-
Notifications
You must be signed in to change notification settings - Fork 0
/
make_plots.R
256 lines (237 loc) · 15.8 KB
/
make_plots.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
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
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
## Source this file to make exploratory plots. Need to load and prep data
## before sourcing
ggwidth <- 9
ggheight <- 7
library(ggplot2)
library(plyr)
### ------------------------------------------------------------
## First section is popbin effect. The 'scalars' object is a vector of
## parameters to save, see main.R for it's definition.
p <- popbins.binwidth.scalars[,c("species", "binwidth", 'F', paste0(scalars, "_RE"))] %>%
melt(id.vars=c('species','F', "binwidth")) %>%
ggplot(aes(x=binwidth, y=value, colour=variable)) + geom_line() +
facet_grid(species~F)+ xlab("Population bin width (cm)") +
ylab("Relative change from base case")
ggsave(paste0("plots/popbins_binwidth.png"),p, width=ggwidth, height=ggheight)
p <- popbins.minsize[,c("species", "minsize", 'F', paste0(scalars, "_RE"))] %>%
melt(id.vars=c('species','F', "minsize")) %>%
ggplot(aes(x=minsize, y=value, colour=variable)) + geom_line() +
facet_grid(species~F)+ xlab("Population min bin (cm)") +
ylab("Relative change from base case")
ggsave(paste0("plots/popbins_minsize.png"),p, width=ggwidth, height=ggheight)
p <- popbins.maxsize[,c("species", "max.Linf.ratio", 'F', paste0(scalars, "_RE"))] %>%
melt(id.vars=c('species','F', "max.Linf.ratio")) %>%
ggplot(aes(x=max.Linf.ratio, y=value, colour=variable)) + geom_line() +
facet_grid(species~F, scales='free_x')+
xlab("Population max bin:Linf ratio") +
ylab("Relative change from base case")
ggsave(paste0("plots/popbins_maxsize.png"),p, width=ggwidth, height=ggheight)
g <- ggplot(popbins.k.scalars, aes(k, depletion_RE, group=sigmaR, color=factor(sigmaR)))+geom_line()
ggsave("plots/popbins_k.png", g, width=ggwidth, height=ggheight)
g <- ggplot(popbins.Linf.scalars, aes(Linf, depletion_RE, group=sigmaR, color=factor(sigmaR)))+geom_line()
ggsave("plots/popbins_Linf.png", g, width=ggwidth, height=ggheight)
g <- ggplot(popbins.Lmin.scalars, aes(Lmin, depletion_RE, group=sigmaR, color=factor(sigmaR)))+geom_line()
ggsave("plots/popbins_Lmin.png", g, width=ggwidth, height=ggheight)
### ------------------------------------------------------------
### For effect binning
for(spp in species){
g <- plot_scalar_boxplot(subset(binning.long.growth, species==spp), x="variable", y='value',
vert='species', vert2="data", rel=TRUE,
horiz2="dbin", horiz="pbin", print=FALSE) +
theme(axis.text.x=element_text(angle=90))
ggsave(paste0("plots/binning_growth_errors_",spp, ".png"),g, width=ggwidth, height=ggheight)
g <- plot_scalar_boxplot(subset(binning.long.figure, species==spp), x="dbin", y='value',
vert='species', horiz="data", rel=TRUE,
vert2="variable", horiz2="binmatch", print=FALSE) +
theme(axis.text.x=element_text(angle=90))
ggsave(paste0("plots/binning_figure_errors_",spp, ".png"),g, width=ggwidth, height=ggheight)
g <- plot_scalar_boxplot(subset(binning.long.selex, species==spp), x="variable", y='value', vert2='species', vert="data", rel=TRUE, horiz2="dbin", horiz="pbin", print=FALSE)+
theme(axis.text.x=element_text(angle=90))
ggsave(paste0("plots/binning_selex_errors_", spp, ".png"), g, width=ggwidth, height=ggheight)
g <- plot_scalar_boxplot(subset(binning.long.management, species==spp), x="variable",
y='value', vert2='species', vert="data", rel=TRUE, horiz2="dbin",
horiz="pbin", print=FALSE)+
theme(axis.text.x=element_text(angle=90))
ggsave(paste0("plots/binning_management_errors_", spp, ".png"),g, width=ggwidth, height=ggheight)
g <- ggplot(subset(binning.unfiltered, species==spp), aes(x=data, y=log_max_grad, color=runtime, size=params_on_bound_em,))+
geom_jitter()+
facet_grid(species~dbin+pbin)+
geom_hline(yintercept=log(.01), col='red')
ggsave(paste0("plots/binning_convergence_", spp, ".png"),g, width=ggwidth, height=ggheight)
## Plot MAREs vs iteration to address stability
g <- ggplot(data=subset(binning.long.growth.mares, species==spp), aes(x=replicate2, y=cMARE, group=variable, color=variable))+
ylab("Centered MARE") +xlab("Replicate")
g <- g+geom_line(lwd=.5, alpha=1)+facet_grid(B~data)+ ylim(-.25, .25)
ggsave(paste0("plots/binning_growth_MAREs_",spp, ".png"), width=ggwidth, height=ggheight, units="in")
g <- ggplot(data=subset(binning.long.management.mares, species==spp), aes(x=replicate2, y=cMARE, group=variable, color=variable))+
ylab("Centered MARE") +xlab("Replicate")
g <- g+geom_line(lwd=.5, alpha=1)+facet_grid(B~data)+ ylim(-.25, .25)
ggsave(paste0("plots/binning_management_MAREs_",spp, ".png"), width=ggwidth, height=ggheight, units="in")
g <- ggplot(data=subset(binning.long.selex.mares, species==spp), aes(x=replicate2, y=cMARE, group=variable, color=variable))+
ylab("Centered MARE") +xlab("Replicate")
g <- g+geom_line(lwd=.5, alpha=1)+facet_grid(B~data)+ ylim(-.25, .25)
ggsave(paste0("plots/binning_selex_MAREs_",spp, ".png"), width=ggwidth, height=ggheight, units="in")
## ## time series plots
## myylim <- ylim(-3,3)
## g <- plot_ts_boxplot(subset(binning.ts, species==spp), y="SpawnBio_re", horiz="data",
## vert="dbin", vert2="pbin", print=FALSE, rel=FALSE)+myylim
## ggsave(paste0("plots/binning_ts_SBB_", spp, ".png"),g, width=ggwidth, height=ggheight)
## g <- plot_ts_boxplot(subset(binning.ts, species==spp), y="Recruit_0_re", horiz="data",
## vert="dbin", vert2="pbin", print=FALSE, rel=FALSE)+myylim
## ggsave(paste0("plots/binning_ts_recruits_", spp, ".png"),g, width=ggwidth, height=ggheight)
## g <- plot_ts_boxplot(subset(binning.ts, species==spp), y="F_re", horiz="data",
## vert="dbin", vert2="pbin", print=FALSE, rel=FALSE)+myylim
## ggsave(paste0("plots/binning_ts_F_", spp, ".png"),g, width=ggwidth, height=ggheight)
g <- ggplot(subset(binning.long.growth2, species==spp & B %in% c("B1","B2", "B3", "B4")), aes(x=B0, y=value, color=B)) + geom_point(alpha=.5)+
facet_wrap('variable', scales='fixed')+xlim(-1,1)+ylim(-1,1) + ggtitle("Pairwise RE with base case for unmatching bins")+
geom_hline(yintercept=0, col=gray(.5))+geom_vline(xintercept=0, col=gray(.5))
ggsave(paste0("plots/binning_scatter_", spp, ".png"),g, width=ggwidth, height=ggheight)
g <- ggplot(subset(binning.long.growth2, species==spp & !B %in% c("B1","B2", "B3", "B4")), aes(x=B0, y=value, color=B)) + geom_point(alpha=.5)+
facet_wrap('variable', scales='fixed')+xlim(-1,1)+ylim(-1,1) +
ggtitle("Pairwise RE with base case for matching bins")+
geom_hline(yintercept=0, col=gray(.5))+geom_vline(xintercept=0, col=gray(.5))
ggsave(paste0("plots/binning_scatter2_", spp, ".png"),g, width=ggwidth, height=ggheight)
}
g <- ggplot(binning.counts, aes(x=data.binwidth, y=pct.converged, color=binmatch, group=binmatch))+
geom_line()+ facet_grid(species~data, scales='fixed') +
theme(axis.text.x=element_text(angle=90))
ggsave("plots/binning_convergence_counts.png", g, width=ggwidth, height=ggheight)
g <- ggplot(binning.counts, aes(x=data.binwidth, y=median.iterations, color=binmatch, group=binmatch))+
geom_point(aes(size=pct.converged))+ facet_grid(species~data, scales='free') + geom_line()+
theme(axis.text.x=element_text(angle=90))
ggsave("plots/binning_convergence_iterations.png", g, width=ggwidth, height=ggheight)
g <- ggplot(binning.runtime, aes(x=dbin, y=median.runtime, color=binmatch, group=binmatch))+
geom_line()+ facet_grid(species~data, scales='free') +
theme(axis.text.x=element_text(angle=90))+
geom_point(aes(size=(replicates)))+
geom_linerange(aes(ymin=lower.runtime, ymax=upper.runtime))
ggsave("plots/binning_median_runtime.png", g, width=ggwidth, height=ggheight)
g <- ggplot(binning.stuck, aes(data.binwidth, pct.stuck, color=which.bound))+geom_point(lwd=1.5)+
facet_grid(species+data+binmatch~variable) + ylim(0,1)
ggsave("plots/binning_stuck_params.png", g, width=1.25*ggwidth, height=ggheight)
## is the max gradient OK?
xx <- binning.unfiltered
levels(xx$species) <- c('cod', 'flatfish', 'rockfish')
xx$model.binwidth <- as.numeric(gsub('pbin=', '', x=xx$pbin))
g <- ggplot(xx, aes(log_max_grad, SSB_MSY_re, color=model.binwidth))+
geom_point(alpha=.5)+
facet_grid(data~species) + ylim(-2,2) + geom_vline(xintercept=log(.1)) +
theme_bw()
ggsave('plots/gradient_checks.png', g, width=9, height=5)
## distribution of relative errors for one binning case
xx <- binning.long.growth
levels(xx$species) <- c('cod', 'flatfish', 'rockfish')
for(var in unique(binning.long.growth$variable )){
g <- ggplot(subset(xx, variable==var & B=='B0' & dbin=='dbin=1'), aes(value))+geom_histogram() +
facet_grid(species~data, scales='free') +
ggtitle('Relative Errors for 1cm data and model bins') +
geom_vline(yintercept=0, color='red') + theme_bw() +
ylab(paste('Relative Error for:', var))
ggsave(paste0('plots/binning_histogram_', var, '.png'), g, width=9, height=6)
}
## ## quick plots of bias adjustment
## ggplot(subset(bias.all.long, species=='flatfish'), aes(x=data, y=value))+geom_point() +
## facet_grid(variable~species+B, scales="free_y")
## ggplot(scenarios.converged, aes(B, pct.converged))+geom_point()+facet_grid(species~data)
## ##### For tail compression
## for(spp in species){
## g <- plot_scalar_boxplot(subset(tcomp.long.growth, species==spp ),
## x="tvalue", y='value', vert='data', vert2="B",
## rel=TRUE, horiz="variable", print=FALSE) + ggtitle(spp)+
## theme(axis.text.x=element_text(angle=90))
## ggsave(paste0("plots/tcomp_growth_errors_",spp,".png"),g, width=ggwidth, height=ggheight)
## g <- plot_scalar_boxplot(subset(tcomp.long.selex, species==spp ),
## x="tvalue", y='value', vert='data', vert2="B",
## rel=TRUE, horiz="variable", print=FALSE) + ggtitle(spp)+
## theme(axis.text.x=element_text(angle=90))
## ggsave(paste0("plots/tcomp_selex_errors_",spp,".png"),g, width=ggwidth, height=ggheight)
## g <- plot_scalar_boxplot(subset(tcomp.long.management, species==spp ),
## x="tvalue", y='value', vert='data', vert2="B",
## rel=TRUE, horiz="variable", print=FALSE) + ggtitle(spp)+
## theme(axis.text.x=element_text(angle=90))
## ggsave(paste0("plots/tcomp_management_errors_",spp,".png"),g, width=ggwidth, height=ggheight)
## }
## g <- ggplot(tcomp.unfiltered,
## aes(x=tvalue, y=log_max_grad, color=runtime,
## size=params_on_bound_em,))+
## geom_jitter()+ facet_grid(species~data)+
## geom_hline(yintercept=log(.01), col='red')+
## theme(axis.text.x=element_text(angle=90))
## ggsave("plots/tcomp_convergence.png",g, width=ggwidth, height=ggheight)
## g <- ggplot(tcomp.counts, aes(x=tvalue, y=pct.converged))+facet_grid(species~data+B)+
## geom_bar(stat='identity', aes(fill=pct.converged)) + theme(axis.text.x=element_text(angle=90))
## ggsave("plots/tcomp_convergence_counts.png", g, width=1.25*ggwidth, height=ggheight)
## g <- ggplot(tcomp.stuck, aes(I, pct.stuck, group=which.bound, color=which.bound))+geom_line(lwd=1.5)+
## facet_grid(species+D~variable) + ylim(0,1)
## ggsave("plots/tcomp_stuck_params.png", g, width=1.25*ggwidth, height=ggheight)
## ### end of compression plots
## ### ------------------------------------------------------------
## ##### For robustification constant
## for(spp in species){
## g <- plot_scalar_boxplot(subset(robust.long.growth, species==spp ),
## x="rvalue", y='value', vert='data', vert2="B",
## rel=TRUE, horiz="variable", print=FALSE) + ggtitle(spp)+
## theme(axis.text.x=element_text(angle=90))
## ggsave(paste0("plots/robust_growth_errors_",spp,".png"),g, width=ggwidth, height=ggheight)
## g <- plot_scalar_boxplot(subset(robust.long.selex, species==spp ),
## x="rvalue", y='value', vert='data', vert2="B",
## rel=TRUE, horiz="variable", print=FALSE) + ggtitle(spp)+
## theme(axis.text.x=element_text(angle=90))
## ggsave(paste0("plots/robust_selex_errors_",spp,".png"),g, width=ggwidth, height=ggheight)
## g <- plot_scalar_boxplot(subset(robust.long.management, species==spp ),
## x="rvalue", y='value', vert='data', vert2="B",
## rel=TRUE, horiz="variable", print=FALSE) + ggtitle(spp)+
## theme(axis.text.x=element_text(angle=90))
## ggsave(paste0("plots/robust_management_errors_",spp,".png"),g, width=ggwidth, height=ggheight)
## }
## g <- ggplot(robust.unfiltered,
## aes(x=rvalue, y=log_max_grad, color=runtime,
## size=params_on_bound_em,))+
## geom_jitter()+ facet_grid(species~data)+
## geom_hline(yintercept=log(.01), col='red')+
## theme(axis.text.x=element_text(angle=90))
## ggsave("plots/robust_convergence.png",g, width=ggwidth, height=ggheight)
## g <- ggplot(robust.counts, aes(x=rvalue, y=pct.converged))+facet_grid(species~data+B)+
## geom_bar(stat='identity', aes(fill=pct.converged)) + theme(axis.text.x=element_text(angle=90))
## ggsave("plots/robust_convergence_counts.png", g, width=1.25*ggwidth, height=ggheight)
## g <- ggplot(robust.stuck, aes(I, pct.stuck, group=which.bound, color=which.bound))+geom_line(lwd=1.5)+
## facet_grid(species+D~variable) + ylim(0,1)
## ggsave("plots/robust_stuck_params.png", g, width=1.25*ggwidth, height=ggheight)
## ### end of robust plots
## ### ------------------------------------------------------------
## ## Temp code to explore correlations, needs to be fixed
## temp <- subset(binning, species=='cod')
## ggplot(temp, aes(L_at_Amin_Fem_GP_1_re, CV_young_Fem_GP_1_re))+
## geom_point()+facet_grid(D~B)+xlim(-1,1)+ylim(-1,1)
## ggplot(temp, aes(VonBert_K_Fem_GP_1_re, CV_young_Fem_GP_1_re))+
## geom_point()+facet_grid(D~B)+xlim(-1,1)+ylim(-1,1)
## ggplot(temp, aes(L_at_Amin_Fem_GP_1_re, VonBert_K_Fem_GP_1_re))+
## geom_point()+facet_grid(D~B)+xlim(-1,1)+ylim(-1,1)
## ggplot(temp, aes(CV_young_Fem_GP_1_re, SSB_MSY_re))+
## geom_point()+facet_grid(D~B)+xlim(-1,1)+ylim(-1,1)
## ggplot(temp, aes(LnQ_base_2_Survey_re, SSB_MSY_re))+
## geom_point()+facet_grid(D~B)+xlim(-1,1)+ylim(-1,1)
## ggplot(temp, aes(CV_young_Fem_GP_1_re, SSB_MSY_re))+
## geom_point()+facet_grid(D~B)+xlim(-1,1)+ylim(-1,1)
## ggplot(temp, aes(VonBert_K_Fem_GP_1_re, SSB_MSY_re))+
## geom_point()+facet_grid(D~B)+xlim(-1,1)+ylim(-1,1)
## ggplot(temp, aes(L_at_Amax_Fem_GP_1_re, SSB_MSY_re))+
## geom_point()+facet_grid(D~B)+xlim(-1,1)+ylim(-1,1)
## ggplot(temp, aes(SizeSel_2P_3_Survey_re, SSB_MSY_re))+
## geom_point()+facet_grid(D~B)+xlim(-1,1)+ylim(-1,1)
## ggplot(temp, aes(SizeSel_2P_1_Survey_re, SSB_MSY_re))+
## geom_point()+facet_grid(D~B)+xlim(-1,1)+ylim(-1,1)
## ggplot(temp, aes(CV_young_Fem_GP_1_re, SSB_Unfished_re))+
## geom_point()+facet_grid(D~B)+xlim(-1,1)+ylim(-1,1)
## ggplot(temp, aes(LnQ_base_2_Survey_re, SSB_Unfished_re))+
## geom_point()+facet_grid(D~B)+xlim(-1,1)+ylim(-1,1)
## ggplot(temp, aes(CV_young_Fem_GP_1_re, SSB_Unfished_re))+
## geom_point()+facet_grid(D~B)+xlim(-1,1)+ylim(-1,1)
## ggplot(temp, aes(VonBert_K_Fem_GP_1_re, SSB_Unfished_re))+
## geom_point()+facet_grid(D~B)+xlim(-1,1)+ylim(-1,1)
## ggplot(temp, aes(L_at_Amax_Fem_GP_1_re, SSB_Unfished_re))+
## geom_point()+facet_grid(D~B)+xlim(-1,1)+ylim(-1,1)
## ggplot(temp, aes(SizeSel_2P_3_Survey_re, SSB_Unfished_re))+
## geom_point()+facet_grid(D~B)+xlim(-1,1)+ylim(-1,1)
## ggplot(temp, aes(SizeSel_2P_1_Survey_re, SSB_Unfished_re))+
## geom_point()+facet_grid(D~B)+xlim(-1,1)+ylim(-1,1)