-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathconsensus_set.py
215 lines (199 loc) · 9.2 KB
/
consensus_set.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
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
import numpy as np
import numpy.ma as ma
import datetime as dt
import pytz
from typing import List
import netCDF4
import tools
from vad import VADSet
from Lidar_functions import consensus_avg
class ConsensusSet(VADSet):
@staticmethod
def create_time_ranges(day: dt.date, size: dt.timedelta
) -> List[dt.datetime]:
""" Create a list of datetimes every given interval for this day """
# currently if the interval size is not divisible into 24 hours, the
# last time range will extend into the following day
start = dt.datetime(day.year, day.month, day.day, tzinfo=pytz.UTC)
end = start + dt.timedelta(days=1)
ranges = []
while (start < end):
ranges.append(start)
start = start + size
return ranges
@staticmethod
def median_from_consensus_idxs(vals: ma.array, idxs: List) -> float:
"""
Calculate the median of an array from the subset of elements at the
same indices as were used for consensus averaging for w.
"""
# avoid "mean of an empty slice" numpy runtime warning
if not idxs:
return np.nan
return ma.median(vals[idxs])
def __init__(self, alt: float, lat: float, lon: float, height: np.array,
time: list, u: ma.array, v: ma.array, w: ma.array,
n_u: ma.array, n_v: ma.array, n_w: ma.array,
residual: ma.array, correlation: ma.array, mean_cnr: ma.array,
wspd: ma.array, wdir: ma.array, window: float,
span: dt.timedelta):
# separating out calculating from vadset into a classmethod in case we
# want another classmethod to read from file in the future
self.alt = alt
self.lat = lat
self.lon = lon
self.height = height
self.stime = time
# generate etime from stime and span, no need to supply it
self.etime = [s + span for s in self.stime]
self.u = u
self.v = v
self.w = w
self.n_u = n_u # number of indices averaged in consensus
self.n_v = n_v
self.n_w = n_w
self.residual = residual
self.correlation = correlation
self.mean_cnr = mean_cnr
self.speed = wspd
self.wdir = wdir
self.window = window # size of consensus buckets
self.span = span # timespan to average over
@classmethod
def from_VADSet(cls, vs: VADSet, window: float, span: dt.timedelta):
# create half-hour time bins
ranges = ConsensusSet.create_time_ranges(vs.stime[0].date(), span)
u_mean = ma.zeros((len(ranges), len(vs.height)))
v_mean = ma.zeros((len(ranges), len(vs.height)))
w_mean = ma.zeros((len(ranges), len(vs.height)))
n_u = ma.zeros((len(ranges), len(vs.height)))
n_v = ma.zeros((len(ranges), len(vs.height)))
n_w = ma.zeros((len(ranges), len(vs.height)))
residual = ma.zeros((len(ranges), len(vs.height)))
correlation = ma.zeros((len(ranges), len(vs.height)))
mean_cnr = ma.zeros((len(ranges), len(vs.height)))
for idx, r in enumerate(ranges):
start = r
end = start + span
bin_idxs = [i for i in range(len(vs.stime)) if vs.stime[i] >= start
and vs.stime[i] < end]
# no vads in this time window
if len(bin_idxs) == 0:
u_mean[idx, :] = np.nan
v_mean[idx, :] = np.nan
w_mean[idx, :] = np.nan
residual[idx, :] = np.nan
correlation[idx, :] = np.nan
mean_cnr[idx, :] = np.nan
n_u[idx, :] = np.nan
n_v[idx, :] = np.nan
n_w[idx, :] = np.nan
continue
u_bin = ma.array([vs.u[i] for i in bin_idxs])
v_bin = ma.array([vs.v[i] for i in bin_idxs])
w_bin = ma.array([vs.w[i] for i in bin_idxs])
res_bin = ma.array([vs.residual[i] for i in bin_idxs])
cor_bin = ma.array([vs.correlation[i] for i in bin_idxs])
mc_bin = ma.array([vs.mean_cnr[i] for i in bin_idxs])
for hgt in range(len(vs.height)):
avg, idxs = consensus_avg(u_bin[:, hgt], window)
u_mean[idx, hgt] = avg
n_u[idx, hgt] = len(idxs)
avg, idxs = consensus_avg(v_bin[:, hgt], window)
v_mean[idx, hgt] = avg
n_v[idx, hgt] = len(idxs)
avg, idxs = consensus_avg(w_bin[:, hgt], window)
w_mean[idx, hgt] = avg
n_w[idx, hgt] = len(idxs)
# take median of idxs used for w to avg res, corr, mean_cnr
residual[idx, hgt] = ConsensusSet.\
median_from_consensus_idxs(res_bin[:, hgt], idxs)
correlation[idx, hgt] = ConsensusSet.\
median_from_consensus_idxs(cor_bin[:, hgt], idxs)
mean_cnr[idx, hgt] = ConsensusSet.\
median_from_consensus_idxs(mc_bin[:, hgt], idxs)
# calculate wspd/wdir from averaged u and v
wspd, wdir = tools.wspd_wdir_from_uv(u_mean, v_mean)
return cls(vs.alt,
vs.lat,
vs.lon,
vs.height,
ranges,
ma.masked_invalid(u_mean),
ma.masked_invalid(v_mean),
ma.masked_invalid(w_mean),
ma.masked_invalid(n_u),
ma.masked_invalid(n_v),
ma.masked_invalid(n_w),
ma.masked_invalid(residual),
ma.masked_invalid(correlation),
ma.masked_invalid(mean_cnr),
ma.masked_invalid(wspd),
ma.masked_invalid(wdir),
window,
span)
@classmethod
def from_file(cls, filename: str):
""" Create a ConsensusSet object from a daily consensus netcdf file """
f = netCDF4.Dataset(filename, 'r')
stime = list(netCDF4.num2date(f.variables['time'][:],
f.variables['time'].units,
only_use_python_datetimes=True,
only_use_cftime_datetimes=False))
# reconstitute endtime from scan duration
# make dates tz aware
stime = [s.replace(tzinfo=pytz.utc) for s in stime]
# scalar values with a missing_value get read in as np.ma.MaskedArray
# if the single value is not masked, but as np.ma.core.MaskedConstant
# if the single value is masked. Instances of MaskedConstant don't
# compare well to each other or to other data types, so if the scalar
# lat/lon/alt is masked, replace it with a null-dimensioned masked
# array with a value of nan.
alt = (ma.array(np.nan) if f.variables['alt'][:] is np.ma.masked
else f.variables['alt'][:])
lat = (ma.array(np.nan) if f.variables['lat'][:] is np.ma.masked
else f.variables['lat'][:])
lon = (ma.array(np.nan) if f.variables['lon'][:] is np.ma.masked
else f.variables['lon'][:])
# reconstitute span by finding the difference between subsequent times
span = stime[1] - stime[0]
return cls(alt,
lat,
lon,
f.variables['height'][:],
stime,
ma.array(f.variables['u'][:]),
ma.array(f.variables['v'][:]),
ma.array(f.variables['w'][:]),
ma.array(f.variables['u_npoints'][:]),
ma.array(f.variables['v_npoints'][:]),
ma.array(f.variables['w_npoints'][:]),
ma.array(f.variables['residual'][:]),
ma.array(f.variables['correlation'][:]),
ma.array(f.variables['mean_snr'][:]),
ma.array(f.variables['wind_speed'][:]),
ma.array(f.variables['wind_direction'][:]),
int(f.consensus_avg_window),
span)
def add_aux_variables(self, nc_file: netCDF4.Dataset):
# number of points used in consensus for uvw
n_u = nc_file.createVariable('u_npoints', 'f', ('time', 'height'))
n_u.missing_value = -9999.0
n_u[:, :] = self.n_u
n_u.long_name = ('Number of points used in consensus averaging window '
'for eastward component of winds')
n_u.units = 'unitless'
n_v = nc_file.createVariable('v_npoints', 'f', ('time', 'height'))
n_v.missing_value = -9999.0
n_v[:, :] = self.n_v
n_v.long_name = ('Number of points used in consensus averaging window '
'for northward component of winds')
n_v.units = 'unitless'
n_w = nc_file.createVariable('w_npoints', 'f', ('time', 'height'))
n_w.missing_value = -9999.0
n_w[:, :] = self.n_w
n_w.long_name = ('Number of points used in consensus averaging window '
'for vertical component of winds')
n_w.units = 'unitless'
def add_attributes(self, nc_file: netCDF4.Dataset):
nc_file.consensus_avg_window = self.window