["MSLP (hPa)", "Max 10m Wind (kts)"]):
# NOTE : Passing x_label messes up x axis formatting
plt.figure(metric)
tcv_plot_helper.plot_tcv_metric_vs_fhr(
forecast_track,
metric,
y_label=yLabel,
time_axis_parameter="epoch_zeta",
label=label,
color=colors[i])
#plt.legend()
#plt.figure("maxwind_value")
#plt.plot(maxwinds, color=colors[i], lw=line_width, label=label)
#plt.legend()
plt.figure("tracks")
bling.decorate_map(m)
# options worth testing: water_color='#99ffff',lake_color='#99ffff',
#continent_color='#cc9966'
# plot legend. The skip_duplicates arg must be True since the plot_track()
# function actually plots multiple lines for each track in order to vary
# the intensity of the color based on the max wind value
bling.create_simple_legend(skip_duplicates=True, position='upper right')
plt.savefig('tracks.png')
plt.figure("mslp_value")
plt.legend(loc='best')
plt.savefig("mslp.png")
plt.figure("maxwind_value")
def plot_track(lats, lons, **kwargs):
'''
Given a list of latitudes and a list of longitudes corresponding to
a hurricane track, plot the track using Basemap.
RETURNS
- the basemap object plotted on
REQUIRED ARGUMENTS
- lats - List of latitude points
- lons - List of longitude points
OPTIONAL KEYWORD ARGUMENTS (and their corresponding functionality):
- line_color - Color to use for the line. Default: black
- line_width - Width to use for the line: Default: 1.5
- line_style - Line style. Default: solid
- zorder - Z order of the line. Default: 0
- label - label to apply to the data being passed
- windspeeds - List of windspeeds corresponding to the lat/lon pairs.
If set, the shade of the track lines will be adjusted
according to the strength of the storm.
*NOTE: The scheme for this is currently very crude.
If you do not want this feature, do not
pass in the windspeeds*
- flagged_idc - List of indices in the lat/lon that are marked as
flagged. If passed in, these points will be indicated
using circular markers
- ax - Axes object to plot on. This overrides the 'basemap' argument.
- basemap - If set, use it as the Basemap object for plotting.
Otherwise instantiate a new one.
- extents - If set and basemap is not passed in, use these extents
for the map. Should be a 4-element list as follows:
[lowerLat, westLon, upperLat, eastLon]
If not passed in, use the global variables defined in
this module
- indicator_freq - If set, mark points corresponding to this
interval.
e.g. if passing in values ever 6 hours and you want
to indicate every 24 hours, set this to 4
- alpha - Alpha level to use for the line
- Plus, whatever arguments are accepted by the decorate_map function,
since it will be called afterwards
'''
# Set basic optional values
# array of arguments to pass to the plot() call
mpl_plot_kwargs = {}
if kwargs.has_key('line_color'):
line_color = kwargs['line_color']
else:
line_color = TRACK_LINE_COLOR
if kwargs.has_key('line_style'):
line_style = kwargs['line_style']
else:
line_style = '-'
if kwargs.has_key('zorder'):
zorder = kwargs['zorder']
else:
zorder = 10 # keep lines in front of land
if kwargs.has_key('alpha'):
line_alpha = kwargs['alpha']
else:
line_alpha = 1.0
if kwargs.has_key('line_width'):
line_width = kwargs['line_width']
else:
line_width = TRACK_LINE_WIDTH
if kwargs.has_key('label'):
mpl_plot_kwargs['label'] = kwargs['label']
# sanity checks
assert len(lats) == len(lons)
#import pdb ;pdb.set_trace()
# Set the axes
if kwargs.has_key('ax'):
ax = kwargs['ax']
else:
ax = gca()
# Instantiate/set the Basemap object
if kwargs.has_key('basemap'):
m = kwargs['basemap']
else:
if kwargs.has_key('extents'):
lowerLat = kwargs['extents'][0]
westLon = kwargs['extents'][1]
upperLat = kwargs['extents'][2]
eastLon = kwargs['extents'][3]
else:
lowerLat = SOUTH_CORNER
westLon = WEST_CORNER
upperLat = NORTH_CORNER
eastLon = EAST_CORNER
m = Basemap(llcrnrlon=westLon,
llcrnrlat=lowerLat,
urcrnrlon=eastLon,
urcrnrlat=upperLat,
projection=MAP_PROJECTION,
resolution=MAP_RESOLUTION,
area_thresh=AREA_THRESHOLD,
ax=ax)
#
# plot track
#
maxwinds = []
if kwargs.has_key('windspeeds') and kwargs['windspeeds'] is not None:
maxwinds = kwargs['windspeeds']
assert len(maxwinds) == len(lats)
rgb = ColorConverter().to_rgb(line_color)
max_maxwind = max(maxwinds)
# To support different colors depending on speed, we need to plot multiple
# small lines. Due to a technicality in Basemap::plot, we need to
# plot in increments of at least 3 points. As a result, the last two points'
# wind speed will not affect the line color
# NOTE That since all points (except the first and last 2) are being plotted
# multiple times, the effect of different colors/alphas is greatly diminished.
# a better temporary scheme is to only do each 3-point 'subline' once
# and use the average of the 3 points' maxwind values to determine the shading.
#for i in range(len(lats) - 2):
subline_length = 3
for i in range(0, len(lats) - 2, subline_length):
# if next subline would have less than 3 elements, extend the
# current subline correspondingly
if (i + subline_length) >= (len(lats) - (subline_length - 1)):
subline_length = subline_length + (len(lats) % subline_length)
if len(maxwinds) > 0:
#scale_factor = maxwinds[i] / max_maxwind
scale_factor = np.mean(
maxwinds[i:i + subline_length + 1]) / max_maxwind
# This scheme only works with black and other mixed colors...
# not the common red,green,blue, etc.
#currColor = [ max(0,c - scale_factor) for c in rgb]
#print currColor
# for now, just use alpha
if line_alpha != 1:
log.debug('Note: currently using a crude scheme for line color'\
' that just modifies the alpha')
currColor = line_color
line_alpha = scale_factor
log.debug(line_alpha)
else:
currColor = line_color
log.info("Commenting out code that sets line alpha according to speed")
'''
m.plot(lons[i:i+subline_length+1], lats[i:i+subline_length+1],
latlon=True, color=currColor,
linewidth=line_width, alpha=line_alpha, ls=line_style,
zorder=zorder,
**mpl_plot_kwargs)
'''
# uncomment below to do the color by speed
m.plot(lons,
lats,
latlon=True,
color=currColor,
linewidth=line_width,
alpha=line_alpha,
ls=line_style,
zorder=zorder,
**mpl_plot_kwargs)
# plot markers on flagged points - this is slow!
if kwargs.has_key('flagged_idc') and len(kwargs['flagged_idc']) > 0:
# Due to an issue with Basemap, cannot indicate flagged
# indices unless there are at least 3 of them, so pad the list
if len(kwargs['flagged_idc']) < 3:
flagged_idc = kwargs['flagged_idc']
flagged_idc.append(flagged_idc[0])
flagged_idc.append(flagged_idc[0])
flaggedLats = []
flaggedLons = []
for x in kwargs['flagged_idc']:
flaggedLats.append(lats[x])
flaggedLons.append(lons[x])
m.scatter(
flaggedLons,
flaggedLats,
latlon=True,
s=35.0,
#facecolors='none', edgecolors=line_color)
facecolors='red',
edgecolors=line_color)
# plot markers every 24 hours
if kwargs.has_key('indicator_freq'):
for i in kwargs['indicator_freq']:
sz = line_width * 8
m.scatter(lons[i], lats[i], latlon=True, color=line_color, s=sz)
bling.decorate_map(m, **kwargs)
return m