|
1
|
|
|
""" |
|
2
|
|
|
Main AsgardpyConfig Operations Module |
|
3
|
|
|
""" |
|
4
|
|
|
|
|
5
|
|
|
import logging |
|
6
|
|
|
from collections.abc import Mapping |
|
7
|
|
|
from pathlib import Path |
|
8
|
|
|
|
|
9
|
|
|
import numpy as np |
|
10
|
|
|
from gammapy.modeling.models import Models, SkyModel |
|
11
|
|
|
|
|
12
|
|
|
__all__ = [ |
|
13
|
|
|
"all_model_templates", |
|
14
|
|
|
"compound_model_dict_converstion", |
|
15
|
|
|
"get_model_template", |
|
16
|
|
|
"recursive_merge_dicts", |
|
17
|
|
|
"deep_update", |
|
18
|
|
|
] |
|
19
|
|
|
|
|
20
|
|
|
CONFIG_PATH = Path(__file__).resolve().parent |
|
21
|
|
|
|
|
22
|
|
|
log = logging.getLogger(__name__) |
|
23
|
|
|
|
|
24
|
|
|
|
|
25
|
|
|
def all_model_templates(): |
|
26
|
|
|
""" |
|
27
|
|
|
Collect all Template Models provided in Asgardpy, and their small tag names. |
|
28
|
|
|
""" |
|
29
|
|
|
template_files = sorted(list(CONFIG_PATH.glob("model_templates/model_template*yaml"))) |
|
30
|
|
|
|
|
31
|
|
|
all_tags = [] |
|
32
|
|
|
for file in template_files: |
|
33
|
|
|
all_tags.append(file.name.split("_")[-1].split(".")[0]) |
|
34
|
|
|
all_tags = np.array(all_tags) |
|
35
|
|
|
|
|
36
|
|
|
return all_tags, template_files |
|
37
|
|
|
|
|
38
|
|
|
|
|
39
|
|
|
def get_model_template(spec_model_tag): |
|
40
|
|
|
""" |
|
41
|
|
|
Read a particular template model yaml filename to create an AsgardpyConfig |
|
42
|
|
|
object. |
|
43
|
|
|
""" |
|
44
|
|
|
all_tags, template_files = all_model_templates() |
|
45
|
|
|
new_model_file = None |
|
46
|
|
|
|
|
47
|
|
|
for file, tag in zip(template_files, all_tags, strict=True): |
|
48
|
|
|
if spec_model_tag == tag: |
|
49
|
|
|
new_model_file = file |
|
50
|
|
|
return new_model_file |
|
51
|
|
|
|
|
52
|
|
|
|
|
53
|
|
|
def check_gammapy_model(gammapy_model): |
|
54
|
|
|
""" |
|
55
|
|
|
For a given object type, try to read it as a Gammapy Models object. |
|
56
|
|
|
""" |
|
57
|
|
|
if isinstance(gammapy_model, Models | SkyModel): |
|
58
|
|
|
models_gpy = Models(gammapy_model) |
|
59
|
|
|
else: |
|
60
|
|
|
try: |
|
61
|
|
|
models_gpy = Models.read(gammapy_model) |
|
62
|
|
|
except KeyError: |
|
63
|
|
|
raise TypeError("%s File cannot be read by Gammapy Models", gammapy_model) from KeyError |
|
64
|
|
|
|
|
65
|
|
|
return models_gpy |
|
66
|
|
|
|
|
67
|
|
|
|
|
68
|
|
|
def recursive_merge_lists(final_config_key, extra_config_key, value): |
|
69
|
|
|
""" |
|
70
|
|
|
Recursively merge from lists of dicts. Distinct function as an auxiliary for |
|
71
|
|
|
the recursive_merge_dicts function. |
|
72
|
|
|
""" |
|
73
|
|
|
new_config = [] |
|
74
|
|
|
|
|
75
|
|
|
for key_, value_ in zip(final_config_key, value, strict=False): |
|
76
|
|
|
key_ = recursive_merge_dicts(key_ or {}, value_) |
|
77
|
|
|
new_config.append(key_) |
|
78
|
|
|
|
|
79
|
|
|
# For example moving from a smaller list of model parameters to a |
|
80
|
|
|
# longer list. |
|
81
|
|
|
if len(final_config_key) < len(extra_config_key): |
|
82
|
|
|
for value_ in value[len(final_config_key) :]: |
|
83
|
|
|
new_config.append(value_) |
|
84
|
|
|
return new_config |
|
85
|
|
|
|
|
86
|
|
|
|
|
87
|
|
|
def recursive_merge_dicts(base_config, extra_config): |
|
88
|
|
|
""" |
|
89
|
|
|
Recursively merge two dictionaries. |
|
90
|
|
|
Entries in extra_config override entries in base_config. The built-in |
|
91
|
|
|
update function cannot be used for hierarchical dicts. |
|
92
|
|
|
|
|
93
|
|
|
Also for the case when there is a list of dicts involved, one has to be |
|
94
|
|
|
more careful. The extra_config may have longer list of dicts as compared |
|
95
|
|
|
with the base_config, in which case, the extra items are simply added to |
|
96
|
|
|
the merged final list. |
|
97
|
|
|
|
|
98
|
|
|
Combined here are 2 options from SO. |
|
99
|
|
|
|
|
100
|
|
|
See: |
|
101
|
|
|
http://stackoverflow.com/questions/3232943/update-value-of-a-nested-dictionary-of-varying-depth/3233356#3233356 |
|
102
|
|
|
and also |
|
103
|
|
|
https://stackoverflow.com/questions/3232943/update-value-of-a-nested-dictionary-of-varying-depth/18394648#18394648 |
|
104
|
|
|
|
|
105
|
|
|
Parameters |
|
106
|
|
|
---------- |
|
107
|
|
|
base_config : dict |
|
108
|
|
|
dictionary to be merged |
|
109
|
|
|
extra_config : dict |
|
110
|
|
|
dictionary to be merged |
|
111
|
|
|
Returns |
|
112
|
|
|
------- |
|
113
|
|
|
final_config : dict |
|
114
|
|
|
merged dict |
|
115
|
|
|
""" |
|
116
|
|
|
final_config = base_config.copy() |
|
117
|
|
|
|
|
118
|
|
|
for key, value in extra_config.items(): |
|
119
|
|
|
if key in final_config and isinstance(final_config[key], list): |
|
120
|
|
|
final_config[key] = recursive_merge_lists(final_config[key], extra_config[key], value) |
|
121
|
|
|
elif key in final_config and isinstance(final_config[key], dict): |
|
122
|
|
|
final_config[key] = recursive_merge_dicts(final_config.get(key) or {}, value) |
|
123
|
|
|
else: |
|
124
|
|
|
final_config[key] = value |
|
125
|
|
|
|
|
126
|
|
|
return final_config |
|
127
|
|
|
|
|
128
|
|
|
|
|
129
|
|
|
def deep_update(d, u): |
|
130
|
|
|
""" |
|
131
|
|
|
Recursively update a nested dictionary. |
|
132
|
|
|
|
|
133
|
|
|
Just like in Gammapy, taken from: https://stackoverflow.com/a/3233356/19802442 |
|
134
|
|
|
""" |
|
135
|
|
|
for k, v in u.items(): |
|
136
|
|
|
if isinstance(v, Mapping): |
|
137
|
|
|
d[k] = deep_update(d.get(k, {}), v) |
|
138
|
|
|
else: |
|
139
|
|
|
d[k] = v |
|
140
|
|
|
return d |
|
141
|
|
|
|
|
142
|
|
|
|
|
143
|
|
|
def compound_model_dict_converstion(dict): |
|
144
|
|
|
""" |
|
145
|
|
|
Given a Gammapy CompoundSpectralModel as a dict object, convert it into |
|
146
|
|
|
an Asgardpy form. |
|
147
|
|
|
""" |
|
148
|
|
|
ebl_abs = dict["model2"] |
|
149
|
|
|
ebl_abs["alpha_norm"] = ebl_abs["parameters"][0]["value"] |
|
150
|
|
|
ebl_abs["redshift"] = ebl_abs["parameters"][1]["value"] |
|
151
|
|
|
ebl_abs.pop("parameters", None) |
|
152
|
|
|
|
|
153
|
|
|
dict["type"] = dict["model1"]["type"] |
|
154
|
|
|
dict["parameters"] = dict["model1"]["parameters"] |
|
155
|
|
|
dict["ebl_abs"] = ebl_abs |
|
156
|
|
|
|
|
157
|
|
|
dict.pop("model1", None) |
|
158
|
|
|
dict.pop("model2", None) |
|
159
|
|
|
dict.pop("operator", None) |
|
160
|
|
|
|
|
161
|
|
|
return dict |
|
162
|
|
|
|
chaimain /
asgardpy
