-
Notifications
You must be signed in to change notification settings - Fork 1
/
vortex.py
executable file
·159 lines (117 loc) · 4.06 KB
/
vortex.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
#!/usr/bin/env python3
#
# read map file, apply to vortex data, compute error
#
import os, time
from netCDF4 import Dataset
from math import pi
import numpy as np
from numpy import sin,cos,arctan2,arcsin,cosh,tanh,sqrt
import scipy as sp
import scipy.sparse as sparse
from plotpoly_mpl import plotpoly
from test_fields import test_fields
# use scipy instead:
# import numba
# @numba.njit()
# def apply_map( data_a, S, row, col,n_b ):
# data_b=numpy.zeros(n_b)
# for k in range(len(S)):
# data_b[row[k]] = data_b[row[k]] + data_a[col[k]] * S[k]
# return data_b
if len(os.sys.argv) < 2:
print("./vortex.py mapfilename.nc [vortex,y16_23,y2_2]")
os.sys.exit(1)
mapfile=(os.sys.argv[1])
plotfile=mapfile.split(".nc")[0]
testfield='vortex'
if len(os.sys.argv) >= 3:
testfield=os.sys.argv[2]
#################################################################################33
#
# main code
#
#################################################################################33
mapf = Dataset(mapfile,"r")
print("reading",mapfile)
S=mapf.variables['S'][:]
row=mapf.variables['row'][:]
col=mapf.variables['col'][:]
row=row-1 # convert to zero indexing
col=col-1
lat_a = mapf.variables['yc_a'][:]
lon_a = mapf.variables['xc_a'][:]
lat_b = mapf.variables['yc_b'][:]
lon_b = mapf.variables['xc_b'][:]
area_b = mapf.variables['area_b'][:]
# should we check units?
#if "radian" in infile.variables["xc_a"].units.lower():
deg_to_rad=pi/180
#print("lat_a",len(lat_a),"min/max",min(lat_a),max(lat_a))
if max(abs(lat_a))>1.1*pi:
#print(" converting to radians")
lat_a = lat_a*deg_to_rad
lon_a = lon_a*deg_to_rad
#print("lat_b",len(lat_b),"min/max",min(lat_b),max(lat_b))
if max(abs(lat_b))>1.1*pi:
#print(" converting to radians")
lat_b = lat_b*deg_to_rad
lon_b = lon_b*deg_to_rad
n_a = len(lat_a)
n_b = len(lat_b)
data_a=np.zeros(n_a)
data_b=np.zeros(n_b)
data_b_exact=np.zeros(n_b)
data_a=test_fields(lon_a,lat_a,testfield)
data_b_exact=test_fields(lon_b,lat_b,testfield)
print("applying mapfile...")
# NOT VALID:
#data_b[row[:]] += data_a[col[:]]*S[:]
# SLOW! ne30np4_to_ne1024pg2: 296s
# tic=time.perf_counter()
# for i in range(len(S)): # very slow
# data_b[row[i]] += data_a[col[i]]*S[i]
# toc=time.perf_counter()
# print(f"python loop:: {toc - tic:0.4f} seconds")
# fastest. ne30np4_to_ne1024pg2: 0.42s
tic=time.perf_counter()
data2 = np.stack( (data_a, np.ones_like(data_a)), axis=1) # combine into (m,2) matrix
data_b2 = sparse.coo_matrix((S, (row,col)), shape=(n_b,n_a)) @ data2 # need scypi
data_b=data_b2[:,0]
mask_b=(data_b2[:,1] != 0)
toc=time.perf_counter()
print(f"apply map via sparse.coo_matrix: {toc - tic:0.4f} seconds")
# # Fast, but not as fast as scypy: 0.61s
# S=np.ma.getdata(S)
# data_a=np.ma.getdata(data_a)
# row=np.ma.getdata(row)
# col=np.ma.getdata(col)
# tic=time.perf_counter()
# data_b=apply_map(data_a,S,row,col,n_b)
# toc=time.perf_counter()
# print(f"apply map via numba: {toc - tic:0.4f} seconds")
# remove points where map(1) == 0
# devide by map(1) to compute correct error:
data_b=data_b[mask_b] / data_b2[mask_b,1]
data_b_exact=data_b_exact[mask_b]
area_b=area_b[mask_b]
# compute error between data_b and data_b_exact
# only compute error where map(1) <> 0
max_err = max( abs(data_b-data_b_exact) ) / max( abs( data_b_exact ))
l2_err = sum( area_b*(data_b-data_b_exact)**2 ) / sum(area_b)
l2_err = sqrt(l2_err)
print(testfield,": pointwise relative error l2=%.3e max=%.3e" % (l2_err,max_err))
# read in the cell polygons
lat_b = mapf.variables['yv_b'][:,:]
lon_b = mapf.variables['xv_b'][:,:]
lat_b = lat_b[mask_b,:]
lon_b = lon_b[mask_b,:]
if np.max(np.abs(lat_b))<1.1*pi:
#print(" converting to degrees")
lat_b = lat_b/deg_to_rad
lon_b = lon_b/deg_to_rad
plotpoly(lat_b,lon_b,data_b,plotfile+'_field.png')
#plotpoly(lat_b,lon_b,data_b-data_b_exact,plotfile+'_error.png',colormap='Spectral')
# error plot zoomed in over UK, with bounds specified
plotpoly(lat_b,lon_b,data_b-data_b_exact,plotfile+'_error.png',clim=(-.005,.005),
colormap='Spectral',xlim=(-35,35),ylim=(10,80))