Repository navigation
Expand file tree
/
Copy pathprmspy.py
More file actions
225 lines (174 loc) · 6.87 KB
/
Copy pathprmspy.py
File metadata and controls
225 lines (174 loc) · 6.87 KB
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
import click
import io
import itertools
import json
import matplotlib.pyplot as plt
import numpy as np
import os
import re
from datetime import datetime
from matplotlib.backends.backend_pdf import PdfPages
from ..scenario import ScenarioSeries
from ..util import load_data, load_statvar, nash_sutcliffe
@click.group()
def prmspy():
"access PRMS-Python functionality from the command line"
click.echo('\n*** Welcome to PRMS-Python! ***\n')
@prmspy.command()
@click.argument('base_data_dir', nargs=1)
@click.option('--params', '-p', nargs=1, type=str, multiple=True,
help='list of parameters to edit')
@click.option('--scale_vals', '-s', nargs=1, type=str, multiple=True,
help='list of scaling values; must be divisible by len(params), '
'e.g. "[0.8, 0.9, 1.0, 1.1, 1.2]"'
)
@click.option('--output-dir', '-o', nargs=1, type=str,
help='directory where scenario data should be written')
@click.option('--title', '-t', nargs=1, type=str,
help='title for this parameter scaling experiment')
@click.option('--description', '-d', nargs=1, type=str,
help='description for this parameter scaling experiment')
@click.option('--run-prms', is_flag=True,
help='run PRMS after creating scenario input data')
@click.option('--prms-exec', '-e', default='prms',
help='PRMS executable to be used, eg prmsV4')
@click.option('--nproc', '-n', default=None, type=int,
help='number of processors to use')
@click.option('--analyze-output', is_flag=True,
help='create analysis of model output in the form of a '
'"Nash-Sutcliffe Matrix"')
def param_scale_sim(base_data_dir,
params,
scale_vals,
output_dir,
title,
description,
run_prms,
prms_exec,
nproc,
analyze_output):
'Provide params and scaling values; run PRMS scenarios'
# scale_vals must be divisible by the number of params
# valid_input = len(scale_vals) % len(params) == 0
print(scale_vals)
scale_vals = [eval(el) for el in scale_vals]
# if not valid_input:
# raise IOError('The length of scale_vals list is not divisible '
# 'by the length of the parameter list')
if not output_dir:
output_dir =\
re.sub(r'\/|\\\\', '-', base_data_dir) + \
'-modified-params-' + datetime.now().isoformat()
os.mkdir(output_dir)
s = ScenarioSeries(
base_data_dir, output_dir, title=title, description=description
)
def _scale_fun(val):
def scale(x):
return x * val
return scale
# assign scaling values to parameter names
# dimensionality correctness guaranteed from valid_input check
pval_len = len(scale_vals)/len(params)
def _slice(idx):
return slice(pval_len*idx, pval_len*(idx+1))
param_val_dict = {
param: zip(itertools.repeat(param), scale_vals[idx])
for idx, param in enumerate(params)
}
# param_val_dict = {
# param: zip(itertools.repeat(param, pval_len), scale_vals[_slice(idx)])
# for idx, param in enumerate(params)
# }
# for every combination we need to iterate over every value of
# each param we have
combinations = itertools.product(*param_val_dict.values())
# now build the scenario_list with re-callable titles
scenario_list = []
for combo in combinations:
scenario_def = {
'title': _build_title(combo)
}
scenario_def.update(
{
param: _scale_fun(val)
for param, val in combo
}
)
scenario_list.append(scenario_def)
s.build(scenario_list)
if run_prms:
s.run(prms_exec=prms_exec, nproc=nproc)
if analyze_output:
print('\n** Sorry, analyze_output has not yet been implemented! **\n')
@prmspy.command()
@click.argument('data_dir', nargs=1)
@click.argument('output_pdf_path', nargs=1)
def nash_sutcliffe_matrix(data_dir, output_pdf_path):
'Save a PDF of the Nash-Sutcliffe created from <data_dir> to <output_pdf_path>'
observed = load_data(
os.path.join(data_dir, 'base_inputs', 'data')
).runoff_1
series_metadata = json.loads(
io.open(os.path.join(data_dir, 'series_metadata.json')).read()
)
modeled_flows = {
title: load_statvar(
os.path.join(data_dir, uu, 'outputs', 'statvar.dat')
).basin_cfs_1
for uu, title in series_metadata['uuid_title_map'].items()
}
titles = list(modeled_flows.keys())
params = [pair.split(':')[0] for pair in titles[0].split('|')]
if len(params) > 2:
print('This command is only supported for two covarying parameters!')
exit(1)
p1_vals = sorted({
float(pair[0].split(':')[1])
for pair in map(lambda t: t.split('|'), titles)
})
p2_vals = sorted({
float(pair[1].split(':')[1])
for pair in map(lambda t: t.split('|'), titles)
})
print(p1_vals)
nash_sutcliffe_mat = np.zeros((len(p1_vals), len(p2_vals)))
for p1_idx, p1_val in enumerate(p1_vals):
for p2_idx, p2_val in enumerate(p2_vals):
coord = (p1_idx, p2_idx)
title = '{p1}:{p1val:.3f}|{p2}:{p2val:.3f}'.format(
p1=params[0], p2=params[1], p1val=p1_val, p2val=p2_val
)
modeled = modeled_flows[title]
nash_sutcliffe_mat[coord] = nash_sutcliffe(observed, modeled)
with PdfPages(output_pdf_path) as pdf:
fig, ax = plt.subplots()
cax = ax.matshow(nash_sutcliffe_mat, cmap='viridis')
ytix = p1_vals
xtix = p2_vals
plt.yticks(range(len(ytix)), ytix)
plt.xticks(range(len(xtix)), xtix)
ax.xaxis.set_ticks_position('bottom')
plt.ylabel(params[0].replace('"', ''))
plt.xlabel(params[1].replace('"', ''))
for i, j in itertools.product(range(len(ytix)), range(len(xtix))):
plt.text(j, i, '%.2f' % nash_sutcliffe_mat[i, j],
horizontalalignment='center',
color='w'
if nash_sutcliffe_mat[i, j] <
np.mean(nash_sutcliffe_mat.flatten())
else 'k')
plt.title('Nash-Sutcliffe Matrix')
plt.grid(b=False)
fig.colorbar(cax)
pdf.savefig()
plt.close()
def _build_title(combo):
'''
Given a list of tuples with parameter/scale_val pairs, build the title
for the parameterization. For example,
>>> combo = [('snow_adj', 0.8), ('rad_trncf', 0.9), ('jh_coef', 1.1)]
>>> assert _build_title(combo) ==\
'"snow_adj":0.800|"rad_trncf":0.900|"jh_coef":1.100'
'''
return '|'.join('"{0}":{1:.3f}'.format(param, val) for param, val in combo)