Extract_KW51_bridge_data_Maes.Lombaerts_2021.qmd

---
title: "Covariate-Adjusted Functional Data Analysis for Structural Health Monitoring"
subtitle: "Download, extract and prepare bridge KW51 data: @Maes.Lombaert_2020 and @Maes.Lombaert_2021"
    
# Freeze computed outputs
freeze: true

# Enable banner style title blocks
title-block-banner: true

# Enable CC licence appendix
license: "CC BY"

# Default for table of contents
toc: true
toc-title: Table of contents
toc-location: left

# Default knitr options
execute:
  echo: true
  message: true
  warning: false
  cache: false

date: "10/07/2024" 
format:
  html:
    embed-resources: false
    code-fold: false
    code-summary: "Code"
    code-tools:
      source: true
      toggle: true
      caption: none

knitr:
  opts_chunk:
    dev:
      - png
      - pdf
      
editor: source

author:
  - name: Philipp Wittenberg
    corresponding: true
    id: pw
    orcid: 0000-0001-7151-8243
    email: pwitten@hsu-hh.de
    affiliation: 
      - name: Helmut Schmidt University
        city: Hamburg
        country: Germany
        url: www.hsu-hh.de
  - name: Lizzie Neumann
    id: ln
    orcid: 0000-0003-2256-1127
    email: neumannl@hsu-hh.de
    affiliation: 
      - name: Helmut Schmidt University
        city: Hamburg
        country: Germany
        url: www.hsu-hh.de
  - name: Alexander Mendler
    id: ln
    orcid: 0000-0002-7492-6194
    email: alexander.mendler@tum.de
    affiliation: 
      - name: Technical University of Munich
        city: Hamburg
        country: Germany
        url: www.hsu-hh.de
  - name: Jan Gertheiss
    id: jg
    orcid: 0000-0001-6777-4746
    email: gertheij@hsu-hh.de
    affiliation: 
      - name: Helmut Schmidt University
        city: Hamburg
        country: Germany
        url: www.hsu-hh.de        
        
        

citation: 
  container-title: arXiv:2408.02106
  doi: 10.48550/arXiv.2408.02106
  
bibliography:
  - literature.bib
link-citations: true

output: github_document
---

# Load libraries

```{r, message=FALSE}
# Package names
packages <- c("dplyr", "lubridate", "R.matlab", "tidyr", "tidyselect", "utils")

# Install packages not yet installed
installed_packages <- packages %in% rownames(installed.packages())
if (any(installed_packages == FALSE)) {
  install.packages(packages[!installed_packages])
}

# Packages loading
invisible(lapply(packages, library, character.only = TRUE))
```

# Data preprocessing

## Raw data import

```{r raw_matlab_data_import, eval=TRUE}
url <- "https://zenodo.org/records/3745914/files/trackedmodes.zip?download=1"
path1 <- tempfile(fileext = ".zip")
if (file.exists(path1))  'file alredy exists' else download.file(url, path1, mode="wb")
unzip(zipfile = path1,exdir = tempdir())
list.import.mat <- R.matlab::readMat(paste0(tempdir(),"/trackedmodes/trackedmodes.mat"))
```

## Extract environmental data

```{r environmental_data, eval=TRUE}
df1 <- list.import.mat$modes[7][[1]] |>
  data.frame() |>
  `colnames<-`(list.import.mat$modes[6][[1]] |> unlist()) |>
  mutate(time=c(list.import.mat$modes[1][[1]])) |>
  mutate(time=as.Date(time, tz="UTC", origin="0000-01-01")) |>
  mutate(time=seq(
    as.POSIXlt(paste0(date(first(time)), " 0:00:00"), tz="UTC"),
    as.POSIXlt(paste0(date(last(time)), " 23:00:00"), tz="UTC"), by = "hour")) |>
  mutate(tempdate=format(time, "%Y-%m-%d")) |>
  group_by(tempdate) |>
  mutate(ind_day=cur_group_id()) |>
  ungroup() |>
  dplyr::select(-(tempdate)) |>
  rename(date=time)
```

## Extract Modal data

```{r modal_data, eval=TRUE}
df2 <- list.import.mat$modes[2][[1]] |>
  data.frame() |>
  'colnames<-' (sprintf("Mode_%02d", 1:14)) |>
  mutate(time=c(list.import.mat$modes[1][[1]])) |>
  mutate(time=as.Date(time, tz="UTC", origin="0000-01-01")) |>
  mutate(time=seq(
    as.POSIXlt(paste0(date(first(time)), " 0:00:00"), tz="UTC"),
    as.POSIXlt(paste0(date(last(time)), " 23:00:00"), tz="UTC"), by = "hour")) |>
  mutate(tempdate=format(time, "%Y-%m-%d")) |>
  group_by(tempdate) |>
  mutate(ind_day=cur_group_id()) |>
  ungroup() |>
  dplyr::select(-(tempdate)) |>
  rename(date=time)
```

## Join data

```{r join_data, eval=TRUE}
df3 <- full_join(df1, df2, by=c("date", "ind_day")) |>
  mutate(intervention=if_else(date<="2019-05-15 0:00:00", 0,
                              if_else(date>="2019-09-27 0:00:00", 2, 1))) |>
  mutate_all(~replace(., is.nan(.), NA))
```

## Create hourly data time series

```{r hourly_times_series, eval=TRUE}
df_hour <- data.frame(date=seq(
    ymd_hms(paste0(date(first(df3$date)), " 0:00:00"), tz="UTC"),
    ymd_hms(paste0(date(last(df3$date)), " 23:59:59"), tz="UTC"), by="hour")) |>
  mutate(tempdate=format(date, "%Y-%m-%d")) |>
  group_by(tempdate) |>
  mutate(ind_hour=row_number()) |>
  ungroup() |>
  dplyr::select(-(tempdate)) |>
  as_tibble()
```

## Join and save data sets

```{r join_save_data, eval=TRUE}
df4 <- left_join(df_hour, df3, by="date") |>
  mutate(td01=ind_hour/24)

saveRDS(df4, "data/Dataset_bridge_KW51.RDS")
```