__author__ = "Markus Pichler"
__credits__ = ["Markus Pichler"]
__maintainer__ = "Markus Pichler"
__email__ = "markus.pichler@tugraz.at"
__version__ = "0.1"
__license__ = "MIT"
import pandas as pd
from .definitions import COL
from .little_helpers import duration_steps_readable
from .sww_utils import guess_freq, rain_events, event_duration
RETURN_PERIOD_COLORS = {
# 0.5: '#e0ffff', # 'lightcyan',
1: '#00ffff', # 'cyan',
2: '#add8e6', # 'lightblue',
5: '#0000ff', # 'blue',
10: '#ffff00', # 'yellow',
20: '#ffa500', # 'orange',
50: '#ff0000', # 'red',
100: '#ff00ff', # 'magenta',
}
[docs]
def idf_bar_axes(ax, idf_table, return_period_colors=RETURN_PERIOD_COLORS):
"""
create
Args:
ax (matplotlib.pyplot.Axes):
idf_table (pandas.DataFrame):
return_period_colors (dict): color of each return period {return period: color}
Returns:
matplotlib.pyplot.Axes:
"""
return_periods = list(return_period_colors.keys())
color_return_period = list(return_period_colors.values())
# legend
from matplotlib.lines import Line2D
custom_lines = [Line2D([0], [0], color=c, lw=4) for c in color_return_period]
names = ['{}a'.format(t) for t in return_periods]
ax.legend(custom_lines, names, bbox_to_anchor=(0., 1.02, 1., .102), loc=3, ncol=len(color_return_period),
mode="expand", borderaxespad=0., title='return periods')
duration_size = len(idf_table.columns)
# labels for the y axis
durations_index = range(duration_size)
dh = 1
ax.set_yticks([i + dh/2 for i in durations_index], minor=True)
ax.set_yticks(list(durations_index), minor=False)
ax.set_yticklabels(duration_steps_readable(idf_table.columns), minor=True)
ax.set_yticklabels([''] * duration_size, minor=False)
ax.set_ylabel('duration of the design rainfall')
# for the relative start time
freq = guess_freq(idf_table.index)
start_period = idf_table.index[0].to_period(freq).ordinal
# idf_table.index = idf_table.index - idf_table.index[0]
min_duration = pd.Timedelta(minutes=1)
for hi, d in enumerate(idf_table.columns):
tn = idf_table[d]
for t in return_periods:
c = return_period_colors[t]
# not really a rain event, but the results are the same
# tab2 = rain_events(tn, ignore_rain_below=t, min_gap=pd.Timedelta(minutes=d))
tab = rain_events(tn, ignore_rain_below=t, min_gap=freq)
# if tab.size != tab2.size:
# print()
if tab.empty:
continue
if 1:
durations = ((event_duration(tab) + freq) / min_duration).tolist()
rel_starts = ((tab[COL.START] - idf_table.index[0]) / min_duration + start_period).tolist()
bar_x = list(zip(rel_starts, durations))
else:
tab[COL.DUR] = event_duration(tab) / min_duration
bar_x = [(r[COL.START] / min_duration + start_period, r[COL.DUR]) for _, r in tab.iterrows()]
ax.broken_barh(bar_x, (hi, dh), facecolors=c)
ax.tick_params(axis='y', which='minor', length=0)
ax.grid(axis='y', which='major')
ax.set_ylim(0, duration_size)
ax.set_xticklabels([])
from matplotlib.ticker import NullFormatter
ax.xaxis.set_major_formatter(NullFormatter())
# ---
duration_steps = idf_table.columns.values
duration_steps_middle_to_long = duration_steps[duration_steps > 2*60]
if duration_steps_middle_to_long.size:
# (k)urzzeitige Summationen, d. h. der Dauerstufen von 5 Minuten bis 2 Stunden
ax.axhline(duration_steps.tolist().index(duration_steps_middle_to_long[0]), color='black')
duration_steps_long = duration_steps_middle_to_long[duration_steps_middle_to_long > 3*60*24]
if duration_steps_long.size:
# (m)ittelfristige Summationen, d. h. der Dauerstufen von 3 Stunden bis 3 Tagen.
ax.axhline(duration_steps.tolist().index(duration_steps_long[0]), color='black')
return ax