forked from google/bi-tempered-loss
-
Notifications
You must be signed in to change notification settings - Fork 0
/
loss_plot.js
139 lines (115 loc) · 4.24 KB
/
loss_plot.js
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
/* Copyright 2019 Google Inc. All Rights Reserved.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
==============================================================================*/
numLossPlotActivations = 100
var dataset = []
function updateDataset() {
dataset = d3.range(numLossPlotActivations).map(function(d) {
x = -5.0 + (10.0 / numLossPlotActivations) * d
activation = tf.tensor([x], [1])
probability = tf.tensor([1.0], [1])
t1 = d3.select("#t1").property("value") / 100.0
t2 = d3.select("#t2").property("value") / 100.0
y = tf.mean(bitemperedBinaryLogisticLoss(activation, probability, t1, t2)).arraySync()
return {
x: x,
y: y,
};
});
}
var logisticLossDataset = d3.range(numLossPlotActivations).map(function(d) {
x = -5.0 + (10.0 / numLossPlotActivations) * d
activation = tf.tensor([x], [1])
probability = tf.tensor([1.0], [1])
y = tf.mean(bitemperedBinaryLogisticLoss(activation, probability, 1.0, 1.0)).arraySync()
return {
x: x,
y: y,
};
});
updateDataset()
function renderLossPlot() {
var width = 240;
var height = 160;
const lossPlotSvg = d3.select('#loss-plot').attr("height", height)
.attr("width", width);
const render = data => {
const title = 'Plot of bi-tempered logistic loss function';
lossPlotSvg.selectAll("*").remove();
const xValue = d => d.x;
const xAxisLabel = 'activations';
const yValue = d => d.y;
const yAxisLabel = 'loss';
const margin = {
top: 30,
right: 50,
bottom: 40,
left: 50
};
const innerWidth = width - margin.left - margin.right;
const innerHeight = height - margin.top - margin.bottom;
const xScale = d3.scaleLinear()
.domain([-5, 5])
.range([0, innerWidth])
.nice();
const yScale = d3.scaleLinear()
.domain([0, 6])
.range([innerHeight, 0])
.nice();
const g = lossPlotSvg.append('g')
.attr('transform', `translate(${margin.left},${margin.top})`);
const xAxis = d3.axisBottom(xScale)
.tickSize(-innerHeight)
.tickPadding(5);
const yAxis = d3.axisLeft(yScale)
.tickSize(-innerWidth)
.ticks(7)
.tickPadding(10);
const yAxisG = g.append('g').call(yAxis);
yAxisG.selectAll('.domain').remove();
yAxisG.append('text')
.attr('class', 'axis-label')
.attr('y', -25)
.attr('x', -innerHeight / 2)
.attr('fill', 'black')
.attr('transform', `rotate(-90)`)
.attr('text-anchor', 'middle')
.text(yAxisLabel);
const xAxisG = g.append('g').call(xAxis)
.attr('transform', `translate(0,${innerHeight})`);
xAxisG.select('.domain').remove();
xAxisG.append('text')
.attr('class', 'axis-label')
.attr('y', 25)
.attr('x', innerWidth / 2)
.attr('fill', 'black')
.text(xAxisLabel);
const lineGenerator = d3.line()
.x(d => xScale(xValue(d)))
.y(d => yScale(yValue(d)))
.curve(d3.curveBasis);
g.append('path')
.style("stroke", "#248ea9")
.attr('class', 'line-path')
.attr('d', lineGenerator(data));
g.append('path')
.attr('class', 'line-path')
.style("stroke", "#ff8b6a")
.style("stroke-dasharray", "4,4")
.attr('d', lineGenerator(logisticLossDataset));
g.append('text')
.attr('class', 'title')
.attr('y', -15)
.text(title);
};
render(dataset)
}
renderLossPlot();