|
1
|
|
|
# Copyright (c) 2020 Stefan Bender |
|
2
|
|
|
# |
|
3
|
|
|
# This module is part of pyspaceweather. |
|
4
|
|
|
# pyspaceweather is free software: you can redistribute it or modify |
|
5
|
|
|
# it under the terms of the GNU General Public License as published |
|
6
|
|
|
# by the Free Software Foundation, version 2. |
|
7
|
|
|
# See accompanying COPYING.GPLv2 file or http://www.gnu.org/licenses/gpl-2.0.html. |
|
8
|
|
|
"""Python interface for space weather indices |
|
9
|
|
|
|
|
10
|
|
|
""" |
|
11
|
|
|
import os |
|
12
|
|
|
from pkg_resources import resource_filename |
|
13
|
|
|
import requests |
|
14
|
|
|
import logging |
|
15
|
|
|
from warnings import warn |
|
16
|
|
|
|
|
17
|
|
|
import numpy as np |
|
18
|
|
|
import pandas as pd |
|
19
|
|
|
|
|
20
|
|
|
__all__ = [ |
|
21
|
|
|
"sw_daily", "ap_kp_3h", "read_sw", |
|
22
|
|
|
"get_file_age", "update_data", |
|
23
|
|
|
"SW_PATH_ALL", "SW_PATH_5Y", |
|
24
|
|
|
] |
|
25
|
|
|
|
|
26
|
|
|
DL_URL_ALL = "https://celestrak.com/SpaceData/SW-All.txt" |
|
27
|
|
|
DL_URL_5Y = "https://celestrak.com/SpaceData/SW-Last5Years.txt" |
|
28
|
|
|
SW_FILE_ALL = os.path.basename(DL_URL_ALL) |
|
29
|
|
|
SW_FILE_5Y = os.path.basename(DL_URL_5Y) |
|
30
|
|
|
SW_PATH_ALL = resource_filename(__name__, os.path.join("data", SW_FILE_ALL)) |
|
31
|
|
|
SW_PATH_5Y = resource_filename(__name__, os.path.join("data", SW_FILE_5Y)) |
|
32
|
|
|
|
|
33
|
|
|
|
|
34
|
|
|
def _dl_file(swpath, url=DL_URL_ALL): |
|
35
|
|
|
with requests.get(url, stream=True) as r: |
|
36
|
|
|
with open(swpath, 'wb') as fd: |
|
37
|
|
|
for chunk in r.iter_content(chunk_size=1024): |
|
38
|
|
|
fd.write(chunk) |
|
39
|
|
|
|
|
40
|
|
|
|
|
41
|
|
|
def get_file_age(swpath, relative=True): |
|
42
|
|
|
for line in open(swpath): |
|
43
|
|
|
if line.startswith("UPDATED"): |
|
44
|
|
|
# closes the file automatically |
|
45
|
|
|
break |
|
46
|
|
|
upd = pd.to_datetime(line.lstrip("UPDATED"), utc=True) |
|
|
|
|
|
|
47
|
|
|
if relative: |
|
48
|
|
|
return pd.Timestamp.utcnow() - upd |
|
49
|
|
|
return upd |
|
50
|
|
|
|
|
51
|
|
|
|
|
52
|
|
|
def update_data( |
|
53
|
|
|
min_age="3h", |
|
54
|
|
|
swpath_all=SW_PATH_ALL, swpath_5y=SW_PATH_5Y, |
|
55
|
|
|
url_all=DL_URL_ALL, url_5y=DL_URL_5Y, |
|
56
|
|
|
): |
|
57
|
|
|
def _update_file(swpath, url, min_age): |
|
58
|
|
|
if not os.path.exists(swpath): |
|
59
|
|
|
logging.info("{0} not found, downloading.".format(swpath)) |
|
60
|
|
|
_dl_file(swpath, url) |
|
61
|
|
|
return |
|
62
|
|
|
if get_file_age(swpath) < pd.Timedelta(min_age): |
|
63
|
|
|
logging.info("not updating '{0}'.".format(swpath)) |
|
64
|
|
|
return |
|
65
|
|
|
logging.info("updating '{0}'.".format(swpath)) |
|
66
|
|
|
_dl_file(swpath, url) |
|
67
|
|
|
|
|
68
|
|
|
# Update the large file after four years |
|
69
|
|
|
# to have some overlap with the 5-year data |
|
70
|
|
|
_update_file(swpath_all, url_all, "4y") |
|
71
|
|
|
# Don't re-download before `min_age` has passed (3h) |
|
72
|
|
|
_update_file(swpath_5y, url_5y, min_age) |
|
73
|
|
|
|
|
74
|
|
|
|
|
75
|
|
|
def read_sw(swpath): |
|
76
|
|
|
kpns = ["Kp{0}".format(i) for i in range(0, 23, 3)] + ["Kpsum"] |
|
77
|
|
|
sw = np.genfromtxt( |
|
78
|
|
|
swpath, |
|
79
|
|
|
skip_header=3, |
|
80
|
|
|
# yy mm dd br rd kp kp kp kp kp kp kp kp Kp ap ap ap ap ap ap ap ap Ap cp c9 is f1 q f2 f3 f4 f5 f6 |
|
81
|
|
|
delimiter=[4, 3, 3, 5, 3, 3, 3, 3, 3, 3, 3, 3, 3, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 2, 4, 6, 2, 6, 6, 6, 6, 6], |
|
82
|
|
|
dtype= "i4,i4,i4,i4,i4,i4,i4,i4,i4,i4,i4,i4,i4,i4,i4,i4,i4,i4,i4,i4,i4,i4,i4,f8,i4,i4,f8,i4,f8,f8,f8,f8,f8", |
|
83
|
|
|
names=[ |
|
84
|
|
|
"year", "month", "day", "bsrn", "rotd", |
|
85
|
|
|
"Kp0", "Kp3", "Kp6", "Kp9", "Kp12", "Kp15", "Kp18", "Kp21", "Kpsum", |
|
86
|
|
|
"Ap0", "Ap3", "Ap6", "Ap9", "Ap12", "Ap15", "Ap18", "Ap21", "Apavg", |
|
87
|
|
|
"Cp", "C9", "isn", "f107_adj", "Q", "f107_81ctr_adj", "f107_81lst_adj", |
|
88
|
|
|
"f107_obs", "f107_81ctr_obs", "f107_81lst_obs" |
|
89
|
|
|
] |
|
90
|
|
|
)[2:-1] |
|
91
|
|
|
sw = sw[sw["year"] != -1] |
|
92
|
|
|
ts = pd.to_datetime([ |
|
93
|
|
|
"{0:04d}-{1:02d}-{2:02d}".format(yy, mm, dd) |
|
94
|
|
|
for yy, mm, dd in sw[["year", "month", "day"]] |
|
95
|
|
|
]) |
|
96
|
|
|
sw_df = pd.DataFrame(sw, index=ts) |
|
97
|
|
|
sw_df[kpns] = 0.1 * sw_df[kpns] |
|
98
|
|
|
return sw_df |
|
99
|
|
|
|
|
100
|
|
|
|
|
101
|
|
|
def sw_daily(swpath_all=SW_PATH_ALL, swpath_5y=SW_PATH_5Y, update_interval="30days"): |
|
102
|
|
|
"""Daily Ap, Kp, and f10.7 index values |
|
103
|
|
|
""" |
|
104
|
|
|
# ensure that the file exists and is up to date |
|
105
|
|
|
if ( |
|
106
|
|
|
not os.path.exists(swpath_all) |
|
107
|
|
|
or not os.path.exists(swpath_5y) |
|
108
|
|
|
): |
|
109
|
|
|
warn("Could not find space weather data, trying to download.") |
|
110
|
|
|
update_data() |
|
111
|
|
|
|
|
112
|
|
|
if ( |
|
113
|
|
|
get_file_age(swpath_all) > pd.Timedelta("5y") |
|
114
|
|
|
or get_file_age(swpath_5y) > pd.Timedelta(update_interval) |
|
115
|
|
|
): |
|
116
|
|
|
warn("Data files *might* be too old, consider running `sw.update_data()`.") |
|
117
|
|
|
|
|
118
|
|
|
df_all = read_sw(swpath_all) |
|
119
|
|
|
df_5y = read_sw(swpath_5y) |
|
120
|
|
|
return pd.concat([df_all[:df_5y.index[0]], df_5y[1:]]) |
|
121
|
|
|
|
|
122
|
|
|
|
|
123
|
|
|
def ap_kp_3h(swpath_all=SW_PATH_ALL, swpath_5y=SW_PATH_5Y, update_interval="30days"): |
|
124
|
|
|
"""3h Ap and Kp index values |
|
125
|
|
|
""" |
|
126
|
|
|
daily_df = sw_daily( |
|
127
|
|
|
swpath_all=swpath_all, swpath_5y=swpath_5y, update_interval=update_interval |
|
128
|
|
|
) |
|
129
|
|
|
ret = daily_df.copy() |
|
130
|
|
|
apns = ["Ap{0}".format(i) for i in range(0, 23, 3)] |
|
131
|
|
|
kpns = ["Kp{0}".format(i) for i in range(0, 23, 3)] |
|
132
|
|
|
for i, (ap, kp) in enumerate(zip(apns, kpns)): |
|
133
|
|
|
ret[ap].index = daily_df[ap].index + pd.Timedelta((i * 3 + 1.5), unit="h") |
|
134
|
|
|
ret[kp].index = daily_df[kp].index + pd.Timedelta((i * 3 + 1.5), unit="h") |
|
135
|
|
|
sw_ap = pd.concat([ret[ap] for ap in apns]) |
|
136
|
|
|
sw_kp = pd.concat([ret[kp] for kp in kpns]) |
|
137
|
|
|
return pd.DataFrame({"Ap": sw_ap, "Kp": sw_kp}) |
|
138
|
|
|
|