R/prep_m_to_m_data.R

In randrescastaneda/joyn: Tool for Diagnosis of Tables Joins and Complementary Join Features

# this needs work


# prep_m_to_m_data <- function(x, y, by){
#
#   # 1) re-order using `by` ----
#   x <- x |>
#     roworderv(
#       by
#     )
#   y <- y |>
#     roworderv(
#       by
#     )
#
#   # 2) n_key ----
#   x$n_key <- ave(
#     x$key_column,
#     x$key_column,
#     FUN = seq_along
#   )
#   y$n_key <- ave(
#     y$key_column,
#     y$key_column,
#     FUN = seq_along
#   )
#
#   # 3) Expansion ----
#   # Compute the size of each group in both data frames
#   x_size <- table(
#     x[[by]]
#   )
#   y_size <- table(
#     y[[by]]
#   )
#   # Expand x and y data frames
#   x_expanded <- x[
#     rep(
#       seq_len(
#         nrow(x)
#       ),
#       y_size[
#         x[[by]]
#       ]
#     ),
#   ]
#   y_expanded <- y[
#     rep(
#       seq_len(
#         nrow(y)
#       ),
#       x_size[
#         y[[by]]
#       ]
#     ),
#   ]
#
#   # 4 and 5) Group `by` then create new count ----
#   x_expanded$m_key <- ave(
#     x_expanded$n_key,
#     x_expanded$n_key,
#     FUN = seq_along
#   )
#   y_expanded$m_key <- ave(
#     y_expanded$n_key,
#     y_expanded$n_key,
#     FUN = seq_along
#   )
#
#   # 6 and 7) Create new m:m key ----
#   x_expanded <- x_expanded |>
#     ftransform(
#       key_m_to_m = paste(
#         n_original_rownumber,
#         m_gp_rownumber,
#         sep = "_"
#       )
#     )
#   y_expanded <- y_expanded |>
#     ftransform(
#       key_m_to_m = paste(
#         m_gp_rownumber,
#         n_original_rownumber,
#         sep = "_"
#       )
#     )
#
#   # Clean final prepped data
#   rownames(x_expanded) <- NULL
#   rownames(y_expanded) <- NULL
#   x_expanded <- x_expanded |>
#     fselect(
#       - c(
#         n_original_rownumber,
#         m_gp_rownumber
#       )
#     )
#   y_expanded <- y_expanded |>
#     fselect(
#       - c(
#         n_original_rownumber,
#         m_gp_rownumber
#       )
#     )
#   # Return list ----
#   return(
#     list(
#       x = x_expanded,
#       y = y_expanded
#     )
#   )
#
# }
#
#
#
# # Simulate and check
# # Set seed for reproducibility
# set.seed(124)
#
# # Create dataset x
# x <- data.frame(
#   key_column = sample(LETTERS[1:2], 7, replace = TRUE),
#   data_x = rnorm(7)
# )
#
# # Create dataset y
# y <- data.frame(
#   key_column = sample(LETTERS[1:2], 7, replace = TRUE),
#   data_y = rnorm(7)
# )
#
# # Print the datasets
# print(x)
# print(y)
#
# by = "key_column"
#
# result <- prep_m_to_m_data(x, y, "key_column")
#
# result
# # Print the results
# print(x)
# print(y)
# print(result$x)
# print(result$y)