-
Notifications
You must be signed in to change notification settings - Fork 0
/
shared.py
196 lines (175 loc) · 5.56 KB
/
shared.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
import datetime
import os.path
from typing import Tuple, List, Optional
import pickle
import numpy as np
import netCDF4
import matplotlib.dates
def read_0d_observations(path: str) -> np.ndarray:
obs = []
for l in open(os.path.join(path)):
if not l.startswith("#"):
dt, value, sd = l.rstrip().rsplit(maxsplit=2)
mu = float(value)
sigma = float(sd)
p25 = mu - 0.67448 * sigma
p75 = mu + 0.67448 * sigma
obs.append(
[datetime.datetime.strptime(dt, "%Y-%m-%d %H:%M:%S"), mu, p25, p75]
)
return np.array(obs).reshape(-1, 4)
def read_result(
path: str, name: str
) -> Tuple[List[datetime.datetime], np.ndarray, np.ndarray, str, str]:
with netCDF4.Dataset(path) as nc:
nctime = nc["time"]
time = netCDF4.num2date(
nctime,
nctime.units,
only_use_cftime_datetimes=False,
only_use_python_datetimes=True,
)
z = -nc["z"][:, :, 0, 0]
ncvar = nc[name]
values = ncvar[..., 0, 0]
return time, z, values, ncvar.long_name, ncvar.units
def read_ensemble_result(
path: str, name: str, N: int
) -> Tuple[List[datetime.datetime], np.ndarray, np.ndarray, str, str]:
pathname, pathext = os.path.splitext(path)
results = []
for i in range(1, N + 1):
time, z, values, long_name, units = read_result(
f"{pathname}_{i:04}{pathext}", name
)
results.append(values)
return time, z, np.array(results), long_name, units
def plot_0d_timeseries(ax, time, values, obs, label: str = "model", extra_series=[]):
low = obs[:, 1] - obs[:, 2]
high = obs[:, 3] - obs[:, 1]
ax.errorbar(
obs[:, 0],
obs[:, 1],
yerr=[low, high],
ecolor="k",
elinewidth=1.0,
fmt=".k",
alpha=0.4,
zorder=-1,
label="observations",
)
icolor = 0
for extra_label, extra_values in extra_series:
ax.plot_date(time, extra_values, "-", color=f"C{icolor}", label=extra_label)
icolor += 1
(series,) = ax.plot_date(time, values, "-", color=f"C{icolor}", label=label)
ax.set_xlim(time[0], time[-1])
ax.xaxis.set_major_formatter(
matplotlib.dates.ConciseDateFormatter(ax.xaxis.get_major_locator())
)
ax.grid()
ax.legend()
return series
def plot_1d_timeseries(ax, time, z, values, *args, cax=None, **kwargs):
fig = ax.figure
time_2d = np.broadcast_to(time[:, np.newaxis], z.shape)
pc = ax.contourf(time_2d, z, values, *args, **kwargs)
cb = fig.colorbar(pc, cax=cax)
ax.set_ylabel("depth (m)")
ax.xaxis.axis_date()
ax.set_xlim(time[0], time[-1])
ax.xaxis.set_major_formatter(
matplotlib.dates.ConciseDateFormatter(ax.xaxis.get_major_locator())
)
ax.grid()
ax.set_ylim(z.max(), z.min())
return pc, cb
def plot_0d_ensemble_timeseries(
ax,
time,
ens,
ref={},
obs=None,
filter_period: int = 1,
plot_spread: bool = True,
label: Optional[str] = None,
):
if filter_period != 1:
import scipy.signal
ens = scipy.signal.medfilt(ens, (1, filter_period))
if obs is not None:
low = obs[:, 1] - obs[:, 2]
high = obs[:, 3] - obs[:, 1]
# ax.plot_date(obs[:,0], obs[:,1], '.k', alpha=0.4, label='observations')
ax.errorbar(
obs[:, 0],
obs[:, 1],
yerr=[low, high],
ecolor="k",
elinewidth=1.0,
fmt=".k",
alpha=0.4,
zorder=-1,
label="observations",
)
label = label or "model"
icolor = 0
for reflabel, refvalues in ref:
ax.plot_date(time, refvalues, "-", color=f"C{icolor}", label=reflabel)
icolor += 1
if plot_spread:
p25 = np.percentile(ens, 25.0, axis=0)
p75 = np.percentile(ens, 75.0, axis=0)
pmin = np.min(ens, axis=0)
pmax = np.max(ens, axis=0)
ax.fill_between(
time,
pmin,
pmax,
alpha=0.2,
label=f"{label}, ensemble min to max",
fc=f"C{icolor}",
)
ax.fill_between(time, p25, p75, fc="w")
ax.fill_between(
time,
p25,
p75,
alpha=0.5,
label=f"{label}, 1st to 3rd ensemble quartile",
fc=f"C{icolor}",
)
for p in (pmin, pmax, p25, p75):
ax.plot_date(time, p, "-k", lw=0.2)
label = f"{label}, ensemble median"
median = np.median(ens, axis=0)
ax.plot_date(time, median, "-", color=f"C{icolor}", label=label)
ax.grid()
ax.legend()
ax.set_xlim(time[0], time[-1])
ax.xaxis.set_major_formatter(
matplotlib.dates.ConciseDateFormatter(ax.xaxis.get_major_locator())
)
def plot_1d_ensemble_timeseries(ax, time, z, ens, *args, cax=None, **kwargs):
fig = ax.figure
time_2d = np.broadcast_to(time[:, np.newaxis], z.shape)
pc = ax.contourf(time_2d, z, np.median(ens, axis=0), *args, **kwargs)
cb = fig.colorbar(pc, cax=cax)
ax.set_ylabel("depth (m)")
ax.grid()
ax.xaxis.axis_date()
ax.set_xlim(time[0], time[-1])
ax.xaxis.set_major_formatter(
matplotlib.dates.ConciseDateFormatter(ax.xaxis.get_major_locator())
)
ax.set_ylim(z.max(), z.min())
return pc, cb
def seed(path: str):
if os.path.isfile(path):
with open(path, "rb") as f:
seed = pickle.load(f)
else:
seed = np.random.SeedSequence()
with open(path, "wb") as f:
pickle.dump(seed.entropy, f)
return seed