-
Notifications
You must be signed in to change notification settings - Fork 17
/
osh5utils_q3d.py
188 lines (137 loc) · 7.82 KB
/
osh5utils_q3d.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
# ******************************************************************************************************
# ******************************************************************************************************
# ******************************************************************************************************
# ******************************************************************************************************
#
# osh5utils_q3d.py
# quasi-3d utilities for pyVisOS
#
# Revision History
# Version 1: First commit, made a subroutine that converts q3d data to full
# 3d data via mode summation
#
#
#
# ******************************************************************************************************
# ******************************************************************************************************
# ******************************************************************************************************
# ******************************************************************************************************
import osh5io
import osh5def
import osh5vis
import osh5utils
import matplotlib.pyplot as plt
import osh5visipy
# ******************************************************************************************************
# ******************************************************************************************************
# ******************************************************************************************************
# ******************************************************************************************************
def filename_re(rundir,plasma_field,mode,fileno):
plasma_field_path=rundir+'/MS/FLD/MODE-{mode:1d}-RE/{plasma_field:s}_cyl_m/{plasma_field:s}_cyl_m-{mode:1d}-re-{fileno:06d}.h5'.format(plasma_field=plasma_field,mode=mode,fileno=fileno)
# print(plasma_field_path)
return(plasma_field_path)
def filename_im(rundir,plasma_field,mode,fileno):
plasma_field_path=rundir+'/MS/FLD/MODE-{mode:1d}-RE/{plasma_field:s}_cyl_m/{plasma_field:s}_cyl_m-{mode:1d}-re-{fileno:06d}.h5'.format(plasma_field=plasma_field,mode=mode,fileno=fileno)
# print(plasma_field_path)
return(plasma_field_path)
def filename_3d(rundir,plasma_field,fileno):
filename=rundir+'/MS/FLD/{plasma_field:s}/{plasma_field:s}-{fileno:06d}.h5'.format(plasma_field=plasma_field,fileno=fileno)
return(filename)
# ******************************************************************************************************
# ******************************************************************************************************
# ******************************************************************************************************
# ******************************************************************************************************
def q3d_to_3d(rundir,plasma_field,fileno,mode_max,x1_min,x1_max,nx1,x2_min,x2_max,nx2,x3_min,x3_max,nx3):
from scipy import interpolate
dx1=(x1_max-x1_min)/(nx1-1)
dx2=(x2_max-x2_min)/(nx2-1)
dx3=(x3_max-x3_min)/(nx3-1)
x1_axis=np.arange(x1_min,x1_max+dx1,dx1)
x2_axis=np.arange(x2_min,x2_max+dx2,dx2)
x3_axis=np.arange(x3_min,x3_max+dx3,dx3)
a = np.zeros((nx1,nx2,nx3),dtype=float)
filename_out = filename_3d(rundir,plasma_field,fileno)
x1 = osh5def.DataAxis(x1_min,x1_max, nx1, attrs={'NAME':'x1', 'LONG_NAME':'x_1', 'UNITS':'c / \omega_0'})
x2 = osh5def.DataAxis(x2_min,x2_max, nx2, attrs={'NAME':'x2', 'LONG_NAME':'x_2', 'UNITS':'c / \omega_0'})
x3 = osh5def.DataAxis(x3_min,x3_max, nx3, attrs={'NAME':'x3', 'LONG_NAME':'x_3', 'UNITS':'c / \omega_0'})
# More attributes associated with the data/simulation. Again no need to worry about the details.
data_attrs = {'UNITS': osh5def.OSUnits('m_e c \omega_0 / e'), 'NAME': plasma_field, 'LONG_NAME': plasma_field}
run_attrs = {'NOTE': 'parameters about this simulation are stored here', 'TIME UNITS': '1/\omega_0',
'XMAX':np.array([1., 15.]), 'XMIN':np.array([0., 10.])}
# Now "wrap" the numpy array into osh5def.H5Data. Note that the data and the axes are consistent and are in fortran ordering
b = osh5def.H5Data(a, timestamp='123456', data_attrs=data_attrs, run_attrs=run_attrs, axes=[x1, x2, x3])
# I am doing mode 0 outside of the loop
fname_re=filename_re(rundir,plasma_field,0,fileno)
# DEBUG
print(fname_re)
# DEBUG
data_re = osh5io.read_h5(fname_re)
print(data_re.shape)
print(data_re.axes[1].ax.shape)
print(data_re.axes[0].ax.shape)
func_re = interpolate.interp2d(data_re.axes[1].ax,data_re.axes[0].ax,data_re,kind='cubic')
for i1 in range(0,nx1):
for i2 in range(0,nx2):
for i3 in range(0,nx3):
z = x1_axis[i1]
x = x3_axis[i3]
y = x2_axis[i2]
r = np.sqrt(x*x+y*y)
# if r != 0:
# cos_th = x/r
# sin_th = y/r
# else:
# cos_th = 1
# sin_th = 0
a[i1,i2,i3]=a[i1,i2,i3]+func_re(z,r)
for i_mode in range(1,mode_max+1):
fname_re=filename_re(rundir,plasma_field,i_mode,fileno)
fname_im=filename_im(rundir,plasma_field,i_mode,fileno)
# DEBUG
print(fname_re)
# DEBUG
if(plasma_field =='e2' or plasma_field == 'e3'):
if (plasma_field == 'e2'):
field_comp = 'e3'
else:
field_comp = 'e2'
data_re_self=osh5io.read_h5(filename_re(rundir,plasma_field,i_mode,fileno))
data_im_self=osh5io.read_h5(filename_im(rundir,plasma_field,i_mode,fileno))
data_re_comp=osh5io.read_h5(filename_re(rundir,field_comp,i_mode,fileno))
data_im_comp=osh5io.read_h5(filename_im(rundir,field_comp,i_mode,fileno))
else:
data_re=osh5io.read_h5(filename_re(rundir,plasma_field,i_mode,fileno))
data_im=osh5io.read_h5(filename_im(rundir,plasma_field,i_mode,fileno))
func_re = interpolate.interp2d(data_re.axes[1].ax,data_re.axes[0].ax,data_re,kind='cubic')
func_im = interpolate.interp2d(data_im.axes[1].ax,data_im.axes[0].ax,data_im,kind='cubic')
for i1 in range(0,nx1):
for i2 in range(0,nx2):
for i3 in range(0,nx3):
z = x1_axis[i1]
x = x3_axis[i3]
y = x2_axis[i2]
r = np.sqrt(x*x+y*y)
if r > 0.000001:
cos_th = x/r
sin_th = y/r
else:
cos_th = 1
sin_th = 0
# start the recursion relation to evaluate cos(n*theta) and sin(n_theta)
sin_n=sin_th
cos_n=cos_th
for int_mode in range(2,i_mode+1):
temp_s=sin_n
temp_c=cos_n
cos_n=temp_c*cos_th-temp_s*sin_th
sin_n=temp_s*cos_th+temp_c*sin_th
#
# here we perform the addition of the N-th mode
# to the data in 3D
#
a[i1,i2,i3]=a[i1,i2,i3]+func_re(z,r)*cos_n-func_im(z,r)*sin_n
osh5io.write_h5(b,filename=filename_out)
# ******************************************************************************************************
# ******************************************************************************************************
# ******************************************************************************************************
# ******************************************************************************************************