R/conv_data.R

In clockplot: Plot Event Times on a 24-Hour Clock

# Convert values to 0.5 to 1

conv_half_1 <- function(x) {
  rx <- max(x) - min(x)
  incrx <- 0.5 / rx
  y <- 0.5 + (x - min(x)) * incrx
  return(y)
}

# Convert data for simplest clock chart
# This will be used to convert data for showing times on clock.
# Length of hands of clock is constant

conv_data <- function(data, time) {
  # Data Preparation
  dt <- data %>%
    mutate(time = hms::parse_hm({{ time }})) %>%
    tidyr::separate_wider_delim(
      cols = time,
      names = c("hr", "mn", "sc"),
      cols_remove = FALSE,
      delim = ":"
    ) %>% # Separate minute
    dplyr::mutate(
      mn = as.numeric(ifelse(is.na(.data$mn), 0, .data$mn)),
      mn = .data$mn / 60,
      # minute = ifelse(.data$minute<10, .data$minute * 5/30, .data$minute * 5/300),
      hr = as.numeric(.data$hr),
      timc = .data$hr + .data$mn,
      time_angle = ifelse(0 <= .data$timc & .data$timc <= 6,
        (6 - .data$timc) * pi / 12,
        (30 - .data$timc) * pi / 12
      ),
      x1 = cos(.data$time_angle) * 0.97,
      y1 = sin(.data$time_angle) * 0.97,
      x0 = rep(0, dim(data)[1]),
      y0 = rep(0, dim(data)[1])
    )
  return(dt)
}

#  Convert data for coloring based on criteria, without modifying length.

conv_data_col <- function(data, time, colby) {
  # Data Preparation
  dt <- data %>%
    mutate(time = hms::parse_hm({{ time }})) %>%
    tidyr::separate_wider_delim(
      cols = time,
      names = c("hr", "mn", "sc"),
      cols_remove = FALSE,
      delim = ":"
    ) %>% # Separate minute
    mutate(
      mn = as.numeric(ifelse(is.na(.data$mn), 0, .data$mn)),
      mn = .data$mn / 60,
      # minute = ifelse(.data$minute<10, .data$minute * 5/30, .data$minute * 5/300),
      hr = as.numeric(.data$hr),
      timc = .data$hr + .data$mn, # Ignore second
      time_angle = ifelse(0 <= .data$timc & .data$timc <= 6,
        (6 - .data$timc) * pi / 12,
        (30 - .data$timc) * pi / 12
      ),
      x1 = cos(.data$time_angle) * 0.95,
      y1 = sin(.data$time_angle) * 0.95,
      x0 = rep(0, dim(data)[1]),
      y0 = rep(0, dim(data)[1]),
      colby = {{ colby }}
    )
  return(dt)
}


# Change length of hands by a quantitative vector, optionally select a
# color for the hands
conv_data_len <- function(data, time, len) {
  len <- dplyr::pull(data, {{ len }})
  # Normalize length. This will be replace by rn
  len <- (len - min(len)) / (max(len) - min(len))
  len[which(len == 0)] <- (
    min(len) + min(len[len != min(len)])) / 2
  len[which(len == 1)] <- (
    max(len) + max(len[len != max(len)])) / 2
  rn <- conv_half_1(len)

  # Data Preparation
  dt <- data %>%
    dplyr::mutate(
      len = rn,
      time = hms::parse_hm({{ time }})
    ) %>%
    tidyr::separate_wider_delim(
      cols = time,
      names = c("hr", "mn", "sc"),
      cols_remove = FALSE,
      delim = ":"
    ) %>% # Separate minute
    dplyr::mutate(
      mn = as.numeric(ifelse(is.na(.data$mn), 0, .data$mn)),
      mn = .data$mn / 60,
      # minute = ifelse(.data$minute<10, .data$minute * 5/30, .data$minute * 5/300),
      hr = as.numeric(.data$hr),
      timc = .data$hr + .data$mn,
      time_angle = ifelse(0 <= .data$timc & .data$timc <= 6,
        (6 - .data$timc) * pi / 12,
        (30 - .data$timc) * pi / 12
      ),
      x1 = cos(.data$time_angle) * rn * 0.95,
      y1 = sin(.data$time_angle) * rn * 0.95,
      x0 = rep(0, dim(data)[1]),
      y0 = rep(0, dim(data)[1])
    )
  return(dt)
}


# Keep factor original order

keep_fct_order <- function(x) factor(x, levels = unique(x))

## Legend lines
# ---
# -----
# --------

lgl <- data.frame(
  x0 = rep(1.1, 3),
  x1 = c(1.2, 1.3, 1.4),
  y0 = c(-0.1, 0, 0.1),
  y1 = c(-0.1, 0, 0.1)
)

# geom_segment(data = lgl, mapping = aes(x=x0, y = y0, xend = x1, yend = y1))