import_operation_haridwar.R

#' Helper function: store MySQL database credentials in ".my.cnf"
#'
#' @param dbname name of database to be accessed
#' @param group name of group
#' @param dir path to the MySQL configuration file (default: file.path(getwd(),
#' ".my.cnf"))
#' @param host hostname of MySQL database (default: NULL)
#' @param user username of MySQL database (default: NULL)
#' @param password password of MySQL database (default: NULL)
#' @param ... further arguments passed to dplyr::src_mysql()
#' @return sets dplyr::src_mysql to work with MySQL config file
#' @importFrom dplyr src_mysql
#' @importFrom dbplyr src_dbi
#' @keywords internal
src_mysql_from_cnf <- function(dbname,
                               group = NULL,
                               dir=file.path(getwd(), ".my.cnf"),
                               host=NULL,
                               user=NULL,
                               password=NULL,
                               ...) {
  dir <- normalizePath(dir)
  if (!(file.exists(dir))) {
    stop(sprintf("No such file '%s'", dir))
  }
  dplyr::src_mysql(
    dbname,
    group = group,
    default.file = dir,
    # explicitly passing null unless otherwise specified.
    host = host,
    user = user,
    password = password,
    ...
  )
}

#' Imports operational data
#'
#' @param mysql_conf path to the MySQL configuration file
#' @return returns data frame operational data from MySQL db
#' @import dplyr
#' @export
import_operation <- function(mysql_conf = file.path(getwd(), ".my.cnf")) {
  sumewa <- src_mysql_from_cnf(
    dbname = "sumewa",
    group = "autarcon",
    dir = mysql_conf
  )


  tbl_live <- dplyr::tbl(sumewa, "live") %>%
    dplyr::filter_(~AnlagenID == 4014) %>%
    dplyr::tbl_df() %>%
    dplyr::rename_("Redox_Out1" = "Red1", "Redox_Out2" = "Red2") %>%
    dplyr::tbl_df()


  tbl_india <- dplyr::tbl(sumewa, "indienmesssystem") %>%
    dplyr::filter_(~AnlagenID == 4013) %>%
    dplyr::select_(
      ~id,
      ~AnlagenID,
      ~time,
      ~Red2,
      ~Flux,
      ~Uz,
      ~Up,
      ~Iz,
      ~Anlauf,
      ~modus,
      ~errcode,
      ~err,
      ~ext,
      ~maintenance,
      ~Filterbetrieb,
      ~Filterpumpe,
      ~BB,
      ~CC,
      ~err_sdcard,
      ~err_modem
    ) %>%
    dplyr::tbl_df() %>%
    dplyr::rename_(
      "Redox_In" = "Red2",
      "Pressure1" = "Flux",
      "Pressure2" = "Uz",
      "DiffPressure" = "Iz",
      "H20Head" = "Up"
    ) %>%
    dplyr::tbl_df()



  tbl_tmp <- plyr::rbind.fill(tbl_live, tbl_india)


  duplicated_datetimes <- names(which(table(tbl_tmp$time) != 1))

  operation <- tbl_tmp[!tbl_tmp$time %in% duplicated_datetimes, ] %>%
    left_join(data.frame(
      AnlagenID = c(4013, 4014),
      LocationName = rep("Haridwar", 2)
    )) %>%
    dplyr::rename_("DateTime" = "time") %>%
    dplyr::mutate_("DateTime" = "as.POSIXct(DateTime,tz = 'UTC')")

  return(operation)
}



if (FALSE) {
  operation <- import_operation()

  ### Calculate additional parameters:
  operation <- operation %>%
    dplyr::mutate_(
      Redox_Out = "(Redox_Out1+Redox_Out2)/2",
      Redox_Diff = "Redox_Out - Redox_In",
      Power_pump = "Up*Ip",
      Power_cell = "Uz*Iz",
      Pump_WhPerCbm = "Power_pump/Flux/1000",
      Cell_WhPerCbm = "Power_cell/Flux/1000"
    )



  ### Aggregation of online data to user defined
  drop.cols <- "LocationName"
  operation_grouped <- kwb.base::hsGroupByInterval(
    data = operation %>%
      dplyr::select_(.dots = setdiff(names(.), drop.cols)),
    interval = 60 * 15,
    tsField = "DateTime",
    FUN = "median"
  )

  ### Plot it
  pdf(file = "report/datenanalyse_haridwar.pdf", width = 7, height = 5)

  operation_grouped_tidy1 <- tidyr::gather(
    operation_grouped[, c("DateTime", "Redox_Out1", "Redox_Out2", "Redox_In")],
    key = "Parameter",
    value = "Value", -DateTime
  )

  p1 <- ggplot(data = operation_grouped_tidy1, aes(x = DateTime, y = Value, col = Parameter)) +
    geom_point() +
    labs(list(x = "Datetime (UTC)", y = "Redox potential (mV)")) +
    theme_bw() +
    theme(legend.position = "top")
  print(p1)


  operation_grouped_tidy2 <- tidyr::gather(
    operation_grouped[, c("DateTime", "Redox_Out", "Redox_In")],
    key = "Parameter",
    value = "Value", -DateTime
  )

  p2 <- ggplot(data = operation_grouped_tidy2, aes(x = DateTime, y = Value, col = Parameter)) +
    geom_point() +
    labs(list(x = "Datetime (UTC)", y = "Redox potential (mV)")) +
    theme_bw() +
    theme(legend.position = "top")
  print(p2)

  p3 <- ggplot(data = operation %>% select(DateTime, Redox_Diff), aes(x = DateTime, y = Redox_Diff)) +
    geom_point() +
    # geom_vline()
    labs(list(
      x = "Datetime (UTC)",
      y = "Redox potential difference: cell outflow - inflow (mV)"
    )) +
    theme_bw() +
    theme(legend.position = "top")
  print(p3)


  backwash <- operation[operation$Anlauf == 90, "DateTime"]

  p4 <- ggplot(data = operation_grouped %>% filter(DiffPressure < 10), aes(x = DateTime, y = DiffPressure)) +
    geom_point() +
    geom_vline(xintercept = as.numeric(backwash), col = "red") +
    labs(list(
      x = "Datetime (UTC)",
      y = "Pressure difference (out - in)"
    )) +
    theme_bw() +
    theme(legend.position = "top")
  print(p4)


  energy_tidy <- operation %>% select(DateTime, Pump_WhPerCbm, Cell_WhPerCbm) %>% gather(key = "Key", value = "Value", -DateTime)
  energy_tidy <- tidyr::separate(
    energy_tidy,
    col = "Key",
    into = c("System component", "Unit"),
    sep = "_"
  )

  energy_title <- sprintf(
    "Based on 15 minute median values of online data\n(period: %s to %s)",
    min(energy_tidy$DateTime),
    max(energy_tidy$DateTime)
  )

  p5 <- ggplot(energy_tidy, aes_string(
    x = "DateTime",
    y = "Value",
    col = "`System component`"
  )) +
    geom_point() +
    geom_vline(xintercept = as.numeric(backwash), col = "red") +
    # facet_wrap(~ `System component`) +
    labs(list(
      x = "Datetime (UTC)",
      y = "Specific energy demand (Wh/m3)",
      title = energy_title
    )) +
    theme_bw() +
    theme(legend.position = "top")
  print(p5)

  p6 <- ggplot(energy_tidy, aes_string(
    x = "`System component`",
    y = "Value",
    col = "`System component`"
  )) +
    geom_jitter(height = 0, width = 0.3, alpha = 0.5) +
    # facet_wrap(~ `System component`) +
    labs(list(
      x = "System component",
      y = "Specific energy demand (Wh/m3)",
      title = energy_title
    )) +
    theme_bw() +
    theme(legend.position = "top")
  print(p6)

  dev.off()
}