-
Notifications
You must be signed in to change notification settings - Fork 6
/
Copy pathplot.py
1309 lines (1131 loc) · 51.7 KB
/
plot.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
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
# -*- coding: utf8 -*-
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <https://www.gnu.org/licenses/>.
# Headless plotting
import matplotlib
matplotlib.use("Agg")
import sys
import os
import math
import subprocess
import numpy as np
import xarray as xr
import pandas as pd
import geopandas as gpd
import shapely.geometry
import seaborn as sns
import toml
import cartopy as ctpy
import scipy.stats
from cartopy.mpl.gridliner import LONGITUDE_FORMATTER, LATITUDE_FORMATTER
import matplotlib.pyplot as plt
from matplotlib.ticker import FuncFormatter, ScalarFormatter
import config
import ev_quantiles
import helper
import postprocessing
# Increase DPI of figures
matplotlib.rcParams['savefig.dpi'] = 200
plot_dir = config.plot['dir']
midas_stations = config.plot['station_map']
basemap = config.plot['base_map']
EXTRACT = dict(latitude=slice(45, -45),
longitude=slice(0, 180))
# Coordinates of study sites
# STUDY_SITES = {'Kampala': (0.317, 32.616),
# 'Kisumu': (0.1, 34.75)}
STUDY_SITES = {
'Jakarta': (-6.2, 106.816), 'Sydney': (-33.865, 151.209),
'Beijing': (39.92, 116.38), 'New Delhi': (28.614, 77.21),
'Niamey': (13.512, 2.125), #'Jeddah': (21.54, 39.173), 'Cape Town': (-33.925278, 18.4238),
'Nairobi': (-1.28, 36.82), #'Brussels': (50.85, 4.35),
'Santiago': (-33.45, -70.67), #'New York City': (40.72, -74.0),
'Mexico City': (19.43, -99.13), # 'Vancouver': (49.25, -123.1),
# 'Vienna': (48.2, 16.367),
'Natal': (-5.78, -35.2),
# 'Surakarta': (-7.57, 110.817), 'Semarang': (-6.97, 110.417)
}
XTICKS = [1, 3, 6, 12, 24, 48, 120, 360]
# color-blind safe qualitative palette from colorbrewer2 (paired)
C_PRIMARY_1 = '#1f78b4'
C_SECONDARY_1 = '#6ba9ca'
C_PRIMARY_2 = '#33a02c'
C_SECONDARY_2 = '#89c653'
METADATA = 'metadata.toml'
with open(METADATA, 'r') as toml_file:
metadata = toml.loads(toml_file.read())
# output
PXR2 = f"pxr2-{metadata['common']['version']}.nc"
PXR4 = f"pxr4-{metadata['common']['version']}.nc"
def set_logd_xticks(ax, xmin=min(XTICKS), xmax=max(XTICKS)):
xticks = [i for i in XTICKS if i <= xmax and i >= xmin]
ax.set_xlim(xmin, xmax)
ax.set_xticks(xticks)
ax.get_xaxis().set_major_formatter(matplotlib.ticker.ScalarFormatter())
def single_map(da, cbar_label, title, fig_name, center=False, reverse=False):
"""https://cbrownley.wordpress.com/2018/05/15/visualizing-global-land-temperatures-in-python-with-scrapy-xarray-and-cartopy/
"""
plt.figure(figsize=(8, 5))
ax_p = plt.gca(projection=ctpy.crs.EqualEarth(), aspect='auto')
if center is True:
if reverse:
cmap = 'RdBu_r'
else:
cmap = 'RdBu'
else:
cmap='viridis'
da.plot.imshow(ax=ax_p, transform=ctpy.crs.PlateCarree(),
robust=True, cmap=cmap, center=center,
cbar_kwargs=dict(orientation='horizontal', label=cbar_label))
ax_p.coastlines(linewidth=.5, color='black')
plt.title(title)
plt.savefig(os.path.join(plot_dir, fig_name))
plt.close()
def multi_maps(da_list, disp_names, value_name, fig_name,
sqr=False, center=False, reverse=False, robust=True,
onebar=True, bar_labels=None):
crs = ctpy.crs.EqualEarth()
# reshape variables as dimension
da_list2 = []
for da in da_list:
if sqr == True:
da_sel = np.square(da).expand_dims('param')
else:
da_sel = da.expand_dims('param')
da_sel.coords['param'] = [da.name]
da_list2.append(da_sel)
da_all = xr.concat(da_list2, 'param')
da_all.attrs['long_name'] = value_name # Color bar title
# Actual plot
# print(da_all['longitude'])
aspect = len(da_all['longitude']) / len(da_all['latitude'])
if center is True:
if reverse:
cmap = 'RdBu_r'
else:
cmap = 'RdBu'
else:
cmap = 'viridis'
if onebar: # One shared colorbar
fg = da_all.plot(col='param', col_wrap=1,
transform=ctpy.crs.PlateCarree(), aspect=aspect,
cmap=cmap, center=center,
robust=robust, extend='both',
subplot_kws=dict(projection=crs)
)
for ax, disp_name in zip(fg.axes.flat, disp_names):
ax.coastlines(linewidth=.5, color='black')
ax.set_title(disp_name)
else:
fig, axes = plt.subplots(len(da_list2), 1, sharey=False, sharex=False,
subplot_kw=dict(projection=crs),
)
for da, ax, disp_name, bar_label in zip(da_list2, axes, disp_names, bar_labels):
da = da.isel(param=0, drop=True)
# print(da)
da.plot.imshow(ax=ax, transform=ctpy.crs.PlateCarree(),
robust=True, cmap=cmap, center=center,
cbar_kwargs=dict(orientation='vertical', label=bar_label))
ax.coastlines(linewidth=.5, color='black')
ax.set_title(disp_name)
plt.tight_layout()
plt.savefig(os.path.join(plot_dir, fig_name))
plt.close()
def get_site_list(ds, name_col, use_only=None):
"""return a dict {name: (lat, lon)}
"""
# Set station coordinates to station name
ds.coords['station'] = ds[name_col]
if not use_only:
use_only = ds[name_col].values
sites = {}
for name, lat, lon in zip(ds[name_col].values,
ds['latitude'].values,
ds['longitude'].values):
if name in use_only:
sites[name] = (lat, postprocessing.convert_lon(lon))
return sites
def plot_point_map(ds, ax):
ds_sel = ds.sel(duration=ds.duration.values[0],
source=ds.source.values[0],)
df = ds_sel.to_dataframe()
df['geometry'] = [shapely.geometry.Point(lon, lat)
for lon, lat in zip(df['longitude'],
df['latitude'])]
gdf = gpd.GeoDataFrame(df, geometry='geometry')
ax.set_global()
ax.coastlines(linewidth=.3, color='0.6', zorder=0)
gdf.plot(ax=ax, markersize=25, transform=ctpy.crs.PlateCarree(), zorder=20)
ax.set_title('Sample sites location')
def plot_scaling_per_site(ds, fig_name):
"""
"""
linesyles = {
# 'MIDAS_location': dict(linestyle='None', linewidth=0, marker='v', markersize=3,
# color=C_PRIMARY_1, label='Location $\mu$ (MIDAS)'),
# 'MIDAS_location_lr': dict(linestyle='solid', linewidth=0.5, marker=None, markersize=0,
# color=C_PRIMARY_1, label='$a d^\\alpha$ (MIDAS)'),
# 'MIDAS_scale': dict(linestyle='None', linewidth=0, marker='v', markersize=3,
# color=C_PRIMARY_2, label='Scale $\sigma$ (MIDAS)'),
# 'MIDAS_scale_lr': dict(linestyle='solid', linewidth=0.5, marker=None, markersize=0,
# color=C_PRIMARY_2, label='$\sigma_D (d/D)^\\beta$ (MIDAS)'),
'ERA5_location': dict(linestyle='None', linewidth=0, marker='o', markersize=2,
color=C_PRIMARY_1, label='Location $\mu$', zorder=20),
'ERA5_scale': dict(linestyle='None', linewidth=0, marker='o', markersize=2,
color=C_PRIMARY_2, label='Scale $\sigma$', zorder=20),
'ERA5_location_lr': dict(linestyle='solid', linewidth=1., marker=None, markersize=0,
color=C_PRIMARY_1, label='$ad^\\alpha$', zorder=10),
'ERA5_scale_lr': dict(linestyle='solid', linewidth=1., marker=None, markersize=0,
color=C_PRIMARY_2, label='$bd^\\beta$', zorder=10),
# 'ERA5_location_lr_daily': dict(linestyle='dashed', linewidth=1., marker=None, markersize=0,
# color=C_SECONDARY_1, label='$ad^\\alpha$ (daily)'),
# 'ERA5_scale_lr_daily': dict(linestyle='dashed', linewidth=1., marker=None, markersize=0,
# color=C_SECONDARY_2, label='$bd^\\beta$ (daily)'),
}
dict_df = postprocessing.ds_to_df(ds, 'station')
# sys.exit() ###
col_num = 2
row_num = math.ceil(len(dict_df) / col_num)
# fig_size = (9, 5) # landscape
fig_size = (6, 9) # portrait
fig = plt.figure(figsize=fig_size)
ax_num = 1
# Draw map
ax_map = fig.add_subplot(row_num, col_num, ax_num,
projection=ctpy.crs.EqualEarth(),
aspect='auto')
plot_point_map(ds, ax_map)
ax_num += 1
sites_ax_list = []
for site_ax_num, (site_name, df) in enumerate(dict_df.items()):
if site_ax_num == 0:
ax = fig.add_subplot(row_num, col_num, ax_num)
first_ax = ax
else:
ax = fig.add_subplot(row_num, col_num, ax_num,
sharex=first_ax, sharey=first_ax)
print(site_name)
# if site_name == 'Jakarta':
# print(df[['ERA5_location_ci_l', 'ERA5_location_ci_h', 'ERA5_location_lr_ci_l', 'ERA5_location_lr_ci_h']].head())
# plot
for col_prefix, styles in linesyles.items():
col_est = col_prefix + '_est'
col_ci_l = col_prefix + '_ci_l'
col_ci_h = col_prefix + '_ci_h'
# plot estimate
df[col_est].plot(ax=ax, label=styles['label'],
title=site_name, zorder=styles['zorder'],
loglog=True,
linestyle=styles['linestyle'], linewidth=styles['linewidth'],
markersize=styles['markersize'],
marker=styles['marker'], color=styles['color'])
# plot error
if col_prefix.endswith('_lr'):
ax.fill_between(df.index, df[col_ci_h], df[col_ci_l],
alpha=0.25, label=styles['label'] + ' 95% CI', linewidth=0,
color=styles['color'], zorder=0)
else:
df[col_ci_h].plot(ax=ax, label=styles['label'] + ' 95% CI',
title=site_name, loglog=True,
zorder=5, alpha=0.9, marker='_',
linestyle=styles['linestyle'], linewidth=styles['linewidth'],
markersize=styles['markersize'],
color=styles['color'])
df[col_ci_l].plot(ax=ax, label='',
title=site_name, loglog=True,
zorder=5, alpha=0.9, marker='_',
linestyle=styles['linestyle'], linewidth=styles['linewidth'],
markersize=styles['markersize'],
color=styles['color'])
lines, labels = ax.get_legend_handles_labels()
ax.set_xlabel('$d$ (hours)')
ax.set_ylabel('$\mu, \sigma$ (mm/h)')
# ax.set_xscale('log')
# ax.set_yscale('log')
sites_ax_list.append(ax)
ax_num += 1
# set ticks
for ax in sites_ax_list:
set_logd_xticks(ax)
ax.use_sticky_edges = False
ax.autoscale()
# lgd_ncol = math.ceil(len(labels) / 4) # talk (landscape)
lgd_ncol = math.ceil(len(labels) / 2) # article (portrait)
lgd = fig.legend(lines, labels, loc='lower center', ncol=lgd_ncol)
plt.tight_layout()
# plt.subplots_adjust(bottom=.2, wspace=None, hspace=None) # Talk (landscape)
plt.subplots_adjust(bottom=.12, wspace=None, hspace=None) # Article (portrait)
plt.savefig(os.path.join(plot_dir, fig_name))
plt.close()
def plot_hourly(ds, sites, fig_name):
ds_list = []
for site_name, site_coord in sites.items():
ds_list.append(ds.sel(latitude=site_coord[0],
longitude=site_coord[1],
method='nearest'))
drop_coords = ['latitude', 'longitude']
ds_sites = xr.concat(ds_list, dim='site').drop(drop_coords)
ds_sites.coords['site'] = list(sites.keys())
p = ds_sites['precipitation'].plot(col='site')
plt.savefig(os.path.join(plot_dir, fig_name))
plt.close()
def probability_plot(ds, sites, fig_name):
# print(ds)
DURATION = 6
keep_var = ['ecdf_gringorten', 'annual_max', 'cdf', 'ev_parameter']
# extract sites values from the dataset
dict_cdf = {}
ds_list = []
for site_name, site_coord in sites.items():
ds_sel = ds.sel(latitude=site_coord[0],
longitude=site_coord[1],
duration=DURATION,
method='nearest', drop=True)
ds_sel = ds_sel.expand_dims('site')
ds_sel.coords['site'] = [site_name]
ds_list.append(ds_sel[keep_var])
# Split CDF into variables
ds_sites = xr.concat(ds_list, dim='site')
ds_cdf = ds_sites['cdf'].to_dataset(dim='ev_fit')
ds_split = xr.merge([ds_sites.drop('cdf'), ds_cdf]).drop('ev_fit')
# All the probabilities into one dimension
da_prob = ds_split.drop('annual_max').to_array('cdf', name='prob')
ds_prob = xr.merge([da_prob, ds_split['annual_max']])
# print(ds_prob)
# compute quantiles for a range of years
ds_params = ds_sites['ev_parameter'].to_dataset(dim='ev_param')
# print(ds_params)
t = np.arange(2, 10000)
return_periods = xr.DataArray(t, dims=['T'], coords={'T': t})
# print(return_periods)
da_q = ev_fit.gev_quantile(return_periods,
ds_params['location'],
ds_params['scale'],
ds_params['shape']
).rename('intensity')
# print(da_q)
# to Pandas dataframe
df_cdf = (ds_prob
.to_dataframe()
.reset_index()
.drop('year', axis='columns')
)
# prob to return period
df_cdf['T'] = 1/(1-df_cdf['prob'])
df_cdf = df_cdf.sort_values('T')
# plot
colwrap = 3
row_num = math.ceil(len(sites) / colwrap)
fig, axes = plt.subplots(row_num, colwrap, sharey=False, sharex=True, figsize=(9, 7))
for ax, site_name in zip(axes.flat, sites.keys()):
df_site = df_cdf.loc[df_cdf['site'] == site_name]
# ECDF
ecdf = df_site.loc[df_cdf['cdf'].str.startswith('ecdf_')]
ecdf.plot(x='T', y='annual_max', logx=True, legend=False, ax=ax,
linestyle='', lw=0.1, marker='x', markeredgecolor='.1',
markersize=5, label='ecdf')
# CDFs
gumbel = da_q.sel(site=site_name, ev_fit='gumbel').to_dataframe().reset_index()
gumbel.plot(x='T', y='intensity', logx=True, legend=False, ax=ax,
c='.2', linewidth=1, linestyle='--', label='Gumbel')
gev = da_q.sel(site=site_name, ev_fit='gev').to_dataframe().reset_index()
gev.plot(x='T', y='intensity', logx=True, legend=False, ax=ax,
label='GEV')
frechet = da_q.sel(site=site_name, ev_fit='frechet').to_dataframe().reset_index()
frechet.plot(x='T', y='intensity', logx=True, legend=False, ax=ax,
label='Fréchet')
ax.set_title(site_name)
ax.set_ylabel('i (mm/h)')
ax.set_xlabel('return period (years)')
lines, labels = ax.get_legend_handles_labels()
lgd_labels = ['Observed values', 'Gumbel', 'GEV', 'Fréchet']
lgd = fig.legend(lines, lgd_labels, loc='lower center', ncol=4)
# plt.tight_layout()
plt.subplots_adjust(bottom=.24, wspace=None, hspace=None)
plt.savefig(os.path.join(plot_dir, fig_name))
plt.close()
def plot_gauges_data(ds, ymin, ymax, fig_name):
"""Plot the completeness of the gauges data
"""
da_year_count = ds['precipitation'].groupby('date.year').count(dim='date')
year_count_sel = da_year_count.sel(year=slice(ymin, ymax))
# plot
p = year_count_sel.plot(vmin=int(365*.9))
ax = plt.gca()
labels = list(da_year_count['code'].values)
ax.set_yticks(range(len(labels)))
ax.set_yticklabels(labels)
plt.tight_layout()
plt.savefig(os.path.join(plot_dir, fig_name))
plt.close()
def plot_gauges_map_from_ds(ds, id_dim, fig_name, global_extent=True):
"""convert dataset in geopandas DataFrame
plot it
"""
# Create a GeoDataFrame
ds_sel = ds.sel(duration=24, year=2000)#['annual_max']
df = ds_sel.to_dataframe().set_index(id_dim, drop=True).drop(axis=1, labels=['annual_max', 'year', 'duration'])
df['geometry'] = [shapely.geometry.Point(lon, lat) for lon, lat in zip(df['longitude'], df['latitude'])]
gdf = gpd.GeoDataFrame(
df, geometry='geometry')#.cx[slice(-2, -1.5), slice(51, 52)]
print(gdf)
# Plot the gauges on a world map
plt.figure(figsize=(12, 12))
if global_extent:
ax_p = plt.gca(projection=ctpy.crs.EqualEarth(), aspect='auto')
ax_p.set_global()
else:
ax_p = plt.gca(projection=ctpy.crs.PlateCarree(), aspect='auto')
gl = ax_p.gridlines(crs=ctpy.crs.PlateCarree(), linewidth=.3,
#color='black', alpha=0.5, linestyle='--',
draw_labels=False,
xlocs=np.arange(-10,10,0.25),
ylocs=np.arange(45,70,0.25),
)
ax_p.coastlines(linewidth=.3, color='black')
gdf.plot(ax=ax_p, markersize=5, transform=ctpy.crs.PlateCarree())
plt.savefig(os.path.join(plot_dir, fig_name))
plt.close()
def fig_gauges_map(fig_name):
source_crs = ctpy.crs.PlateCarree()
display_crs = ctpy.crs.OSGB()
midas = gpd.read_file(midas_stations)
ne_land = gpd.read_file(basemap)
# Extent of the UK
xmin = -7
xmax = 2
ymin = 49.5
ymax = 59.5
aspect = (xmax-xmin) / (ymax-ymin)
fig_width = 5
fig = plt.figure(figsize=(fig_width, fig_width*aspect))
ax_p = fig.gca(projection=display_crs,
aspect=aspect)
ticks = [0,1,2]
ne_land.plot(ax=ax_p, color='', edgecolor='0.2',
linewidth=0.4, alpha=0.4, transform=source_crs)
midas.plot(ax=ax_p, color='k', marker='x', markersize=15,
linewidth=1.5, label='MIDAS station', transform=source_crs)
ax_p.set_extent([xmin, xmax, ymin, ymax], crs=source_crs)
gl = ax_p.gridlines(crs=source_crs, linewidth=.3,
alpha=0.4,
)
fig_path = os.path.join(plot_dir, fig_name)
plt.savefig(fig_path)
subprocess.call(['pdfcrop', '--margins', '10', fig_path, fig_path])
plt.close()
def hexbin(da1, da2, xlabel, ylabel, fig_name):
x = da1.values.flat
y = da2.values.flat
xq = np.nanquantile(x, [0.01, 0.99])
yq = np.nanquantile(y, [0.01, 0.99])
xmin, xmax = xq[0], xq[1]
ymin, ymax = xq[0], xq[1]
fig, ax = plt.subplots(figsize=(4, 3))
# ax.axis([xmin, xmax, ymin, ymax])
hb = ax.hexbin(x, y, gridsize=20, extent=[xmin, xmax, ymin, ymax],
mincnt=100, bins=None)
ax.set_xlabel(xlabel)
ax.set_ylabel(ylabel)
ax.plot([xmin, xmax], [ymin, ymax], color='k', linewidth=0.5)
norm = matplotlib.colors.Normalize(vmin=1000, vmax=10000)
cb = plt.colorbar(hb, spacing='uniform', orientation='vertical',
label='Number of cells', norm=norm, extend='both')
# cb.set_label('# of cells')
plt.tight_layout()
plt.savefig(os.path.join(plot_dir, fig_name))
plt.close()
def fig_map_KS(ds):
"""create maps of the Kolmogorov-Smirnov / Lilliefors test statistic
"""
alpha = 0.05
DURATION = 2
Dcrit = ds['Dcrit'].sel(significance_level=alpha).values
Dmean = ds['KS_D'].mean(dim='duration')
single_map(Dmean,
# title='Lilliefors test statistic (1979-2018, mean on $d$, $\\alpha=${})'.format(alpha),
title='',
cbar_label="Kolmogorov-Smirnov $D$ ($D_{{\mathrm{{crit}}}} = {:.2f}$)".format(Dcrit),
center=Dcrit,
reverse=True,
fig_name='D_1979-2018_{}_dmean.png'.format(int(alpha*100)))
def fig_map_filliben(ds):
"""create maps of the Kolmogorov-Smirnov / Lilliefors test statistic
"""
alpha = 0.05
DURATION = 2
r_crit = ds['filliben_crit'].sel(significance_level=alpha).values
r_mean = ds['filliben_stat'].mean(dim='duration')
single_map(r_mean,
# title='Lilliefors test statistic (1979-2018, mean on $d$, $\\alpha=${})'.format(alpha),
title='',
cbar_label="Filliben's $r$ ($r_{{\mathrm{{crit}}}} = {:.2f}$)".format(r_crit),
center=r_crit,
reverse=False,
fig_name='filliben_1979-2018_{}_dmean.png'.format(int(alpha*100)))
def fig_maps_gev24h(ds):
da_loc = ds['location'].sel(duration=24)
da_scale = ds['scale'].sel(duration=24)
print(da_loc)
print(da_scale)
multi_maps([da_loc, da_scale],
['Location $\mu$', 'Scale $\sigma$'],
'Parameter value', 'gev_params_24h_1979-2018_twobars.png',
onebar=False,
bar_labels=[da_loc.attrs['units'], da_scale.attrs['units']]
)
def fig_maps_gev_scaled24h(ds):
da_loc = ds['gev_scaled'].sel(ci='estimate', duration=24, ev_param='location').rename('location')
da_scale = ds['gev_scaled'].sel(ci='estimate', duration=24, ev_param='scale').rename('scale')
da_shape = ds['gev_scaled'].sel(ci='estimate', duration=24, ev_param='shape').rename('shape')
multi_maps([da_loc, da_scale, da_shape],
['Location $\mu$', 'Scale $\sigma$', 'Shape $\kappa$'],
'Parameter value', 'gev_params_scaled_24h_1979-2018.png')
def fig_maps_r(ds):
multi_maps(ds.sel(scaling_extent=b'daily'), ['location_line_rvalue', 'scale_line_rvalue'],
['Location', 'Scale'],
'$r$', 'gumbel_r_daily.png', sqr=False)
def fig_maps_rsquared(ds):
da_scaling = ds['gev_scaling'].sel(ci='estimate', scaling_param='rsquared')
da_loc_r2 = da_scaling.sel(ev_param='location').rename('location')
da_scale_r2 = da_scaling.sel(ev_param='scale').rename('scale')
multi_maps([da_loc_r2, da_scale_r2],
['Location', 'Scale'],
'$r^2$', 'gev_r2.png', sqr=False)
def fig_maps_spearman(ds):
da_scaling = ds['gev_scaling'].sel(ci='estimate', scaling_param='spearman')
da_loc_rho = da_scaling.sel(ev_param='location').rename('location')
da_scale_rho = da_scaling.sel(ev_param='scale').rename('scale')
multi_maps([da_loc_rho, da_scale_rho],
['Location', 'Scale'],
'$\\rho$', 'gev_rho.png', sqr=False, robust=False)
def fig_scaling_differences_all(ds_era, ds_midas, fig_name):
quantiles = [0.025, 0.5, 0.975]
ds_era_sel = ds_era.sel(latitude=ds_midas['latitude'],
longitude=postprocessing.convert_lon(ds_midas['longitude']),
method='nearest')
q_extent_dict = {
'Whole world': postprocessing.get_quantile_dict(quantiles,
ERA5=(ds_era, ['latitude', 'longitude']),
),
'MIDAS stations': postprocessing.get_quantile_dict(quantiles,
ERA5=(ds_era_sel, ['station']),
MIDAS=(ds_midas, ['station'])
),
}
# Flatten the dict
q_list = []
for extent, df_dict in q_extent_dict.items():
for param, df_list in df_dict.items():
q_list.append({'extent':extent,
'param':param,
'df_list':df_list})
col_num = len(q_extent_dict)
row_num = 2
fig, axes = plt.subplots(row_num, col_num, figsize=(6, 3),
sharey='row', sharex=True)
ylim = {'location': [-0.25, 0.75], 'scale': [-0.6, 1.1]}
# param in rows, extent in columns
for row_num, (param, ax_row) in enumerate(zip(['location', 'scale'], axes)):
for extent, ax in zip(q_extent_dict, ax_row):
df_list = [i['df_list'] for i in q_list
if i['extent'] == extent
and i['param'] == param][0]
plot_scaling_differences(param, df_list, ax, ylim=False)
if row_num == 0:
ax.set_title(extent)
# set ticks
for ax in axes.flat:
set_logd_xticks(ax)
# add a big axes, hide frame
# fig.add_subplot(111, frameon=False)
# # hide tick and tick label of the big axes
# plt.tick_params(labelcolor='none', top=False,
# bottom=False, left=False, right=False)
# plt.grid(False)
# ylabel = '$\log_{{10}}( d^{{\eta}} / (\mu, \sigma) )$'
# plt.ylabel(ylabel)
handles, labels = ax.get_legend_handles_labels()
fig.legend(handles, labels, loc='lower center', ncol=2)
plt.tight_layout()
plt.subplots_adjust(bottom=0.35, left=0.15, wspace=None, hspace=None)
plt.savefig(os.path.join(plot_dir, fig_name), bbox='tight')
plt.close()
def plot_scaling_differences(param, df_list, ax, ylim=False):
"""
Plot quantiles of differences
"""
param_symbols = {'location': '\mu_d', 'scale': '\sigma_d'}
gradient_symbols = {'location': '\\alpha', 'scale': '\\beta'}
intercept_symbols = {'location': 'a', 'scale': 'b'}
colors = {'ERA5': C_PRIMARY_1, 'MIDAS': C_PRIMARY_2,
'MIDAS mean': C_PRIMARY_2, 'MIDAS stations': 'grey'}
for source, df in df_list:
quantiles = df.columns.values
q_min = quantiles[0]
q_max = quantiles[-1]
title = '{} ${}$'.format(param.title(), param_symbols[param])
fill_label = '{:.0%} of {}'.format(q_max-q_min, source)
df[0.5].plot(logx=True, logy=False, ax=ax,
linewidth=1, color=colors[source], zorder=10,
# title=title,
label='{} median'.format(source))
ax.fill_between(df.index, df[q_min], df[q_max],
facecolor=colors[source], alpha=0.2, zorder=1,
label=fill_label)
ax.axhline(0, linewidth=0.1, color='0.5')
# ax.axvline(24, linewidth=0.1, color='0.5')
if ylim:
ax.set_ylim(ylim[param])
# ax.set_yticks([ 0.0, 0.5])
set_logd_xticks(ax)
ax.set_xlabel('$d$ (hours)')
# ylabel = '$\log_{{10}}( {i}d^{g} / {p} )$'.format(g=gradient_symbols[param],
# p=param_symbols[param],
# i=intercept_symbols[param])
ylabel = "${i}d^{g} - {p}$".format(g=gradient_symbols[param],
p=param_symbols[param],
i=intercept_symbols[param])
ax.set_ylabel(ylabel)
ax.yaxis.set_major_formatter(FuncFormatter(lambda y, _: '{:.0%}'.format(y)))
def fig_scaling_ratio_map(ds):
# scaling_ratio = ds['scaling_ratio'].sel(scaling_extent=b'all')
alpha = ds['gev_scaling'].sel(scaling_param='slope', ci='estimate', ev_param='location')
beta = ds['gev_scaling'].sel(scaling_param='slope', ci='estimate', ev_param='scale')
scaling_ratio = alpha / beta
single_map(scaling_ratio,
title="",
cbar_label='$\\alpha / \\beta$',
center=1.0,
fig_name='scaling_ratio1979-2018.png')
def fig_scaling_gradients_maps(ds):
multi_maps([ds['alpha'], ds['beta']],
['Location scaling gradient $\\alpha$', 'Scale scaling gradient $\\beta$'],
'Parameter value', 'scaling_gradients_1979-2018_twobars.png',
onebar=False,
bar_labels=['$\\alpha$', '$\\beta$'])
def fig_scaling_intersects_maps(ds):
multi_maps([ds['a'], ds['b']],
['a', 'b'],
'Parameter value', 'scaling_intersects_1979-2018_twobars.png',
onebar=False,
bar_labels=['a', 'b'])
def fig_scaling_gradients_ratio_maps(ds):
"""print maps of the ratio between the regression slopes obtain from daily and all durations
"""
ds_daily = ds.sel(scaling_extent=b'daily')
ds_all = ds.sel(scaling_extent=b'all')
ds['loc_ratio'] = ds_all['location_line_slope'] / ds_daily['location_line_slope']
ds['scale_ratio'] = ds_all['scale_line_slope'] / ds_daily['scale_line_slope']
multi_maps(ds,
var_names=['loc_ratio', 'scale_ratio'],
disp_names=['$\\alpha_{all} / \\alpha_{daily}$',
'$\\beta_{all} / \\beta_{daily}$'],
value_name="Ratio", fig_name='scaling_gradients_ratio.png',
sqr=False, center=.5, reverse=True)
def fig_scaling_hexbin(ds):
da1 = ds['gev_scaling'].sel(scaling_param='slope', ci='estimate', ev_param='location')
da2 = ds['gev_scaling'].sel(scaling_param='slope', ci='estimate', ev_param='scale')
hexbin(da1, da2,
'$\\alpha$', '$\\beta$',
'scaling_gradients_hexbin.png')
def table_r_sigcount(ds, alpha, dim):
"""Count the number of points with p > alpha (i.e. not significant).
"""
ds_r = ds[['location_line_pvalue', 'scale_line_pvalue']].sel(scaling_extent=[
b'daily', b'all'])
print(ds_r.where(ds_r > alpha).count(dim=dim).load())
def table_count_noGEVfit(ds):
da_loc = ds['gev'].sel(ci='estimate', ev_param='location')
num_null = (~np.isfinite(da_loc)).sum(dim=['latitude', 'longitude'])
print(num_null.load())
def table_count_noscalingfit(ds):
da_scaling_slope = ds['gev_scaling'].sel(ci='estimate', ev_param=['location', 'scale'], scaling_param='slope')
num_null = np.isnan(da_scaling_slope).sum(dim=['latitude', 'longitude'])
print(num_null.load())
def table_rsquared_quantiles(ds, dim):
ds_rsquared = ds['gev_scaling'].sel(ci='estimate', ev_param=['location', 'scale', 'shape'], scaling_param='rsquared')
q = ds_rsquared.load().quantile([0.01, 0.05], dim=dim)
print(q)
def table_gof_quantiles(ds, dim):
q_levels = [0.001, 0.01, 0.05, 0.1, 0.9, 0.95,0.99,0.999]
q_D = ds['KS_D'].load().quantile(q_levels)
print("Dcrit: {}".format(ds['Dcrit'].load()))
print(q_D)
q_r = ds['filliben_stat'].load().quantile(q_levels)
print("r_crit: {}".format(ds['filliben_crit'].load()))
print(q_r)
def table_gof_count(ds):
"""count the number of cells x durations without good fit.
"""
n_all = ds['filliben_stat'].count().load()
nofit = (ds['filliben_stat'] < ds['filliben_crit'].sel(significance_level=0.05))
n_nofit = xr.where(nofit, 1., np.nan).count().load()
print(n_nofit/n_all)
def table_sizing_values(ds, coords, d):
"""Values over for the culvert sizing exercise.
"""
ds_sel = ds.sel(latitude=coords[0],
longitude=postprocessing.convert_lon(coords[1]),
method='nearest').load()
ds_sel = ds_sel.sel(ci=['estimate', '0.025', '0.975'],
duration=d,
ev_param=['location', 'scale'])
# fitted
da_fitted = ds_sel['gev']
print(da_fitted)
print(ds_sel['gev_scaled'])
print(ds_sel['gev_scaling'].sel(scaling_param=['slope', 'intercept']))
def prepare_midas_mean(ds_era, ds_midas, ds_midas_mean):
keep_vars = ['location', 'scale', 'log_location', 'log_scale',
'location_line_slope', 'scale_line_slope',
'location_line_intercept', 'scale_line_intercept']
# convert station names to UTF8
names_str = np.array([b.decode('utf8') for b in ds_midas['src_name'].values])
ds_midas['src_name'].values = names_str
# Set station coordinates to station name
ds_midas.coords['station'] = ds_midas['src_name']
ds_midas = ds_midas.drop('src_name')
ds_list = []
for pair_name in ds_midas_mean['pair']:
pair_name = str(pair_name.values)
ds_pair_list = []
# Find MIDAS stations in pairs
for i, s in enumerate(pair_name.split('--')):
site_num = i+1
ds_s = ds_midas[keep_vars].sel(station=s, drop=True)
site_lat = ds_s['latitude'].values
site_lon = postprocessing.convert_lon(ds_s['longitude'].values)
# Add a source coordinate
ds_s = ds_s.expand_dims(['source', 'pair'])
ds_s.coords['source'] = ['MIDAS s'+str(site_num)]
ds_s.coords['pair'] = [pair_name]
ds_pair_list.append(ds_s.drop(['latitude', 'longitude']))
# add the source dim to the other two datasets
ds_mean = ds_midas_mean[keep_vars].sel(pair=pair_name)
ds_mean = ds_mean.expand_dims(['pair', 'source'])
ds_mean.coords['source'] = ['MIDAS mean']
ds_pair_list.append(ds_mean)
ds_era_extract = ds_era[keep_vars].sel(latitude=site_lat,
longitude=site_lon,
method='nearest', drop=True)
ds_era_extract.coords['scaling_extent'] = ds_mean['scaling_extent']
ds_era_extract = ds_era_extract.expand_dims(['pair', 'source'])
ds_era_extract.coords['pair'] = [pair_name]
ds_era_extract.coords['source'] = ['ERA5']
ds_pair_list.append(ds_era_extract)
# Join all ds of the cell in a single dataset
ds_site = xr.concat(ds_pair_list, dim='source')
ds_list.append(ds_site)
# Join all along the pair coordinate
ds_all = xr.concat(ds_list, dim='pair')
# Calculate regression lines
for p in ['location', 'scale']:
dur = ds_all['duration']
slope = ds_all['{}_line_slope'.format(p)]
intercept = ds_all['{}_line_intercept'.format(p)]
linereg_var = '{}_lr'.format(p)
ds_all[linereg_var] = (10**intercept) * dur**slope
return ds_all
def fig_midas_mean(ds_pairs, fig_name):
dict_df = postprocessing.ds_to_df(ds_pairs, 'pair')
# # From seaborn colorblind
# C = ['#0173b2', '#de8f05', '#029e73', '#cc78bc']
C = {'ERA5': C_PRIMARY_1, 'MIDAS mean': C_PRIMARY_2,
'MIDAS s1': 'grey','MIDAS s2': 'grey'}
M = {'ERA5': 'o', 'MIDAS mean': 'v',
'MIDAS s1': 'x','MIDAS s2': 'x'}
# MS = {'ERA5': 4, 'MIDAS mean': 5,
# 'MIDAS s1': 2,'MIDAS s2': 2}
linesyles = {
'{}_{}': dict(linestyle='None', linewidth=0, marker='o', markersize=4,
label='{}'),
# '{}_{}_lr_all': dict(linestyle='dashed', linewidth=1.5, marker=None, markersize=0,
# label='{} (all)'),
'{}_{}_lr_daily': dict(linestyle='-', linewidth=1., marker=None, markersize=0,
label='{} (daily)'),
}
ax_list = []
col_num = 2
row_num = len(dict_df) + 1
fig_size = (3.5*col_num, 2.2*row_num)
# fig = plt.figure(figsize=fig_size)
# ax_num = 1
fig, axes = plt.subplots(row_num, col_num, figsize=fig_size,
sharey=False, sharex=True)
for row_idx, ((pair_name, df), ax_row) in enumerate(zip(dict_df.items(), axes[:-1])):
print(pair_name)
for param_name, ax in zip(['location', 'scale'], ax_row):
ax_list.append(ax)
for source in ds_pairs['source'].values:
for col_base, style in linesyles.items():
col = col_base.format(source, param_name)
if source.startswith('MIDAS s'):
if 'lr' in col:
linestyle = 'dashed'
marker = None
markersize = None
if source == 'MIDAS s1':
label = 'MIDAS stations (daily)'
else:
label = ''
else:
linestyle = 'none'
marker = 'x'
linewidth = .5
markersize = 2
if source == 'MIDAS s1':
label = 'MIDAS stations'
else:
label = ''
else:
linestyle = style['linestyle']
linewidth = style['linewidth']
marker = M[source]
markersize = style['markersize']
label = style['label'].format(source)
df[col].plot(ax=ax, label=label,
loglog=True,
linestyle=linestyle, linewidth=linewidth,
markersize=markersize, marker=marker,
color=C[source])
# except KeyError:
# continue
if row_idx == 0:
ax.set_title(param_name)
if param_name == 'location':
ax.set_ylabel('{}'.format(pair_name))
lines, labels = ax.get_legend_handles_labels()
ax.set_xlabel('$d$ (hours)')
## Difference plot ##
# build the dict of ds for calculating quantiles
sites_list = []
dict_quantiles = {}
for s in ds_pairs.source.values:
ds_source = ds_pairs.sel(source=s)
if str(s).startswith('MIDAS s'):
sites_list.append(ds_source)
elif str(s).startswith('MIDAS'):
dict_quantiles['MIDAS mean'] = (ds_source, ['pair'])
elif str(s).startswith('ERA'):
dict_quantiles['ERA5'] = (ds_source, ['pair'])
ds_stations = xr.concat(sites_list, dim='pair')
dict_quantiles['MIDAS stations'] = (ds_stations, ['pair'])
quantiles = [0.01, 0.5, 0.99]
q_dict = get_quantile_dict(quantiles, **dict_quantiles)
# param in rows, extent in columns
for param, df_list, ax_diff in zip(['location', 'scale'], q_dict.values(), axes[-1]):
# ax_diff = fig.add_subplot(row_num, col_num, ax_num)
plot_scaling_differences(param, df_list, ax_diff)
# ax_diff.set_title(param)
ax_list.append(ax_diff)
# set ticks
for ax in ax_list:
set_logd_xticks(ax)
lgd = fig.legend(lines, labels, loc='lower center', ncol=3)
plt.tight_layout()
plt.subplots_adjust(bottom=.1, wspace=None, hspace=None)
plt.savefig(os.path.join(plot_dir, fig_name))
plt.close()
def scatter_intensity(da_i, fig_name):
"""
"""
df_i = da_i.to_dataframe()
# Split intensities in two columns, one for each source
da_list = []
for source in da_i['source'].values:
series_i = df_i.xs(source, level='source').rename(columns={'intensity':source})
da_list.append(series_i)
df_i_s = pd.concat(da_list, axis='columns').reset_index()
# Plot some durations on facetgrid
dur_sel = [1,6,24,240]
df_i_s = df_i_s.loc[np.in1d(df_i_s['duration'], dur_sel)]
fg = sns.FacetGrid(df_i_s, sharex=False, sharey=False, row='T', col='duration')
fg = fg.map(plt.plot, 'ERA5_scaled_rlm', 'MIDAS', color=C_PRIMARY_2, label='ERA5 scaled LTS')
fg = fg.map(plt.scatter, 'ERA5_scaled', 'MIDAS', color=C_PRIMARY_2, label='ERA5 scaled')
fg = fg.map(plt.plot, 'ERA5_rlm', 'MIDAS', color=C_PRIMARY_1, label='ERA5 LTS')
fg = fg.map(plt.scatter, 'ERA5', 'MIDAS', color=C_PRIMARY_1, label='ERA5')
for ax in fg.axes.flatten():
lines, labels = ax.get_legend_handles_labels()
fig = plt.gcf()
lgd = fig.legend(lines, labels, loc='lower center', ncol=4)
plt.tight_layout()
plt.subplots_adjust(bottom=.04, wspace=None, hspace=None)
plt.savefig(os.path.join(plot_dir, fig_name))
plt.close()
def plot_ARF(da, fig_name):
width = 7
aspect = 1.2