-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathchina_italia_brasil.R
125 lines (113 loc) · 4.39 KB
/
china_italia_brasil.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
source("covid_est_exponencial.R") ## pra garantir dados o mais atualizados o possivel
library(zoo) ## Manipulacao de time series, agora nao muito importante, mas pode ser depois
################################################################################
## Italia
################################################################################
italia.n <- as.matrix(data[data$Country.Region=="Italy",-(1:4)])
dim(italia.n) <- NULL
## Junta às datas
italia <- zoo(x = italia.n, order.by= datas)
## Um plot da serie temporal completa
plot(italia, type="p")
## Minimo de casos para o dia zero
italia.n.casos <- 15
## N de dias para incluir no ajuste
dias.final <- 7
## Ajuste a partir do dia zero
italia.dia.zero <- min(which(italia>=italia.n.casos, arr.ind=TRUE))
## Ajuste do modelo aos primeiros 7 dias a partir do dia zero
x <- 0:(dias.final-1)
y <- italia[italia.dia.zero:(italia.dia.zero+dias.final-1)]
## Ajuste do modelo Poisson
## Inicio do periodo
italia.inicio.fit <- glm(y~x, family=poisson)
## Ultimos 7 dias
y <- italia[(length(italia)-dias.final+1):length(italia)]
italia.fim.fit <- glm(y~x, family=poisson)
## Coeficientes
coef(italia.inicio.fit)[2]
coef(italia.fim.fit)[2]
## tempos de duplicacao ##
## Inicio
log(2)/coef(italia.inicio.fit)[2]
## Atual
log(2)/coef(italia.fim.fit)[2]
## Intervalos de confianca
## Inicio
log(2)/confint(italia.inicio.fit)[2,]
## Fim
log(2)/confint(italia.fim.fit)[2,]
## Italia oms
italia.raw <- read.csv("https://covid.ourworldindata.org/data/full_data.csv", as.is=TRUE) %>%
filter(location=="Italy") %>%
select(date,total_cases)
italia.oms <- zoo(x = italia.raw$total_cases,
order.by = as.Date(italia.raw$date, "%Y-%m-%d"))
## Verificando imapcto dos ajustes lineares
## N de casos dia zero
it.oms.n <- 75
italia.oms.76 <- fitP.zoo(italia.oms[min(which(italia.oms>it.oms.n, arr.ind=TRUE)):length(italia.oms)], only.coef=FALSE)
italia.oms.fim <- fitP.zoo(italia.oms[(length(italia.oms)-7): length(italia.oms)], only.coef=FALSE)
## Previstos pelos dois modelos
newdata <- data.frame(ndias=rev(max(time(italia.oms))-time(italia.oms[which(italia.oms>it.oms.n, arr.ind=TRUE)])))
it.fit.76 <- predict(italia.oms.76, newdata=newdata, type="response")
## plot
plot(italia.oms[min(which(italia.oms>it.oms.n, arr.ind=TRUE)):length(italia.oms)], log="y", type="p")
lines(zoo(it.fit.76, time(italia.oms[min(which(italia.oms>it.oms.n, arr.ind=TRUE)):length(italia.oms)])),col="blue")
################################################################################
## Brasil
################################################################################
brasil.raw <- read.csv("brazil_wikipedia_timeseries.csv", as.is=TRUE)
## Converte para time series
brasil <- zoo(x=brasil.raw$casos.acumulados,
order.by = as.Date(brasil.raw$dia, "%d-%m-%Y"))
plot(brasil, log="x", type="p")
## Ultimos 7 dias
y <- brasil[(length(brasil)-dias.final+1):length(brasil)]
brasil.fim.fit <- glm(y~x, family=poisson)
## Coeficientes
coef(brasil.fim.fit)[2]
## tempos de duplicacao ##
## Atual
log(2)/coef(brasil.fim.fit)[2]
## Intervalos de confianca
## Fim
log(2)/confint(brasil.fim.fit)[2,]
################################################################################
## China
################################################################################
china.raw <- read.csv("https://covid.ourworldindata.org/data/full_data.csv", as.is=TRUE) %>%
filter(location=="China") %>%
select(date,total_cases)
china <- zoo(x = china.raw$total_cases,
order.by = as.Date(china.raw$date, "%Y-%m-%d"))
## Um plot da serie temporal completa
plot(china, type="p")
## Minimo de casos para o dia zero
china.n.casos <- 15
## N de dias para incluir no ajuste
dias.final <- 7
## Ajuste a partir do dia zero
china.dia.zero <- min(which(china>=china.n.casos, arr.ind=TRUE))
## Ajuste do modelo aos primeiros 7 dias a partir do dia zero
x <- 0:(dias.final-1)
y <- china[china.dia.zero:(china.dia.zero+dias.final-1)]
## Ajuste do modelo Poisson
## Inicio do periodo
china.inicio.fit <- glm(y~x, family=poisson)
## Ultimos 7 dias
y <- china[(length(china)-dias.final+1):length(china)]
china.fim.fit <- glm(y~x, family=poisson)
## Coeficientes
coef(china.inicio.fit)[2]
coef(china.fim.fit)[2]
## tempos de duplicacao ##
## Inicio
log(2)/coef(china.inicio.fit)[2]
## Atual
log(2)/coef(china.fim.fit)[2]
## Intervalos de confianca
## Inicio
log(2)/confint(china.inicio.fit)[2,]
## Fim
log(2)/confint(china.fim.fit)[2,]