-
Notifications
You must be signed in to change notification settings - Fork 9
/
penalty_gifs.R
127 lines (113 loc) 路 3.96 KB
/
penalty_gifs.R
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
## Produces an animation showing how penalties result in smoother splines
## packages required
library('ggplot2')
library('ggborderline')
library('dplyr')
library('mgcv')
library('gganimate')
library('magick')
library('rsvg')
library('transformr')
library('viridis')
theme_set(theme_classic(base_size = 18, base_family = 'serif'))
## generate some data
set.seed(7)
N <- 400
y <- mvgam::sim_mvgam(T = N, prop_missing = 0,
prop_train = 1,
family = gaussian(),
n_series = 1,
trend_model = 'GP',
trend_rel = 0.95)$data_train$y
x <- 1:N
data <- data.frame(x = x,
y = y)
plot(y ~ x, data = data)
bestmod <- gam(y ~ s(x, k = 40), data = data)
plot(bestmod)
best_sp <- unname(bestmod$sp)
mod <- gam(y ~ s(x, k = 100), data = data,
sp = 0.9)
plot(mod)
# fit gams with different penalties
penalties <- c(1000, 0.00000001,
50, 0.0000001,
20, 0.00001, 10,
0.00000005, 3,
0.000005, 0.15, 2,
0.000009, 0.5,
0.0000098, 0.3,
0.000001, 0.009,
0.00001, 0.008,
0.0001, 0.006,
0.0005, 0.005,
0.002, 0.001,
0.0007, 0.005,
0.00099, 0.0015,
0.000995, 0.001)
line_dat <- do.call(rbind, lapply(seq_along(penalties), function(sp){
mod <- gam(y ~ s(x, k = 100), data = data,
sp = penalties[sp])
modpreds <- predict(mod,
newdata = data.frame(x = seq(1, N, length.out = 400)),
type = 'response')
data.frame(pred = modpreds,
x_line = seq(1, N, length.out = 400),
penalty = penalties[sp],
draw = sp)
}))
p1 <- ggplot(data, aes(x = x, y = y)) +
geom_point(alpha = 0.4) +
geom_borderline(data = line_dat, aes(x = x_line, y = pred),
linewidth = 1.5, bordercolour = "white") +
theme(legend.position = 'none',
axis.line = element_line(size = 1),
axis.ticks = element_line(colour = "black", size = 1)) +
ylab('f(Time)') + xlab('Time')
p1
gif1 <- animate(
p1 + transition_states(draw,
transition_length = 2,
state_length = 2) +
ease_aes('cubic-in-out'),
nframes = 160, device = 'svg', renderer = magick_renderer(),
width = 8, height = 5)
line_dat$penalty <- log(line_dat$penalty)
line_dat$dens <- dnorm(line_dat$penalty, -6.9, 6)
line_dat %>%
dplyr::select(penalty, draw, dens) %>%
dplyr::distinct() -> plot_dat
stats_df <- data.frame(prior = rnorm(100000, -6.9, 6))
p2 <- ggplot(plot_dat, aes(x = penalty)) +
geom_segment(aes(x = penalty,
xend = penalty,
y = 0, yend = dens),
size = 1.5, col = 'white') +
geom_segment(aes(x = penalty,
xend = penalty,
y = 0, yend = dens),
size = 1.3, col = 'darkred') +
geom_point(aes(x = penalty, y = dens),
shape = 21, size = 2,
col = 'white', fill = 'darkred') +
theme(legend.position = 'none',
axis.line = element_line(size = 1),
axis.ticks = element_line(colour = "black", size = 1)) +
xlab(expression(rho)) + ylab('log(likelihood)')
p2
gif2 <- animate(
p2 + transition_states(draw,
transition_length = 2,
state_length = 2) +
shadow_mark(exclude_layer = c(1,2)) +
ease_aes('cubic-in-out'),
nframes = 160, device = 'svg', renderer = magick_renderer(),
width = 8, height = 5)
new_gif <- image_append(c(gif1[1], gif2[1]))
for(i in 2:160){
combined <- image_append(c(gif1[i], gif2[i]))
new_gif <- c(new_gif, combined)
}
new_gif
anim_save(filename = 'day1/resources/penalty_spline.gif',
animation = new_gif)