-
Notifications
You must be signed in to change notification settings - Fork 0
/
birth_seasonality.py
executable file
·219 lines (204 loc) · 9.29 KB
/
birth_seasonality.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
#!/usr/bin/python3
'''Analyze and plot the results of varying the level of birth
seasonality. This requires the file `birth_seasonality.h5`, which is
built by `birth_seasonality_run.py`.'''
from matplotlib import colors, pyplot, ticker
import numpy
import seaborn
import statsmodels.nonparametric.api
import plot_common
import stats
def load():
filename = 'birth_seasonality.h5'
return plot_common.get_extinction_time(filename)
def plot_median(df, CI=0.5):
row = dict(acute=0, chronic=1)
column = dict(enumerate(range(3), 1))
levels = [CI / 2, 1 - CI / 2]
with seaborn.axes_style('darkgrid'):
fig, axes = pyplot.subplots(len(row), len(column),
sharex='col', sharey='row')
for ((model, SAT), group) in df.groupby(['model', 'SAT']):
i, j = row[model], column[SAT]
ax = axes[i, j]
by = 'birth_seasonal_coefficient_of_variation'
times = group.groupby(by).time
median = times.median()
ax.plot(median.index, median,
color=plot_common.SAT_colors[SAT])
CI_ = times.quantile(levels).unstack()
ax.fill_between(CI_.index, CI_[levels[0]], CI_[levels[1]],
color=plot_common.SAT_colors[SAT],
alpha=0.5)
if ax.is_first_row():
ax.set_title(f'SAT{SAT}', fontsize='medium')
else:
ax.set_title('')
if ax.is_last_row():
if j == 1:
ax.set_xlabel('Birth seasonal\ncoefficient of variation')
else:
ax.set_xlabel('')
if ax.is_first_col():
ax.set_ylabel(
f'{model.capitalize()} model\n\nextinction time (y)')
for ax in axes[:, -1]:
ax.set_ylim(bottom=0)
fig.align_labels()
fig.tight_layout()
def plot_survival(df):
row = dict(enumerate(range(3), 1))
column = dict(acute=0, chronic=1)
fig, axes = pyplot.subplots(3, 2, sharex='col', sharey='row')
for ((model, SAT), group) in df.groupby(['model', 'SAT']):
i, j = row[SAT], column[model]
ax = axes[i, j]
for (b, g) in group.groupby('birth_seasonal_coefficient_of_variation'):
survival = stats.get_survival(g, 'time', 'observed')
ax.step(survival.index, survival,
where='post',
label=f'birth_seasonal_coefficient_of_variation {b}')
def plot_kde(df):
row = dict(enumerate(range(3), 1))
column = dict(acute=0, chronic=1)
with seaborn.axes_style('darkgrid'):
fig, axes = pyplot.subplots(3, 2, sharex='col')
for ((model, SAT), group) in df.groupby(['model', 'SAT']):
i, j = row[SAT], column[model]
ax = axes[i, j]
for (b, g) in group.groupby(
'birth_seasonal_coefficient_of_variation'):
ser = g.time[g.observed]
proportion_observed = len(ser) / len(g)
if proportion_observed > 0:
kde = statsmodels.nonparametric.api.KDEUnivariate(ser)
kde.fit(cut=0)
x = kde.support
y = proportion_observed * kde.density
else:
x, y = [], []
label = b if i == j == 0 else ''
ax.plot(x, y, label=label, alpha=0.7)
ax.yaxis.set_major_locator(ticker.NullLocator())
if ax.is_first_row():
ax.set_title(f'{model.capitalize()} model',
fontdict=dict(fontsize='medium'))
if ax.is_last_row():
ax.set_xlim(left=0)
ax.set_xlabel('extinction time (y)')
if ax.is_first_col():
ylabel = '\ndensity' if i == 1 else ''
ax.set_ylabel(f'SAT{SAT}{ylabel}')
leg = fig.legend(loc='center left', bbox_to_anchor=(0.8, 0.5),
title='Birth seasonal\ncoefficient of\nvariation')
fig.tight_layout(rect=(0, 0, 0.82, 1))
def plot_kde_2d(df):
persistence_time_max = dict(acute=0.5, chronic=10)
bscovs = (df.index
.get_level_values('birth_seasonal_coefficient_of_variation')
.unique()
.sort_values())
bscov_baseline = bscovs[len(bscovs) // 2]
width = 390 / 72.27
height = 0.8 * width
rc = plot_common.rc.copy()
rc['figure.figsize'] = (width, height)
rc['xtick.labelsize'] = rc['ytick.labelsize'] = 7
rc['axes.labelsize'] = 8
rc['axes.titlesize'] = 9
nrows = 2 + 1
ncols = 3
height_ratios = (1, 1, 0.5)
w_pad = 8 / 72
with pyplot.rc_context(rc):
fig = pyplot.figure(constrained_layout=True)
fig.set_constrained_layout_pads(w_pad=w_pad)
gs = fig.add_gridspec(nrows, ncols,
height_ratios=height_ratios)
axes = numpy.empty((nrows, ncols), dtype=object)
axes[0, 0] = None # Make sharex & sharey work for axes[0, 0].
for row in range(nrows):
for col in range(ncols):
# Columns share the x scale.
sharex = axes[0, col]
# Rows share the y scale.
sharey = axes[row, 0]
axes[row, col] = fig.add_subplot(gs[row, col],
sharex=sharex,
sharey=sharey)
for (i, (model, group_model)) in enumerate(df.groupby('model')):
persistence_time = numpy.linspace(0, persistence_time_max[model],
301)
for (j, (SAT, group_SAT)) in enumerate(group_model.groupby('SAT')):
ax = axes[i, j]
density = numpy.zeros((len(persistence_time),
len(bscovs)))
proportion_observed = numpy.zeros_like(bscovs, dtype=float)
grouper = group_SAT.groupby(
'birth_seasonal_coefficient_of_variation')
for (k, (b, g)) in enumerate(grouper):
ser = g.time[g.observed]
nruns = len(g)
proportion_observed[k] = len(ser) / nruns
density[:, k] = plot_common.get_density(ser,
persistence_time)
cmap = plot_common.get_cmap_SAT(SAT)
# Use raw `density` for color,
# but plot `density * proportion_observed`.
norm = colors.Normalize(vmin=0, vmax=numpy.max(density))
ax.imshow(density * proportion_observed,
cmap=cmap, norm=norm, interpolation='bilinear',
extent=(min(bscovs), max(bscovs),
min(persistence_time), max(persistence_time)),
aspect='auto', origin='lower', clip_on=False)
# ax shares the xaxis with ax_po.
ax.xaxis.set_tick_params(which='both',
labelbottom=False, labeltop=False)
ax.xaxis.offsetText.set_visible(False)
ax.xaxis.set_minor_locator(ticker.AutoMinorLocator(2))
ax.yaxis.set_major_locator(
ticker.MultipleLocator(max(persistence_time) / 5))
ax.yaxis.set_minor_locator(ticker.AutoMinorLocator(2))
if ax.is_first_col():
ax.set_ylabel('Extinction time (y)')
if ax.is_first_row():
ax.set_title(f'SAT{SAT}')
if model == 'chronic':
ax_po = axes[-1, j]
ax_po.plot(bscovs, 1 - proportion_observed,
color=plot_common.SAT_colors[SAT],
clip_on=False, zorder=3)
ax_po.set_xlabel(
'Birth seasonal\ncoefficient of variation')
ax_po.xaxis.set_minor_locator(ticker.AutoMinorLocator(2))
ax_po.yaxis.set_major_formatter(
ticker.PercentFormatter(xmax=1))
ax_po.yaxis.set_minor_locator(ticker.AutoMinorLocator(2))
if ax_po.is_first_col():
ax_po.set_ylabel('Persisting 10 y')
for ax in fig.axes:
ax.axvline(bscov_baseline,
color='black', linestyle='dotted', alpha=0.7)
ax.autoscale(tight=True)
if not ax.is_first_col():
ax.yaxis.set_tick_params(which='both',
labelleft=False, labelright=False)
ax.yaxis.offsetText.set_visible(False)
for sp in ('top', 'right'):
ax.spines[sp].set_visible(False)
fig.align_ylabels()
label_x = 0
label_kws = dict(fontsize=8,
rotation=90,
horizontalalignment='left',
verticalalignment='center')
fig.text(label_x, 0.79, 'Acute model', **label_kws)
fig.text(label_x, 0.31, 'Carrier model', **label_kws)
fig.savefig('birth_seasonality.pdf')
if __name__ == '__main__':
df = load()
# plot_median(df)
# plot_survival(df)
# plot_kde(df)
plot_kde_2d(df)
pyplot.show()