-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathxr_interp_spare_to_grid.py
307 lines (242 loc) · 12.3 KB
/
xr_interp_spare_to_grid.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
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
"""
"""
from scipy.interpolate import SmoothSphereBivariateSpline,LSQSphereBivariateSpline
# from maihr_core.maihr_model import get_inputdata_maihr
from scipy import signal
import time
# import fatiando
import numpy as np
import xarray as xr
import matplotlib.pyplot as plt
from joblib import Parallel, delayed
from toolbox.utils import tqdm_joblib
from tqdm import tqdm
import pandas as pd
from scipy import interpolate
from scipy.interpolate import griddata
# from fatiando import gridder
from toolbox.geo import Inverse_weighted_interpolation
def xr_SmoothSphereBivariateSpline_sparse_to_grid(ds,
lat_grid=None,
lon_grid=None,
res=0.25, # grau
s=None,
lat_90=False,
lon_180=False,
s_factor=1,
n_process=20,
virtual_mean_border=False,
smooth_bivariate=False):
ds = ds.copy()
ds = ds.dropna('time',how='all')
assert 'latitude' in list(ds.coords), 'must have latitude as coords'
assert 'longitude' in list(ds.coords), 'must have latitude as coords'
assert 'time' in list(ds.coords), 'must have time as dim'
assert 'points' in list(ds.coords), 'must have points as dim'
# U = U.transpose('latitude','longitude')
###############
# Project to scipy function domain
###############
if lat_90:
ds = ds.assign_coords(latitude=('points', (ds.latitude.values + 90) * np.pi / 180))
else:
print('latitude are between 0 and 180 degree')
if lon_180:
ds = ds.assign_coords(longitude=('points', (ds.longitude.values + 180) * np.pi / 180))
else:
print('longitude are between 0 and 360 degree')
##############################
# New grid position
#############################
if lat_grid is None:
latmin_rad = ds.latitude.min()
latmax_rad = ds.latitude.max()
lonmin_rad = ds.longitude.min()
lonmax_rad = ds.longitude.max()
res_radian = res*np.pi/180 # radian
step_lat_rad = (latmax_rad - latmin_rad) / res_radian
step_lon_rad = (lonmax_rad - lonmin_rad) / res_radian
lat_grid = np.linspace(latmin_rad, latmax_rad, step_lat_rad)
lon_grid = np.linspace(lonmin_rad, lonmax_rad, step_lon_rad)
positions_x, positions_y = np.meshgrid(lon_grid,lat_grid)
##############
# Virtual mean border to ensure extrapolation stability
##############
if virtual_mean_border:
nb_points_border = 3
nb_points_along_border = 3
lon_right = np.linspace(lonmax_rad, lonmax_rad+(nb_points_border*res_radian), nb_points_border)
lon_left = np.linspace(lonmin_rad-(nb_points_border*res_radian), lonmin_rad, nb_points_border)
lat_top = np.linspace(latmax_rad, latmax_rad+(nb_points_border*res_radian), nb_points_border)
lat_bottom = np.linspace(latmin_rad-(nb_points_border*res_radian), latmin_rad, nb_points_border)
position_right_x, position_right_y = np.meshgrid(lon_right,np.linspace(latmin_rad, latmax_rad, nb_points_along_border))
position_left_x,position_left_y = np.meshgrid(lon_left,np.linspace(latmin_rad, latmax_rad, nb_points_along_border))
position_bottom_x, position_bottom_y = np.meshgrid(np.linspace(lonmin_rad, lonmax_rad, nb_points_along_border),lat_bottom)
position_top_x,position_top_y = np.meshgrid(np.linspace(lonmin_rad, lonmax_rad, nb_points_along_border),lat_top)
ds_position_right = xr.DataArray(dims=['points','time'],coords={'points':range(position_right_x.size),
'longitude':('points',position_right_x.ravel()),
'latitude':('points',position_right_y.ravel())
,'time':ds.time.values})
ds_position_left = xr.DataArray(dims=['points','time'],coords={'points':range(position_left_x.size),
'longitude':('points',position_left_x.ravel()),
'latitude':('points',position_left_y.ravel()),'time':ds.time.values})
ds_position_top = xr.DataArray(dims=['points','time'],coords={'points':range(position_top_x.size),
'longitude':('points',position_top_x.ravel()),
'latitude':('points',position_top_y.ravel()),'time':ds.time.values})
ds_position_bottom = xr.DataArray(dims=['points','time'],coords={'points':range(position_bottom_x.size),
'longitude':('points',position_bottom_x.ravel()),
'latitude':('points',position_bottom_y.ravel()),'time':ds.time.values})
# ds_position_right = ds_position_right.expand_dims('time')
ds_position_right = ds_position_right.fillna(ds.mean('points'))
ds_position_left = ds_position_left.fillna(ds.mean('points'))
ds_position_top = ds_position_top.fillna(ds.mean('points'))
ds_position_bottom = ds_position_bottom.fillna(ds.mean('points'))
ds = xr.concat([ds,ds_position_left,ds_position_right,ds_position_top,ds_position_bottom],dim='points')
##############
# Apply scipy function
###############
def apply_spline_func(name, element,s,s_factor):
d = element.dropna('points', how='all')
nb_test = 30
delta_increase= 0.15
s_factores = np.linspace(s_factor,s_factor+(nb_test*(s_factor*delta_increase)),nb_test)
i=0
j=0
while i<1:
try:
if s is None:
ss = int(d.std() * len(d.points)) * s_factor
else:
ss = int(s) * s_factor
interpolator_y = SmoothSphereBivariateSpline(d.latitude.values,
d.longitude.values,
d.values,
s=ss)
i=1
print(str(ss))
except:
j = j + 1
s_factor =s_factores[j]
# print(f'could not compute')
va = interpolator_y.ev(positions_y.ravel(), positions_x.ravel()).reshape(positions_y.shape)
ds_interp_rad = xr.DataArray(va,dims=['latitude','longitude'], coords={'latitude':LAT,'longitude':LON})
ds_interp_rad = ds_interp_rad.assign_coords({'time':name})
ds_interp_rad = ds_interp_rad.expand_dims('time')
ds_interp_rad = ds_interp_rad.assign_coords({'s': s})
ds_interp_rad = ds_interp_rad.expand_dims('s')
return ds_interp_rad
print('AppLy spline')
list_ds_interp_rad = Parallel(n_jobs=n_process, prefer='threads')(delayed(apply_spline_func)(X_name, X_element,s,s_factor) for X_name, X_element in list(ds.groupby('time')))
print('concat time spline')
ds_interp_grid = xr.concat(list_ds_interp_rad, dim='time')
print(ds_interp_grid)
print('done')
###############
# Reconstruct original domain
###############
if lat_90:
ds_interp_grid = ds_interp_grid.assign_coords(latitude=(ds_interp_grid.latitude.values * 180 / np.pi) - 90)
if lon_180:
ds_interp_grid = ds_interp_grid.assign_coords(longitude=(ds_interp_grid.longitude.values * 180 / np.pi) -180)
print('done interp grid')
return ds_interp_grid
def xr_fatiando_interp(ds, lat_grid=None, lon_grid=None, res=None):
if lat_grid is None:
latmin_rad = ds.lat.min()
latmax_rad = ds.lat.max()
lonmin_rad = ds.lon.min()
lonmax_rad = ds.lon.max()
step_lat_rad = (latmax_rad - latmin_rad) / res
step_lon_rad = (lonmax_rad - lonmin_rad) / res
lat_grid = np.linspace(latmin_rad, latmax_rad, step_lat_rad)
lon_grid = np.linspace(lonmin_rad, lonmax_rad, step_lon_rad)
positions_x, positions_y = np.meshgrid(lon_grid,lat_grid)
shape = positions_y.shape
ds = ds.dropna('time',how='all')
ds_group = ds.groupby('time')
def interp(d):
try:
d = d[1]
time = d.time
d = d.dropna('sta_id',how='any')
xp, yp, cubic = gridder.interp(d.lat.values,
d.lon.values,
d.values, shape,
area = (positions_y[0,0],positions_y[-1,0],positions_x[0,0],positions_x[0,-1]),
algorithm='cubic',
extrapolate=True)
xp = xp.reshape(shape)
yp = yp.reshape(shape)
cubic = cubic.reshape(shape)
ds_interp = xr.DataArray(cubic,dims=['lat','lon'],coords={'lat':xp[:,0],'lon':yp[0,:]})
ds_interp = ds_interp.assign_coords({'time':time})
ds_interp = ds_interp.expand_dims('time')
except:
ds_interp = None
return ds_interp
list_ds_interp_rad = Parallel(n_jobs=40, prefer='threads')(delayed(interp)(d) for d in tqdm(ds_group))
list_ds_interp_rad = [ d for d in list_ds_interp_rad if d is not None]
print('concat interp')
ds_interp_grid = xr.concat(list_ds_interp_rad, dim='time')
print('done interp')
return ds_interp_grid
def xr_idx_interp(ds_y_for_spline, lon, lat,p=2):
"""
https://rafatieppo.github.io/post/2018_07_27_idw2pyr/
:param ds_y_for_spline:
:param lon:
:param lat:
:return:
"""
ds_y_for_spline = ds_y_for_spline.load()
time= ds_y_for_spline.time
print('interp idx')
Lon, Lat = np.meshgrid(lon, lat)
ds_group = ds_y_for_spline.groupby('time')
out_shape = Lon.shape
Lon = Lon.ravel()
Lat = Lat.ravel()
def interp_idw(dd,Lon,Lat, out_shape,lat,lon,p):
# try:
if 'points' in dd[1].dims:
dd = dd[1].dropna('point', how='all')
if 'sta_id' in dd[1].dims:
dd = dd[1].dropna('sta_id', how='all')
ds_interp = Inverse_weighted_interpolation(dd.lon.values,
dd.lat.values,
dd.values,
Lon,
Lat,p)
ds_interp = xr.DataArray(np.array(ds_interp).reshape(out_shape), dims=['lat', 'lon'],
coords={'lon':lon,
'lat':lat})
# except:
# ds_interp=None
return ds_interp
list_interped = Parallel(n_jobs=40, prefer='processes')(delayed(interp_idw)(dd,Lon,Lat, out_shape,lat,lon,p) for dd in tqdm(ds_group))
list_interped = [d for d in list_interped if d is not None]
print('concat interp')
ds_interp_grid = xr.concat(list_interped, dim='time')
ds_interp_grid = ds_interp_grid.assign_coords(time=time)
print('done interp')
return ds_interp_grid
def xr_smooth(ds, kernel_size,n_jobs=40):
kernel = np.ones((kernel_size, kernel_size))
kernel= kernel/(kernel_size*kernel_size)
ds_group = ds.groupby('time')
def smooth(d,kernel):
d = d[1]
time = d.time
grad = signal.convolve2d(d, kernel, boundary='symm', mode='same')
ds_interp = xr.DataArray(grad, dims=['lat', 'lon'], coords={'lat': d.lat.values, 'lon':d.lon.values})
ds_interp = ds_interp.assign_coords({'time': time})
ds_interp = ds_interp.expand_dims('time')
return ds_interp
list_ds_interp_rad = Parallel(n_jobs=n_jobs, prefer='processes')(delayed(smooth)(d,kernel) for d in tqdm(ds_group))
list_ds_interp_rad = [ d for d in list_ds_interp_rad if d is not None]
print('concat interp')
ds_interp_grid = xr.concat(list_ds_interp_rad, dim='time')
print('done interp')
return ds_interp_grid
if __name__ == '__main__':
pass