-
Notifications
You must be signed in to change notification settings - Fork 0
/
chapter-3-2-plots.R
113 lines (96 loc) · 6.07 KB
/
chapter-3-2-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
require(DBI)
require(dplyr)
require(ggplot2)
require(GGally)
require(ggpubr)
hdbc <- odbcConnect("cbip", case="nochange")
# Figure 4
ggplot(gdf14, aes(x=v, y=avgThold)) +
geom_point(aes(size=okr / cGmina, col=s)) +
scale_size(range=c(0.1, 1)) +
scale_color_gradient2(high="green", low="red", mid="blue", midpoint=0.2675) +
labs(y = "t", size = "c", color = "s") + theme(legend.position = "top")
# Figure 5
ggplot(gdf14, aes(x=v, y=avgThold)) +
geom_point(aes(size=okr / cGmina, col=npMeanS)) +
scale_size(range=c(0.1, 1)) +
scale_color_gradient2(high="green", low="red", mid="blue", midpoint=0.2675) +
labs(y = "t", size = "c", color = "s") + theme(legend.position = "top")
# Figure 6
# Figure 7
# Figure 8
ggplot(filter(gdf14, okr > 0), aes(normV, s, col = uq + 1, size=(okr / cGmina)^2)) +
geom_point() + xlab(expression(lambda)) +
scale_size(range = c(1/5, 1)) +
scale_color_gradient2(high="darkblue", mid="blue", low="red", midpoint=4.4, trans="log", labels = scales::number_format(accuracy = 1)) +
theme(legend.title = element_blank()) +
theme(legend.position="top", legend.key.width=unit(1,"cm"))
# Figure 9
ggplot(filter(gdf14, okr > 0), aes(normV, s, col = uqNorm + 1, size=(okr / cGmina)^2)) +
geom_point() + xlab(expression(lambda)) +
scale_size(range = c(1/5, 1)) +
scale_color_gradient2(high="darkblue", mid="blue", low="red", midpoint=4.4, trans="log", labels = scales::number_format(accuracy = 1)) +
theme(legend.title = element_blank()) +
theme(legend.position="top", legend.key.width=unit(1,"cm"))
# Figure 10
marginDF <- sqlQuery(hdbc, "SELECT g1.rok, g1.gmina, g1.nazwa, g1.lista, g1.mandatow - g2.mandatow AS sDiff, w.margin FROM gerry.gerry AS g1
INNER JOIN gerry.gerry AS g2 ON g2.rok = g1.rok AND g2.gmina = g1.gmina AND g2.rankS = 2
INNER JOIN (
SELECT rok, gmina, lista, RANK() OVER (PARTITION BY rok, gmina, lista ORDER BY zw - runUp DESC) AS mrank,
zw - runUp AS margin
FROM gerry.gerryWybory WHERE mandat = 1
) AS w ON g1.rok = w.rok AND g1.gmina = w.gmina AND g1.lista = w.lista AND w.mrank = g1.mandatow - g2.mandatow
WHERE g1.rok != 9999 AND g1.rankS = 1 ORDER BY sDiff DESC;")
lplot <- ggplot(filter(wyb, mandat==1), aes(x = (zw - runUp) / 1)) +
geom_density() + xlab(expression(m[k])) + xlim(0, 500) + theme(axis.title.y=element_blank())
rplot <- ggplot(marginDF, aes(x = margin)) +
geom_density() + xlab(expression(M)) + xlim(0, 500) + theme(axis.title.y=element_blank())
ggarrange(lplot, rplot)
# Figure 11
pvcdf <- with(filter(gdf14, okr >= cGmina / 2), ecdf(pValue))
nvcdf <- with(filter(gdf14, okr >= cGmina / 2), ecdf(phi))
uwcdf <- with(filter(gdf14, okr >= cGmina / 2), ecdf(uqNorm))
g12 <- ggplot(filter(gdf14, okr >= cGmina / 2), aes(qnorm(pvcdf(pValue)), qnorm(nvcdf(phi)), col=s)) + geom_point(size=0.7)
g21 <- ggplot(filter(gdf14, okr >= cGmina / 2), aes(qnorm(nvcdf(phi)), qnorm(pvcdf(pValue)), col=s)) + geom_point(size=0.7)
g13 <- ggplot(filter(gdf14, okr >= cGmina / 2), aes(qnorm(pvcdf(pValue)), qnorm(uwcdf(uqNorm)), col=s)) + geom_point(size=0.7)
g31 <- ggplot(filter(gdf14, okr >= cGmina / 2), aes(qnorm(uwcdf(uqNorm)), qnorm(pvcdf(pValue)), col=s)) + geom_point(size=0.7)
g23 <- ggplot(filter(gdf14, okr >= cGmina / 2), aes(qnorm(nvcdf(phi)), qnorm(uwcdf(uqNorm)), col=s)) + geom_point(size=0.7)
g32 <- ggplot(filter(gdf14, okr >= cGmina / 2), aes(qnorm(uwcdf(uqNorm)), qnorm(nvcdf(phi)), col=s)) + geom_point(size=0.7)
g11 <- ggplot(gdf14) + geom_density(mapping = aes(pValue), size=1.1)
g22 <- ggplot(gdf14) + geom_density(mapping = aes(phi), size=1.1)
g33 <- ggplot(gdf14) + geom_density(mapping = aes(uqNorm), size=1.1)
ggmatrix(list(g11, g12, g13, g21, g22, g23, g31, g32, g33), 3, 3,
xAxisLabels = c("P", "Phi", "u"),
yAxisLabels = c("P", "Phi", "u"),
byrow = FALSE,
title = NULL, xlab = NULL, ylab = NULL,
showYAxisPlotLabels = FALSE, showXAxisPlotLabels = FALSE
)
# Figure 12
dbcon <- dbConnect(odbc::odbc(), "cbip", timeout = 300)
suspects <- dbGetQuery(dbcon, "SELECT * FROM gerry.suspects")
suspects <- rename(suspects, P = svDirichlet, phi = svKernel, u = svOutlier)
pcdf <- with(filter(suspects, gcUnopp <= 5), ecdf(P))
ncdf <- with(filter(suspects, gcUnopp <= 5), ecdf(phi))
ucdf <- with(filter(suspects, gcUnopp <= 5), ecdf(u))
bcdf <- with(filter(suspects, gcUnopp <= 5), ecdf(biasL1))
g12 <- ggplot(filter(suspects, gcUnopp <= 5), aes(qnorm(pcdf(P)), qnorm(ncdf(phi)), col=komGmina)) + geom_point(size=0.7)
g21 <- ggplot(filter(suspects, gcUnopp <= 5), aes(qnorm(ncdf(phi)), qnorm(pcdf(P)), col=komGmina)) + geom_point(size=0.7)
g13 <- ggplot(filter(suspects, gcUnopp <= 5), aes(qnorm(pcdf(P)), qnorm(ucdf(u)), col=komGmina)) + geom_point(size=0.7)
g31 <- ggplot(filter(suspects, gcUnopp <= 5), aes(qnorm(ucdf(u)), qnorm(pcdf(P)), col=komGmina)) + geom_point(size=0.7)
g14 <- ggplot(filter(suspects, gcUnopp <= 5), aes(qnorm(pcdf(P)), qnorm(bcdf(biasL1)), col=komGmina)) + geom_point(size=0.7)
g41 <- ggplot(filter(suspects, gcUnopp <= 5), aes(qnorm(bcdf(biasL1)), qnorm(pcdf(P)), col=komGmina)) + geom_point(size=0.7)
g23 <- ggplot(filter(suspects, gcUnopp <= 5), aes(qnorm(ncdf(phi)), qnorm(ucdf(u)), col=komGmina)) + geom_point(size=0.7)
g32 <- ggplot(filter(suspects, gcUnopp <= 5), aes(qnorm(ucdf(u)), qnorm(ncdf(phi)), col=komGmina)) + geom_point(size=0.7)
g24 <- ggplot(filter(suspects, gcUnopp <= 5), aes(qnorm(ncdf(phi)), qnorm(bcdf(biasL1)), col=komGmina)) + geom_point(size=0.7)
g42 <- ggplot(filter(suspects, gcUnopp <= 5), aes(qnorm(bcdf(biasL1)), qnorm(ncdf(phi)), col=komGmina)) + geom_point(size=0.7)
g11 <- ggplot(suspects) + geom_density(mapping = aes(P), size=1.1)
g22 <- ggplot(suspects) + geom_density(mapping = aes(phi), size=1.1)
g33 <- ggplot(filter(suspects, !is.na(u))) + geom_density(mapping = aes(u), size=1.1)
g44 <- ggplot(filter(suspects, !is.na(biasL1))) + geom_density(mapping = aes(biasL1), size=1.1)
ggmatrix(list(g11, g12, g13, g21, g22, g23, g31, g32, g33), 3, 3,
xAxisLabels = c("P", "Phi", "u"),
yAxisLabels = c("P", "Phi", "u"),
title = NULL, xlab = NULL, ylab = NULL, byrow=FALSE,
showYAxisPlotLabels = FALSE, showXAxisPlotLabels = FALSE
)