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
|
from typing import Optional
import numpy as np
TRIM_COLOR = "orange"
BURN_COLOR = "red"
def convergence_plot(
convergence: np.ndarray,
dof: float,
cutoff: float = 0.25,
trim_index: Optional[int] = None,
burn_index: Optional[int] = None,
max_points: Optional[int] = None,
) -> dict:
"""
Generate a convergence plot from the convergence data.
Parameters:
- convergence: A 2D numpy array where the first column contains the best values
and the remaining columns contain the population values.
- dof: Degrees of freedom for normalization.
Returns:
- A dictionary suitable for JSON serialization containing the plot data and layout.
"""
# Normalize the population data
normalized_pop = 2 * convergence / dof
best, pop = normalized_pop[:, 0], normalized_pop[:, 1:]
ni, npop = pop.shape
x = np.arange(1, ni + 1)
tail = int(cutoff * ni)
nout = ni - tail
step = nout // max_points if (max_points is not None and nout > max_points) else 1
x_out = x[tail::step]
best_out = best[tail::step]
pop_out = pop[tail::step, :]
# print(f"Convergence plot: {ni} steps, {npop} population, {nout} output, step={step}")
traces = []
layout = {}
hovertemplate = "(%{{x}}, %{{y}})<br>{label}<extra></extra>"
if npop == 5:
if (trim_index is not None and trim_index > tail) or (burn_index is not None and burn_index > tail):
layout["shapes"] = []
layout["annotations"] = []
# fig['data'].append(dict(type="scattergl", x=x, y=pop[tail:,4].tolist(), name="95%", mode="lines", line=dict(color="lightblue", width=1), showlegend=True, hovertemplate=hovertemplate.format(label="95%")))
if trim_index is not None and trim_index > tail:
layout["shapes"].append(
{
"line": {"color": TRIM_COLOR, "dash": "dash", "width": 2},
"type": "line",
"x0": trim_index,
"x1": trim_index,
"xref": "x",
"y0": 0,
"y1": 0.95,
"yref": "y domain",
},
)
layout["annotations"].append(
{
"text": "←trim<br>use→",
"x": trim_index,
"y": 0.95,
"xref": "x",
"xanchor": "left",
"yref": "y domain",
"showarrow": False,
"font": {"color": TRIM_COLOR, "size": 12},
},
)
if burn_index is not None and burn_index > tail:
layout["shapes"].append(
{
"line": {"color": BURN_COLOR, "dash": "dash", "width": 2},
"type": "line",
"x0": burn_index,
"x1": burn_index,
"xref": "x",
"y0": 0,
"y1": 1,
"yref": "y domain",
}
)
layout["annotations"].append(
{
"text": "←burn<br>samples→",
"x": burn_index,
"y": 1.0,
"xref": "x",
"xanchor": "left",
"yref": "y domain",
"showarrow": False,
"font": {"color": BURN_COLOR, "size": 12},
}
)
traces.append(
dict(
type="scattergl",
x=x_out,
y=pop_out[:, 3],
mode="lines",
line=dict(color="lightgreen", width=0),
showlegend=False,
hovertemplate=hovertemplate.format(label="80%"),
)
)
traces.append(
dict(
type="scattergl",
x=x_out,
y=pop_out[:, 1],
name="20% to 80% range",
fill="tonexty",
mode="lines",
line=dict(color="lightgreen", width=0),
hovertemplate=hovertemplate.format(label="20%"),
)
)
traces.append(
dict(
type="scattergl",
x=x_out,
y=pop_out[:, 2],
name="population median",
mode="lines",
line=dict(color="green"),
opacity=0.5,
)
)
traces.append(
dict(
type="scattergl",
x=x_out,
y=pop_out[:, 0],
name="population best",
mode="lines",
line=dict(color="green", dash="dot"),
)
)
traces.append(
dict(
type="scattergl",
x=x_out,
y=best_out,
name="best",
line=dict(color="red", width=1),
mode="lines",
)
)
layout.update(
dict(
template="simple_white",
legend=dict(x=1, y=1, xanchor="right", yanchor="top"),
)
)
layout.update(dict(title=dict(text="Convergence", xanchor="center", x=0.5)))
layout.update(dict(uirevision=1))
layout.update(
dict(
xaxis=dict(
title={"text": "Steps"},
showline=True,
showgrid=False,
zeroline=False,
)
)
)
layout.update(
dict(
yaxis=dict(
title={"text": "Normalized <i>\u03c7</i><sup>2</sup>"}, showline=True, showgrid=False, zeroline=False
)
)
)
return {
"data": traces,
"layout": layout,
}
|
