/
diagnostics.py
223 lines (172 loc) · 7.48 KB
/
diagnostics.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
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
import numpy as np
import matplotlib.pyplot as plt
from cycler import cycler
import figures_tools as ftools
from itertools import product as iproduct
from astropy.io import fits
class Diagnostic(object):
def __init__(self, results, metadata, drpall):
self.results = results
self.metadata = metadata
self.drpall = drpall
def make_diag_figure(self, xnames, ynames):
nobj = len(xnames)
# initialize subplot size
gs, fig = ftools.gen_gridspec_fig(
nobj, add_row=False, border=(0.6, 0.6, 0.2, 0.4),
space=(0.6, 0.35))
# set up subplot interactions
gs_geo = gs.get_geometry()
fgrid_r, fgrid_c = tuple(list(range(n)) for n in gs_geo)
gs_iter = iproduct(fgrid_r, fgrid_c)
# set up kwargs for matplotlib errorbar
# prefer to change color, then marker
colors_ = ['C{}'.format(cn) for cn in range(10)]
markers_ = ['o', 'v', 's', 'P', 'X', 'D', 'H']
# iterate through rows & columns (rows change fastest)
# which correspond to different quantities
for (i, (ri, ci)) in enumerate(gs_iter):
if i >= len(xnames):
continue
# choose axis
ax = fig.add_subplot(gs[ri, ci])
kwarg_cycler = cycler(marker=markers_) * \
cycler(c=colors_)
xqty = xnames[i]
yqty = ynames[i]
# now iterate through results hdulists
for (j, (result, kwargs)) in enumerate(
zip(self.results, kwarg_cycler)):
kwargs['label'] = result[0].header['PLATEIFU']
ax = self._add_log_offset_plot(
j, xqty=xqty, yqty=yqty, ax=ax, **kwargs)
ax.tick_params(labelsize=5)
if i == 0:
handles_, labels_ = ax.get_legend_handles_labels()
plt.figlegend(
handles=handles_, labels=labels_,
loc='upper right', prop={'size': 4.})
fig.suptitle('PCA fitting diagnostics', size=8.)
return fig
def get_log_yerrs(self, i, qty):
'''
get the log-ratio between the measured & ground-truth
'''
# retrieve appropriate entry from results attribute
P50, l_unc, u_unc = tuple(
map(np.squeeze, np.split(self.results[i][qty].data,
3, axis=0)))
logscale = self.results[i][qty].header['LOGSCALE']
# if the quantity's on a log scale already, return as-is
if (logscale) or ('log' in qty):
return P50, l_unc, u_unc
# otherwise, log everything, subtract, and return
logP50 = np.log10(P50)
log_l_unc = logP50 - np.log10(P50 - l_unc)
log_u_unc = np.log10(P50 + u_unc) - logP50
return logP50, log_l_unc, log_u_unc
def get_meas_vs_truth(self, i, qty):
logP50, log_l_unc, log_u_unc = self.get_log_yerrs(i, qty)
truth = self.results[i][qty].header['TRUTH']
# compare measured range to truth
P50_ratio = np.log10(10.**logP50 / 10.**truth)
l_unc_ratio = np.log10(10.**(logP50 - log_l_unc) / 10.**truth)
u_unc_ratio = np.log10(10.**(logP50 + log_u_unc) / 10.**truth)
return P50_ratio, l_unc_ratio, u_unc_ratio
def _munge_xqty(self, xqty, i, arr_len):
# get x qty
if 'drpall' in xqty:
# fetch value from drpall
_, drpall_col = xqty.split('-')
plateifu = self.results[i][0].header['PLATEIFU']
x = self.drpall.loc[plateifu][drpall_col] * \
np.ones(arr_len)
xlabel = r'${{\rm {}}}$'.format(
drpall_col.replace('_', '\_'))
elif 'hdr' in xqty:
# fetch value from header
_, hdr_i, hdr_key = xqty.split('-')
x = self.results[i][int(hdr_i)].header[hdr_key] * \
np.ones(arr_len)
xlabel = r'${{\rm {}}}$'.format(
hdr_key.replace('_', '\_'))
else:
x_hdu = self.results[i][xqty]
if xqty in self.metadata.colnames:
xstr = self.metadata[xqty].meta.get('TeX', xqty).strip('$')
else:
xstr = xqty
# check if "truth" value is available: if not, use array
xtruth = x_hdu.header.get('TRUTH', None)
if xtruth is not None:
x = np.ones(arr_len) * xtruth
xlabel = r'${{ \rm {0} }}$'.format(xstr)
else:
# if value used is not true value, denote with tilde
xlabel = r'${{ \rm \tilde{{{0}}} }}$'.format(xstr)
xdata = x_hdu.data
if len(xdata.shape) == 3:
x = xdata[0, ...]
else:
x = xdata
return x, xlabel
def _munge_yqty(self, yqty, i):
P50_ratio, l_unc_ratio, u_unc_ratio = self.get_meas_vs_truth(
i, yqty)
if 'log' in yqty:
ylabel = r'$\tilde{{{0}}} - {0}$'.format(
self.metadata[yqty].meta.get('TeX', yqty).strip('$'))
else:
ylabel = r'$\log_{{10}}{{\frac{{\tilde{{{0}}}}}{{{0}}}}}$'.format(
self.metadata[yqty].meta.get('TeX', yqty).strip('$'))
return P50_ratio, l_unc_ratio, u_unc_ratio, ylabel
def _add_log_offset_plot(self, i, xqty, yqty, ax, **kwargs):
'''
plot ratio of best-fit vs some other qty
'''
P50_ratio, l_unc_ratio, u_unc_ratio, ylabel = self._munge_yqty(yqty, i)
x, xlabel = self._munge_xqty(xqty, i, arr_len=P50_ratio.shape)
mask = self.results[i]['MASK'].data.astype(bool)
badfit = ~self.results[i]['SUCCESS'].data.astype(bool)
badPDF = (self.results[i]['GOODFRAC'].data[0] * \
self.results[i]['GOODFRAC'].header['NMODELS']) < 10
P50_ratio = np.ma.array(
P50_ratio, mask=np.logical_or.reduce((mask, badfit, badPDF)))
ax.scatter(x, P50_ratio, s=1., edgecolor='None',
alpha=0.6, vmin=0., vmax=1., **kwargs)
ax.axhline(0., c='k', linewidth=0.25)
ax.set_ylabel(ylabel, size='x-small')
ax.set_xlabel(xlabel, size=5)
ax.set_ylim([-1.25, 1.25])
if xqty == 'SNRMED':
ax.set_xscale('log')
ax.set_xlim([.1, 1.1 * ax.get_xlim()[1]])
return ax
if __name__ == '__main__':
from find_pcs import *
from glob import glob
cosmo = WMAP9
warn_behav = 'ignore'
dered_method = 'supersample_vec'
dered_kwargs = {'nper': 5}
CSPs_dir = '/usr/data/minhas2/zpace/CSPs/CSPs_CKC14_MaNGA_20180501-1/'
mpl_v = 'MPL-6'
drpall = m.load_drpall(mpl_v, index='plateifu')
drpall = drpall[drpall['nsa_z'] != -9999]
lsf = ut.MaNGA_LSF.from_drpall(drpall=drpall, n=2)
pca_kwargs = {'lllim': 3700. * u.AA, 'lulim': 8800. * u.AA,
'lsf': lsf, 'z0_': .04}
pca, K_obs = setup_pca(
fname='pca.pkl', base_dir=CSPs_dir, base_fname='CSPs',
redo=False, pkl=True, q=10, fre_target=.005, nfiles=50,
pca_kwargs=pca_kwargs)
# find appropriate files
hdulists = list(map(
fits.open, glob(os.path.join(CSPs_dir, 'fakedata/results/*/*_res.fits'))))
# make diag plots
plt.close('all')
diag = Diagnostic(results=hdulists, metadata=pca.metadata, drpall=drpall)
diag_fig = diag.make_diag_figure(
xnames=['SNRMED', 'tau_V', 'tau_V mu', 'MLV'],
ynames=['MLV', 'MLV', 'MLV', 'MLV'])
diag_fig.savefig('diagplot.png');