Seasonality chart with ggplot2

Question

I have an example graph that shows the seasonality of a given per month over the years. I would like to adapt the chart to days of the week, I have the example code, but I’m not able to adapt it.

The idea is to check if there is seasonality in the number of new cases on weekdays. For example, if there are one day or days that have more confirmed cases and the other hypothesis is to check whether on weekends there is a decrease in that number.

Sample chart:

Code used:

devData <- subset(df, Series=="Deviation", select=c(Series, Year, Month, Values))
meanDevData <- subset(df, Series=="Mean Deviation", select=c(Series, Year, Month, Values))
medDevData <- subset(df, Series=="Median Deviation", select=c(Series, Year, Month, Values))

ggplot(df,aes(Year,Values,colour=Series)) +
  geom_point(data=devData,size=I(2),alpha=I(0.6)) + 
  geom_line(data=meanDevData,size=I(1.5),alpha=I(0.6)) + 
  geom_line(data=medDevData,size=I(1.5),alpha=I(0.4)) + 
  theme_grey(base_size=15) +
  theme(legend.title = element_blank(), legend.position=c(.15,.9), axis.title.y=element_blank(),axis.text.x=element_blank()) + 
  ggtitle("UKRPI Additive Deviation") + facet_grid(. ~ Month) + 
  xlab(paste("Years:",min(df$Year),"to", max(df$Year)))

Website link with the post: https://www.clarusft.com/exploring-seasonality-in-a-time-series-with-rs-ggplot2/

Database I’m using: https://covid.saude.gov.br/

Remembering that I do not have the sample database to do tests. If you can help, I really appreciate.

Answer 1

First of all, the data must be prepared with casosNovos == 0 and with aggregation functions to calculate averages and medians.

The function to be used to read the file is read.csv2 once the decimals are separated by commas.

df <- read.csv2("arquivo_geral_covid.csv")

df$data <- as.Date(df$data)
df$Dia <- format(df$data, format = "%u")
df$Dia <- factor(df$Dia, labels = c("seg", "ter", "qua", "qui", "sex", "sab", "dom"))
dados <- subset(df, casosNovos > 0)
dados_agg <- aggregate(casosNovos ~ Dia, dados, function(x) c(Mean = mean(x), Median = median(x)))
dados_agg <- cbind(dados_agg[1], dados_agg[[2]])

Now is to use the clean data dados for the points and the aggregated basis dados_agg for horizontal lines. The scale of the axis of y is logarithmic, to change this just comment scale_y_continuous.

ggplot(dados, aes(data, casosNovos, group = Dia)) +
  geom_point(alpha = 0.3) + 
  geom_hline(data = dados_agg, 
             aes(yintercept = Mean, group = Dia),
             color = "green") + 
  geom_hline(data = dados_agg, 
             aes(yintercept = Median, group = Dia),
             color = "blue") + 
  scale_y_continuous(trans = "log10") +
  theme_grey(base_size = 15) +
  theme(legend.title = element_blank(), 
        axis.title.y = element_blank(), 
        axis.text.x = element_blank()) + 
  ggtitle("CoVID-19 - Novos casos por semana") + 
  facet_wrap( ~ Dia) + 
  xlab("Dia")