def display_trend_plots(func, kwargs):
"""
调用func函数绘制单个模式不同起报时间, 同一预报时刻的对比.
"""
# prepare function keyword
initTime = kwargs.pop('initTime')
fhour = kwargs.pop('fhour')
frange = kwargs.pop('frange')
runBack = kwargs.pop('runBack')
runStep = kwargs.pop('runStep')
# prepare initTime and forecast hours
initTime = func(getInitTime=True)
fhours = check_frange(frange)
if fhours is None:
fhours = [fhour]
# get the plots
allPlots = []
allLabels = []
for fh in fhours:
validTime = initTime + timedelta(hours=fh)
for irun in range(runBack):
kwargs['initTime'] = initTime - timedelta(hours=runStep * irun)
kwargs['fhour'] = fh + runStep * irun
plot = func(**kwargs)
if plot is None: continue
allPlots.append(plot)
allLabels.append(validTime.strftime('%y%m%d%H'))
# display in jupyter
try:
# code run in jupyter notebook or not
get_ipython()
allPlots = np.asarray(allPlots, dtype=np.object)
allLabels = np.asarray(allLabels)
ipyplot.plot_class_tabs(allPlots,
allLabels,
max_imgs_per_tab=15,
img_width=kwargs['width'])
except Exception:
return allPlots, allLabels
def display_compare_plots(func, kwargs):
"""
调用func绘图函数绘制多个模式, 多个预报时效的预报图形.
"""
# prepare function keyword
models = kwargs.pop('models')
kwargs['noshow'] = True
# get all model names
allModels = func(getModels=True)
# get all model plots
allPlots = []
allLabels = []
for model in allModels:
if models is not None:
if model not in models: continue
kwargs['model'] = model
plots, labels = func(**kwargs)
allPlots.append(plots)
allLabels.append(labels)
allPlots = [item for sublist in allPlots for item in sublist]
allLabels = [item for sublist in allLabels for item in sublist]
# display in jupyter
try:
# code run in jupyter notebook or not
get_ipython()
allPlots = np.asarray(allPlots, dtype=np.object)
allLabels = np.asarray(allLabels)
ipyplot.plot_class_tabs(allPlots,
allLabels,
max_imgs_per_tab=15,
img_width=kwargs['width'])
except Exception:
return allPlots, allLabels
def draw_multiple_plots(func, kwargs):
"""
调用func绘图函数绘制多个预报时效的图形, 并根据运行环境显示为
动画或列出每张图形.
"""
# generate forecast hours
fhours = check_frange(kwargs['frange'])
if fhours is None:
return None
# call funcation
plots = []
fhourLabels = []
for fhour in fhours:
kwargs['fhour'] = fhour
kwargs['frange'] = None
noshow = kwargs['noshow'] # force the plot function return tuple.
kwargs['noshow'] = True
plot, fhourLabel = func(**kwargs)
kwargs['noshow'] = noshow # retore the noshow keyword parameter.
if plot is None: continue
fhourLabels.append(fhourLabel)
plots.append(plot)
# don't show in the jupyter
if kwargs['noshow']:
return plots, fhourLabels
try:
# code run in jupyter notebook or not
get_ipython()
if kwargs['show'].upper() == 'ANIMATION':
# generate the gif figure
# bf = io.BytesIO()
# return Image(bf.getvalue())
f, outfile = tempfile.mkstemp(".gif")
os.close(f)
plots[0].save(outfile,
format='GIF',
append_images=plots[1:],
save_all=True,
duration=400,
loop=0)
image = Image(outfile)
os.unlink(outfile)
return image
else:
# show all plot figure
plots = np.asarray(plots, dtype=np.object)
fhourLabels = np.asarray(fhourLabels)
if kwargs['show'].upper() == 'TAB':
ipyplot.plot_class_tabs(plots,
fhourLabels,
max_imgs_per_tab=3,
img_width=plots[0].width)
else:
ipyplot.plot_images(plots,
fhourLabels,
img_width=kwargs['width'])
except Exception:
return plots, fhourLabels