R/data-actions.R

In bagasbgy/quantrecipes: Extra recipes for quantitative analysis

# example dataset ---------------------------------------------------------

#' @title Example trading actions
#'
#' @details
#'
#'  This dataset provide trading actions data, which
#'  created using [btcusdt][btcusdt] data and [step_zigzag()][step_zigzag]
#'  for an example trading actions.
#'
#' @source the close prices are retrieved from [KuCoin](https://kucoin.com)
#'  market's API.
#'
#' @format An object of class `tibble`.
#'
#' @usage data(actions)
#'
#' @examples
#'
#' # import library
#' library(quantrecipes)
#'
#' # print the data
#' actions
#'
#' @docType data

"actions"

# script used to acquire the data -----------------------------------------

# # import libs
# library(quantrecipes)
#
# # get zigzag
# rec <- recipe(. ~ ., data = tail(btcusdt, 12 * 24 * 3)) %>%
#   step_zigzag(close, change = 0.5, state = TRUE, span = c(1, -1)) %>%
#   prep()
#
# # create example actions data
# actions <- juice(rec) %>%
#   mutate(portfolio = case_when(
#     zigzag_swing == "up" ~ "buy",
#     zigzag_swing == "down" ~ "sell"
#   ))
#
# first_action <- actions %>%
#   drop_na() %>%
#   pull(portfolio) %>%
#   first()
#
# if (first_action == "buy") {
#
#   first_notna <- min(which(!is.na(actions$portfolio)))
#
#   actions[first_notna, "portfolio"] <- "hold"
#
# }
#
# actions <- actions %>%
#   mutate(portfolio = c("buy", .data$portfolio[-1])) %>%
#   mutate(portfolio = ifelse(is.na(.data$portfolio), "hold", .data$portfolio))
#
# actions <- actions %>%
#   mutate(benchmark = "hold") %>%
#   mutate(benchmark = c("buy", .data$benchmark[-1]))
#
# actions <- actions %>%
#   select(datetime, close, benchmark, portfolio)
#
# # save the data
# usethis::use_data(actions, overwrite = TRUE)