MJacobs1985

#### IMPORT LIBRARIES ####
rm(list = ls())
library(brms)
library(dplyr) 
library(ggplot2)
library(rethinking)
library(dmetar)
library(meta)
library(tidybayes)
library(ggridges)

MJacobs1985

rm(list = ls())
require(dplyr)
require(ggplot2)
require(ISOweek)
require(scales)
library(readr)
library(timetk)
rivm.data <- read_csv("COVID-19_casus_landelijk_2020-07-01.csv")
rivm.death <- rivm.data %>%
  dplyr::filter(Deceased == "Yes") ## Extract deaths data only

MJacobs1985

library(readr)
library(ggplot2)
library(gganimate)
library(tidyverse)
library(ggrepel)
library(lubridate)
library(tidymodels)
library(modeltime)
library(tsbox)
library(TSstudio)

MJacobs1985

rm(list = ls())
library(tidyverse)
library(readr)
library(ggthemes)
library(lubridate)
library(zoo)
library(gganimate)
library(cowplot)

getwd()

MJacobs1985

plot(ts(mymts[,1]))
# DWT with level of 2^3
data_dwt <- repr_dwt(mymts[,1], level = 6); length(data_dwt)
# first 84 DFT coefficients are extracted and then inverted
data_dft <- repr_dft(mymts[,1], coef = 63); length(data_dft)
# first 84 DCT coefficients are extracted and then inverted
data_dct <- repr_dct(mymts[,1], coef = 63); length(data_dct)
# Classical PAA
data_paa <- repr_paa(mymts[,1], q = 65, func = mean); length(data_paa)

MJacobs1985

mymts<-dataset%>%
  dplyr::select(VO2i_methode1,DO2i_methode1,RQ_methode1)%>%
  ts(., frequency=31557600)
plot(ts(mymts))
mod.foreca <- foreca(mymts, n.comp = 3, plot = TRUE, spectrum.control=list(method="pspectrum"))
mod.foreca
plot(mod.foreca)
biplot(mod.foreca)
summary(mod.foreca)
mod.foreca$scores <- ts(mod.foreca$scores, start = start(mymts),freq = frequency(mymts))

MJacobs1985

nums <- unlist(lapply(dataset, is.numeric))  
datasetnums<-dataset[ , nums]
corr <- round(cor(datasetnums), 1)
ggcorrplot(corr, hc.order = TRUE, 
           type = "lower", 
           lab = TRUE, 
           lab_size = 3, 
           title="Correlogram of dataset", 
           ggtheme=theme_bw)

MJacobs1985

wide<-dataset%>%dplyr::select(ID,Meting,Hb)%>%spread(Meting, Hb)
wide<-as.data.frame(wide)
wide_cld <- kml::cld(traj = wide, 
                     idAll = wide[,1], 
                     timeInData = 2:125, 
                     maxNA = 2)
class(wide_cld)
wide_klm<-kml::kml(wide_cld, 
nbRedrawing = 50,
toPlot='traj')

MJacobs1985

nums <- unlist(lapply(dataset, is.numeric))  
datasetnums<-dataset[ , nums]
TSdist::EuclideanDistance(dataset$VCO2i, dataset$Hb)
diss <- TSdist::TSDatabaseDistances(datasetnums, distance="euclidean", diag=T, upper=T) %>% 
  as.matrix()
diss %>% 
  reshape::melt() %>% 
  ggplot2::ggplot()+
  ggplot2::geom_tile(ggplot2::aes(x = X1, y = X2, fill = value))+
  ggplot2::scale_fill_viridis_c()

MJacobs1985

try <- dataset %>%
  dplyr::select(VO2cdi, Meting, ID) %>%
  drop_na() %>% # must drop NAs for clustering to work
  glimpse()
spread_try <- try %>%
  spread(ID,VO2cdi)  %>%
  glimpse()
VO2_cluster <- t(spread_try[-1])
VO2_cluster_dist <- dist(VO2_cluster , method="euclidean")  
fit <- hclust(VO2_cluster_dist, method="ward.D")