixed or len(u) == 0:
ctr_u, ctr_v = 20, 20
size = 120
else:
ctr_u = u.mean()
ctr_v = v.mean()
size = max(u.max() - u.min(), v.max() - v.min()) + 20
size = max(120, size)
min_u = ctr_u - size / 2
max_u = ctr_u + size / 2
min_v = ctr_v - size / 2
max_v = ctr_v + size / 2
now = datetime.utcnow()
img_age = now - data['time']
age_cstop = min(_total_seconds(img_age) / sat_age, 1) * 0.4
age_color = mpl.cm.get_cmap('hot')(age_cstop)[:-1]
age_str = "Image created on %s (%s old)" % (
now.strftime("%d %b %Y %H%M UTC"), _fmt_timedelta(img_age))
pylab.figure(figsize=(10, 7.5), dpi=150)
fig_wid, fig_hght = pylab.gcf().get_size_inches()
fig_aspect = fig_wid / fig_hght
axes_left = 0.05
axes_bot = 0.05
axes_hght = 0.9
axes_wid = axes_hght / fig_aspect
pylab.axes((axes_left, axes_bot, axes_wid, axes_hght))
_plot_background(min_u, max_u, min_v, max_v)
_plot_data(data, parameters)
_plot_param_table(parameters, web=web)
pylab.xlim(min_u, max_u)
pylab.ylim(min_v, max_v)
pylab.xticks([])
pylab.yticks([])
if not archive:
pylab.title(img_title, color=age_color)
pylab.text(0.,
-0.01,
age_str,
transform=pylab.gca().transAxes,
ha='left',
va='top',
fontsize=9,
color=age_color)
else:
pylab.title(img_title)
if web:
web_brand = "http://www.autumnsky.us/vad/"
pylab.text(1.0,
-0.01,
web_brand,
transform=pylab.gca().transAxes,
ha='right',
va='top',
fontsize=9)
pylab.savefig(img_file_name, dpi=pylab.gcf().dpi)
pylab.close()
if web:
bounds = {
'min_u': min_u,
'max_u': max_u,
'min_v': min_v,
'max_v': max_v
}
print(json.dumps(bounds))
def _plot_data(data, parameters):
storm_dir, storm_spd = parameters['storm_motion']
bl_dir, bl_spd = parameters['bunkers_left']
br_dir, br_spd = parameters['bunkers_right']
mn_dir, mn_spd = parameters['mean_wind']
u, v = vec2comp(data['wind_dir'], data['wind_spd'])
alt = data['altitude']
storm_u, storm_v = vec2comp(storm_dir, storm_spd)
bl_u, bl_v = vec2comp(bl_dir, bl_spd)
br_u, br_v = vec2comp(br_dir, br_spd)
mn_u, mn_v = vec2comp(mn_dir, mn_spd)
seg_idxs = np.searchsorted(alt, _seg_hghts)
try:
seg_u = np.interp(_seg_hghts, alt, u, left=np.nan, right=np.nan)
seg_v = np.interp(_seg_hghts, alt, v, left=np.nan, right=np.nan)
ca_u = np.interp(0.5, alt, u, left=np.nan, right=np.nan)
ca_v = np.interp(0.5, alt, v, left=np.nan, right=np.nan)
except ValueError:
seg_u = np.nan * np.array(_seg_hghts)
seg_v = np.nan * np.array(_seg_hghts)
ca_u = np.nan
ca_v = np.nan
mkr_z = np.arange(16)
mkr_u = np.interp(mkr_z, alt, u, left=np.nan, right=np.nan)
mkr_v = np.interp(mkr_z, alt, v, left=np.nan, right=np.nan)
for idx in range(len(_seg_hghts) - 1):
idx_start = seg_idxs[idx]
idx_end = seg_idxs[idx + 1]
if not np.isnan(seg_u[idx]):
pylab.plot([seg_u[idx], u[idx_start]], [seg_v[idx], v[idx_start]],
'-',
color=_seg_colors[idx],
linewidth=1.5)
if idx_start < len(
data['rms_error']) and data['rms_error'][idx_start] == 0.:
# The first segment is to the surface wind, draw it in a dashed line
pylab.plot(u[idx_start:(idx_start + 2)],
v[idx_start:(idx_start + 2)],
'--',
color=_seg_colors[idx],
linewidth=1.5)
pylab.plot(u[(idx_start + 1):idx_end],
v[(idx_start + 1):idx_end],
'-',
color=_seg_colors[idx],
linewidth=1.5)
else:
pylab.plot(u[idx_start:idx_end],
v[idx_start:idx_end],
'-',
color=_seg_colors[idx],
linewidth=1.5)
if not np.isnan(seg_u[idx + 1]):
pylab.plot([u[idx_end - 1], seg_u[idx + 1]],
[v[idx_end - 1], seg_v[idx + 1]],
'-',
color=_seg_colors[idx],
linewidth=1.5)
for upt, vpt, rms in list(zip(u, v,
data['rms_error']))[idx_start:idx_end]:
rad = np.sqrt(2) * rms
circ = Circle((upt, vpt), rad, color=_seg_colors[idx], alpha=0.05)
pylab.gca().add_patch(circ)
pylab.plot(mkr_u, mkr_v, 'ko', ms=10)
for um, vm, zm in zip(mkr_u, mkr_v, mkr_z):
if not np.isnan(um):
pylab.text(um,
vm - 0.1,
str(zm),
va='center',
ha='center',
color='white',
size=6.5,
fontweight='bold')
try:
pylab.plot([storm_u, u[0]], [storm_v, v[0]], 'c-', linewidth=0.75)
pylab.plot([u[0], ca_u], [v[0], ca_v], 'm-', linewidth=0.75)
except IndexError:
pass
if not (np.isnan(bl_u) or np.isnan(bl_v)):
pylab.plot(bl_u, bl_v, 'ko', markersize=5, mfc='none')
pylab.text(bl_u + 0.5,
bl_v - 0.5,
"LM",
ha='left',
va='top',
color='k',
fontsize=10)
if not (np.isnan(br_u) or np.isnan(br_v)):
pylab.plot(br_u, br_v, 'ko', markersize=5, mfc='none')
pylab.text(br_u + 0.5,
br_v - 0.5,
"RM",
ha='left',
va='top',
color='k',
fontsize=10)
if not (np.isnan(mn_u) or np.isnan(mn_v)):
pylab.plot(mn_u, mn_v, 's', color='#a04000', markersize=5, mfc='none')
pylab.text(mn_u + 0.6,
mn_v - 0.6,
"MEAN",
ha='left',
va='top',
color='#a04000',
fontsize=10)
smv_is_brm = (storm_u == br_u and storm_v == br_v)
smv_is_blm = (storm_u == bl_u and storm_v == bl_v)
smv_is_mnw = (storm_u == mn_u and storm_v == mn_v)
if not (np.isnan(storm_u) or np.isnan(storm_v)) and not (
smv_is_brm or smv_is_blm or smv_is_mnw):
pylab.plot(storm_u, storm_v, 'k+', markersize=6)
pylab.text(storm_u + 0.5,
storm_v - 0.5,
"SM",
ha='left',
va='top',
color='k',
fontsize=10)
def _plot_data(data, parameters):
storm_dir, storm_spd = parameters['storm_motion']
bl_dir, bl_spd = parameters['bunkers_left']
br_dir, br_spd = parameters['bunkers_right']
u, v = vec2comp(data['wind_dir'], data['wind_spd'])
alt = data['altitude']
storm_u, storm_v = vec2comp(storm_dir, storm_spd)
bl_u, bl_v = vec2comp(bl_dir, bl_spd)
br_u, br_v = vec2comp(br_dir, br_spd)
seg_idxs = np.searchsorted(alt, _seg_hghts)
seg_u = np.interp(_seg_hghts, alt, u, left=np.nan, right=np.nan)
seg_v = np.interp(_seg_hghts, alt, v, left=np.nan, right=np.nan)
ca_u = np.interp([0.5], alt, u, left=np.nan, right=np.nan)
ca_v = np.interp([0.5], alt, v, left=np.nan, right=np.nan)
for idx in xrange(len(_seg_hghts) - 1):
idx_start = seg_idxs[idx]
idx_end = seg_idxs[idx + 1]
if not np.isnan(seg_u[idx]):
pylab.plot([seg_u[idx], u[idx_start]], [seg_v[idx], v[idx_start]],
'-',
color=_seg_colors[idx],
linewidth=1.5)
if idx_start < len(
data['rms_error']) and data['rms_error'][idx_start] == 0.:
# The first segment is to the surface wind, draw it in a dashed line
pylab.plot(u[idx_start:(idx_start + 2)],
v[idx_start:(idx_start + 2)],
'--',
color=_seg_colors[idx],
linewidth=1.5)
pylab.plot(u[(idx_start + 1):idx_end],
v[(idx_start + 1):idx_end],
'-',
color=_seg_colors[idx],
linewidth=1.5)
else:
pylab.plot(u[idx_start:idx_end],
v[idx_start:idx_end],
'-',
color=_seg_colors[idx],
linewidth=1.5)
if not np.isnan(seg_u[idx + 1]):
pylab.plot([u[idx_end - 1], seg_u[idx + 1]],
[v[idx_end - 1], seg_v[idx + 1]],
'-',
color=_seg_colors[idx],
linewidth=1.5)
for upt, vpt, rms in zip(u, v, data['rms_error'])[idx_start:idx_end]:
rad = np.sqrt(2) * rms
circ = Circle((upt, vpt), rad, color=_seg_colors[idx], alpha=0.05)
pylab.gca().add_patch(circ)
pylab.plot([storm_u, u[0], ca_u], [storm_v, v[0], ca_v],
'c-',
linewidth=0.75)
if not (np.isnan(bl_u) or np.isnan(bl_v)):
pylab.plot(bl_u, bl_v, 'ko', markersize=5, mfc='none')
pylab.text(bl_u + 0.5,
bl_v - 0.5,
"LM",
ha='left',
va='top',
color='k',
fontsize=10)
if not (np.isnan(br_u) or np.isnan(br_v)):
pylab.plot(br_u, br_v, 'ko', markersize=5, mfc='none')
pylab.text(br_u + 0.5,
br_v - 0.5,
"RM",
ha='left',
va='top',
color='k',
fontsize=10)
if not (
np.isnan(storm_u) or np.isnan(storm_v)
) and storm_u != bl_u and storm_v != bl_v and storm_u != br_u and storm_v != br_v:
pylab.plot(storm_u, storm_v, 'k+', markersize=6)
pylab.text(storm_u + 0.5,
storm_v - 0.5,
"SM",
ha='left',
va='top',
color='k',
fontsize=10)
def plot_vwp(data,
times,
parameters,
fname=None,
add_hodo=False,
fixed=False,
web=False,
archive=False):
img_title = "%s VWP valid ending %s" % (
data[0].rid, times[0].strftime("%d %b %Y %H%M UTC"))
if fname is not None:
img_file_name = fname
else:
img_file_name = "%s_vwp.png" % data[0].rid
sat_age = 6 * 3600
now = datetime.utcnow()
img_age = now - times[0]
age_cstop = min(_total_seconds(img_age) / sat_age, 1) * 0.4
age_color = mpl.cm.get_cmap('hot')(age_cstop)[:-1]
age_str = "Image created on %s (%s old)" % (
now.strftime("%d %b %Y %H%M UTC"), _fmt_timedelta(img_age))
fig_aspect = 2.5714
fig_wid = 24
fig_hght = fig_wid / fig_aspect
pylab.figure(figsize=(fig_wid, fig_hght), dpi=200)
axes_left = 0.01
axes_bot = 0.02
axes_hght = 0.94
axes_wid = axes_hght / fig_aspect
pylab.axes((axes_left, axes_bot, 0.99, axes_hght))
_plot_vwp_background(times)
_plot_vwp_data(data)
#_plot_param_table(parameters, web=web)
pylab.xlim(0, 1.)
pylab.ylim(0, 1.)
pylab.xticks([])
pylab.yticks([])
pylab.box(False)
if not archive:
pylab.title(img_title, color=age_color)
pylab.text(x_start,
1.03,
age_str,
transform=pylab.gca().transAxes,
ha='left',
va='top',
fontsize=9,
color=age_color)
else:
pylab.title(img_title)
if web:
web_brand = "http://www.autumnsky.us/vad/"
pylab.text(1.0,
-0.01,
web_brand,
transform=pylab.gca().transAxes,
ha='right',
va='top',
fontsize=9)
if add_hodo:
inset_ax = inset_axes(pylab.gca(),
width="30%",
height="55%",
loc='upper left',
bbox_to_anchor=(0.63, 0, 0.85, 1),
bbox_transform=pylab.gca().transAxes)
u, v = vec2comp(data[0]['wind_dir'], data[0]['wind_spd'])
if fixed or len(u) == 0:
ctr_u, ctr_v = 20, 20
size = 120
else:
ctr_u = u.mean()
ctr_v = v.mean()
size = max(u.max() - u.min(), v.max() - v.min()) + 20
size = max(120, size)
min_u = ctr_u - size / 2
max_u = ctr_u + size / 2
min_v = ctr_v - size / 2
max_v = ctr_v + size / 2
_plot_background(min_u, max_u, min_v, max_v)
_plot_data(data[0], parameters)
_plot_param_table(parameters, web=web)
inset_ax.set_xlim(min_u, max_u)
inset_ax.set_ylim(min_v, max_v)
inset_ax.set_xticks([])
inset_ax.set_yticks([])
pylab.savefig(img_file_name, dpi=pylab.gcf().dpi)
pylab.close()
if web:
bounds = {
'min_u': min_u,
'max_u': max_u,
'min_v': min_v,
'max_v': max_v
}
print(json.dumps(bounds))