analyses_de_pl/plots_data.R at revision · KITmetricslab/analyses_de_pl · GitHub

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
# Some plots showing the time series of cases and deaths in DE + PL

library(here)
library(plotrix)

setwd("/home/johannes/Documents/Projects/intermediate_results/R")
source("functions_paper.R")


Sys.setlocale(category = "LC_TIME", locale = "en_US.UTF8")

path_hub <- "/home/johannes/Documents/Projects/covid19-forecast-hub-de"
source(paste0(path_hub, "/code/R/auxiliary_functions.R"))

# read in data:
dat_ecdc <- read.csv(paste0(path_hub, "/app_forecasts_de/data/truth_to_plot_ecdc.csv"), colClasses = list(date = "Date"))
dat_jhu <- read.csv(paste0(path_hub, "/app_forecasts_de/data/truth_to_plot_jhu.csv"), colClasses = list(date = "Date"))


par(las = 1)
options(scipen=5)


# custom function to plot time series:
plot_ts <- function(dat_ecdc, dat_jhu, variable, main = "", col_gm = "black", col_pl = "red", points = FALSE,
                    cex.points = 0.9, highlight = NULL, col_highlight = "lightgrey", ylim = NULL, add = FALSE, ...){
  dat_ecdc_gm <- subset(dat_ecdc, location == "GM")
  dat_ecdc_pl <- subset(dat_ecdc, location == "PL")
  dat_jhu_gm <- subset(dat_jhu, location == "GM")
  dat_jhu_pl <- subset(dat_jhu, location == "PL")

  if(!add){
    if(is.null(ylim)) ylim <- c(0, 1.2*max(c(dat_ecdc_gm[, variable],
                                             dat_jhu_gm[, variable],
                                             dat_ecdc_pl[, variable],
                                             dat_jhu_pl[, variable]), na.rm = TRUE))

    plot(dat_ecdc_gm$date, dat_ecdc_gm[, variable], type = "l", xlab = "", ylab = "", col = "white", ylim = ylim, ...)
    if(!is.null(highlight)) rect(xleft = highlight[1], ybottom = 0, xright = highlight[2], ytop = 10^7, col = col_highlight, border = NA)
  }

  lines(dat_ecdc_gm$date, dat_ecdc_gm[, variable], type = "l", col = col_gm)
  lines(dat_jhu_gm$date, dat_jhu_gm[, variable], type = "l", lty = 2, col = col_gm)
  lines(dat_ecdc_pl$date, dat_ecdc_pl[, variable], type = "l", col = col_pl, cex = 0.7)
  lines(dat_jhu_pl$date, dat_jhu_pl[, variable], type = "l", lty = 2, col = col_pl)

  if(points){
    points(dat_ecdc_gm$date, dat_ecdc_gm[, variable], pch = 15, xlab = "", ylab = "", cex = cex.points, col = col_gm)
    points(dat_jhu_gm$date, dat_jhu_gm[, variable], pch = 17, xlab = "", ylab = "", col = col_gm, cex = cex.points)
    points(dat_ecdc_pl$date, dat_ecdc_pl[, variable], pch = 15, xlab = "", ylab = "", cex = cex.points, col = col_pl)
    points(dat_jhu_pl$date, dat_jhu_pl[, variable], pch = 17, xlab = "", ylab = "", cex = cex.points, col = col_pl)
  }

  mtext(side = 3, main, cex = 0.8, line = 0.3)
}

# helper function to add axes:
add_axes <- function(xticks, xticks_labelled){
  axis(1, at = xticks, labels = rep("", length(xticks)), tck=-0.015)
  axis(1, at = xticks_labelled, labels = xticks_labelled, tck = -0.03)
  axis(2); box()
}
# ticks:
xticks <- seq(from = as.Date("2020-09-12"), by = 7, length.out = 16)
xticks_labelled <- xticks[c(1, 5, 9, 13)]

# define area to highlight:
highlight <- c(as.Date("2020-10-12"), as.Date("2020-12-19"))


pdf("../figures/time_series.pdf", width = 8.5, height = 5.5)

# structure plot area
layout(matrix(1:6, ncol = 2), heights = c(1.4, 2.6, 2.1))

# Cases:

# small plot showing also first wave:
par(mar = c(2.5, 4, 2, 2), las = 1)
plot_ts(dat_ecdc, dat_jhu, "inc_case", highlight = highlight, points = TRUE,
        main = "Weekly incident cases", cex.points = 0.7)

# large time series plot:
par(las = 1, mar = c(2.5, 4, 1, 2))
plot_ts(dat_ecdc, dat_jhu, "inc_case", highlight = highlight,
        xlim = highlight + c(-38, 14), axes = FALSE,
        points = TRUE)
add_axes(xticks = xticks, xticks_labelled = xticks_labelled)

# Add letters/numbers for events:
letter_in_circle(as.Date("2020-09-03"), 200000, "1", col = "red", col_line = "lightgrey") # PL: strict test criteria
letter_in_circle(as.Date("2020-10-10"), 200000, "2", col = "red", col_line = "lightgrey") # PL: entire country classified as "yellow zone"
letter_in_circle(as.Date("2020-10-24"), 200000, "3", col = "red") # PL: entire country classified as "red zone" (partial school closure, restaurants closed, gatherings restrictied)
letter_in_circle(as.Date("2020-10-31"), 180000, "4", col = "red") # PL: relaxed test criteria
letter_in_circle(as.Date("2020-11-02"), 200000, "a") # DE: semi-lockdown
letter_in_circle(as.Date("2020-11-03"), 180000, "b") # DE: new test strategy
letter_in_circle(as.Date("2020-11-07"), 200000, "5", col = "red") # PL: reinforced restrictions (extension of school closures, shooping malls closed)
letter_in_circle(as.Date("2020-11-25"), 200000, "6", col = "red") # PL: bulk reporting in PL
letter_in_circle(as.Date("2020-11-28"), 180000, "7", col = "red") # PL: Malls reopened
letter_in_circle(as.Date("2020-12-01"), 200000, "c") # DE: reinforcement of contact restrictions https://www.tagesschau.de/inland/corona-plan-bundeslaender-beschluss-103.html
letter_in_circle(as.Date("2020-12-16"), 200000, "d") # DE: full lockdown (incl school closure)

# avoid overplotting with vertival dashed lines
plot_ts(dat_ecdc, dat_jhu, "inc_case", add = TRUE, points = TRUE)

# explanation of events:
plot(NULL, xlim = 0:1, ylim = 0:1, axes = FALSE, xlab = "", ylab = "")

legend("center", legend = c("Measures in Germany",
                            "a - semi-lockdown (restaurants closed, work f. home encouraged,",
                            "   no mass gatherings, inter-househ. contacts restr.; 2020-11-02)",
                            "b - new testing strategy announced (stricter criteria on symptoms;",
                            "  2020-11-03)",
                            "c - reinforced restrictions on inter-househ. contacts (2020-12-01)",
                            "d - start of full lockdown (schools and non-essential shops;",
                            "  closed; 2020-12-16)"), bty = "n")

# Deaths:

# small plot showing also first wave:
par(mar = c(2.5, 4, 2, 2), las = 1)
plot_ts(dat_ecdc, dat_jhu, "inc_death", highlight = highlight, points = TRUE,
        main = "Weekly incident deaths", cex.points = 0.7)

# large time series plot:
par(las = 1, mar = c(2.5, 4, 1, 2))
plot_ts(dat_ecdc, dat_jhu, "inc_death", highlight = highlight,
        xlim = highlight + c(-38, 14), ylim = c(0, 5200),  axes = FALSE,
        points = TRUE)
add_axes(xticks = xticks, xticks_labelled = xticks_labelled)

# events:
letter_in_circle(as.Date("2020-09-03"), 5000, "1", col = "red", col_line = "lightgrey") # PL: stricter testing criteria
letter_in_circle(as.Date("2020-10-10"), 5000, "2", col = "red", col_line = "lightgrey") # PL: entire country classified as "yellow zone"
letter_in_circle(as.Date("2020-10-24"), 5000, "3", col = "red") # PL: entire country classified as "red zone" (partial school closure, restaurants closed, gatherings restrictied)
letter_in_circle(as.Date("2020-10-31"), 4500, "4", col = "red") # PL: relaxed test criteria
letter_in_circle(as.Date("2020-11-02"), 5000, "a") # DE: semi-lockdown
letter_in_circle(as.Date("2020-11-03"), 4500, "b") # DE: new test strategy
letter_in_circle(as.Date("2020-11-07"), 5000, "5", col = "red") # PL: reinforced restrictions (extension of school closures, shooping malls closed)
letter_in_circle(as.Date("2020-11-25"), 5000, "6", col = "red") # PL: bulk reporting in PL
letter_in_circle(as.Date("2020-11-28"), 4500, "7", col = "red") # PL: Malls reopened
letter_in_circle(as.Date("2020-12-01"), 5000, "c") # DE: reinforcement of contact restrictions https://www.tagesschau.de/inland/corona-plan-bundeslaender-beschluss-103.html
letter_in_circle(as.Date("2020-12-16"), 5000, "d") # DE: full lockdown (incl school closure)

legend("bottomright", legend = c("Germany", "Poland", "", "ECDC", "JHU"),
       lty = c(1, 1, NA, 1, 2), col = c("black", "red", NA, "black", "black"), bty = "n",
       pch = c(15, 15, NA, 15, 17))

plot_ts(dat_ecdc, dat_jhu, "inc_death", add = TRUE, points = TRUE)

# Explanation of measures in Poland:
plot(NULL, xlim = 0:1, ylim = 0:1, axes = FALSE, xlab = "", ylab = "")

legend("center", legend = c("Measures in Poland",
                            "1 - stricter testing criteria (four symptoms needed; 2020-09-03)",
                            "2 - entire country classified yellow zone (mask wearing and",
                            "  limitations on gatherings; 2020-10-10)",
                            "3 - entire country classified red zone (partial school closure,",
                            "  restaurants closed, gatherings restricted; 2020-10-24)",
                            "4 - eased testing criteria (one symptom sufficient; 2020-10-31)",
                            "5 - reinforced restrictions (extension of school closure, ",
                            "  shopping malls closed; 2020-11-07)",
                            "6 - bulk reporting of 22,000 cases (2020-11-24)",
                            "7 - shopping malls re-opened (2020-11-28)"),
       bty = "n", text.col = "red")

dev.off()

# plot by age, DE (for supplement):
# read in age-stratified data from Hub
dat_ecdc_age_case <- read.csv(paste0(path_hub, "/data-truth/RKI/by_age/truth_RKI-Incident Cases by Age_Germany.csv"), colClasses = list(date = "Date"))
dat_ecdc_age_death <- read.csv(paste0(path_hub, "/data-truth/RKI/by_age/truth_RKI-Incident Deaths by Age_Germany.csv"), colClasses = list(date = "Date"))

# transform to weekly:
dat_ecdc_case_all <- inc_truth_to_weekly(subset(dat_ecdc_age_case, location == "GM"))
dat_ecdc_case_80 <- inc_truth_to_weekly(subset(dat_ecdc_age_case, age_group == "A80+" &
                                                 location == "GM"))
dat_ecdc_case_others <- inc_truth_to_weekly(subset(dat_ecdc_age_case, age_group != "A80+" &
                                                         location == "GM"))


dat_ecdc_death_all <- inc_truth_to_weekly(subset(dat_ecdc_age_death, location == "GM"))
dat_ecdc_death_80 <- inc_truth_to_weekly(subset(dat_ecdc_age_death, age_group == "A80+" &
                                                 location == "GM"))
dat_ecdc_death_others <- inc_truth_to_weekly(subset(dat_ecdc_age_death, age_group != "A80+" &
                                                     location == "GM"))

# Plot:

pdf("../figures/time_series_age.pdf", width = 12.5, height = 6)

# Cases:
par(mfrow = 1:2, las = 1, mar = c(4, 4, 1.5, 1))
plot(dat_ecdc_case_all$date, dat_ecdc_case_all$value, xlab = "", ylab = "",
     xlim = highlight + c(-28, 14), type = "o", pch = 15,
     ylim = c(0, 200000), axes = FALSE)
add_axes(xticks = xticks, xticks_labelled = xticks_labelled)

mtext("Weekly incident cases by age, Germany", line = 0.3, cex = 1.7)
rect(xleft = highlight[1], ybottom = 0, xright = highlight[2], ytop = 10^7, col = "lightgrey", border = NA)

letter_in_circle(as.Date("2020-11-02"), 200000, "a") # DE: semi-lockdown
letter_in_circle(as.Date("2020-11-03"), 180000, "b") # DE: new test strategy
letter_in_circle(as.Date("2020-12-01"), 200000, "c") # DE: reinforcement of contact restrictions https://www.tagesschau.de/inland/corona-plan-bundeslaender-beschluss-103.html
letter_in_circle(as.Date("2020-12-16"), 200000, "d") # DE: full lockdown (incl school closure)

lines(dat_ecdc_case_all$date, dat_ecdc_case_all$value, type = "o", pch = 15)
lines(dat_ecdc_case_80$date, dat_ecdc_case_80$value, type = "o", lty = 3)
lines(dat_ecdc_case_others$date, dat_ecdc_case_others$value, type = "o", lty = 4, pch = 4)

legend("topleft", lty = c(1, 4, 3), legend = c("total", "younger than 80", "80 or older"),
       bty = "n", pch = c(15, 3, 1))

# Deaths:
plot(dat_ecdc_death_all$date, dat_ecdc_death_all$value, type = "l", xlab = "", ylab = "", col = "white",
     xlim = highlight + c(-28, 14), ylim = c(0, 5000), axes = FALSE)
add_axes(xticks = xticks, xticks_labelled = xticks_labelled)
mtext("Weekly incident deaths by age, Germany", line = 0.3, cex = 1.7)
rect(xleft = highlight[1], ybottom = 0, xright = highlight[2], ytop = 10^7, col = "lightgrey", border = NA)

letter_in_circle(as.Date("2020-11-02"), 5000, "a") # DE: semi-lockdown
letter_in_circle(as.Date("2020-11-03"), 4500, "b") # DE: new test strategy
letter_in_circle(as.Date("2020-12-01"), 5000, "c") # DE: reinforcement of contact restrictions https://www.tagesschau.de/inland/corona-plan-bundeslaender-beschluss-103.html
letter_in_circle(as.Date("2020-12-16"), 5000, "d") # DE: full lockdown (incl school closure)

lines(dat_ecdc_death_all$date, dat_ecdc_death_all$value, type = "o", pch = 15)
lines(dat_ecdc_death_80$date, dat_ecdc_death_80$value, type = "o", lty = 3)
lines(dat_ecdc_death_others$date, dat_ecdc_death_others$value, type = "o", lty = 4, pch = 4)

dev.off()