nwasc:

#### Sub- Functions for AAMMAPs Arial and ___ ####
##
# Used after 'transect' csv is read in as df

# create a three col. lookup table with
#### Creat a lookup table indexed by transect_id, with new rev transect_id and Source_name
#### added a drop track flag for clarity
####
#### AMAPPS combine transects

Create.transect.revision.lookup <- function(
                                            .df = cts.dat,
                                            .target_id = target_id,
                                            source_name = "segmented_transect_id",
                                            orginal_id = "transect_id") { # missing transect_dist = 'transect_distance_nb'
  # Organize arguments
  # Make .df a dataframe so it plays nice.
  # .df = data.frame(.df)

  # Select row and columns !!!!!!transect dist subsituted for number of secs.
  Target_rows <- .df %>%
    filter(., dataset_id %in% .target_id) %>%
    select(dataset_id, start_dt, segmented_transect_id, transect_id, time_from_midnight_start, time_from_midnight_stop) %>%
    mutate(transect_dist = time_from_midnight_stop - time_from_midnight_start) %>%
    select(dataset_id, start_dt, source_name = segmented_transect_id, orginal_id = transect_id, transect_dist)


  # Create combined transect Name
  lookup_table <- Target_rows %>%
    mutate(rev_names = substring(source_name, 0, nchar(source_name) - 4)) %>%

    # comute and choose lngest segment
    select(rev_names, orginal_id, transect_dist) %>%
    arrange(rev_names, orginal_id, transect_dist) %>%
    group_by(rev_names) %>%
    mutate(
      max_dist = max(transect_dist),
      drop_track = max_dist != transect_dist,
      rev_ids = orginal_id[1]
    ) %>%
    ungroup() %>%
    select(orginal_id, rev_ids, rev_names, drop_track) %>%
    rename(transect_id = orginal_id)

  lookup_table
}

# lookup_table = Create.transect.revision.lookup()

#### Function to edit the transect file
roll_up_transect <- function(
                             .df = cts.dat,
                             lookup = lookup_table) {
  new_df <- .df %>%
    left_join(., lookup, by = "transect_id") %>% # Join the lookup to transect file
    mutate(
      drop_track = ifelse(is.na(drop_track), FALSE, drop_track),
      transect_id = ifelse(!is.na(rev_ids), rev_ids, transect_id), # is the !is.na needed?
      segmented_transect_id = ifelse(!is.na(rev_names), rev_names, segmented_transect_id)
    ) %>%
    filter(drop_track != TRUE) %>%
    select(-c(rev_ids, rev_names, drop_track))


  new_df
}

# transect2 = roll_up_transect(transect, lookup_table)

#### Function to edit track file (similiar to above)
roll_up_track <- function(
                          .df = shp.pre,
                          lookup = lookup_table) {
  new_df <- .df %>%
    left_join(., lookup, by = "transect_id") %>% # Join the lookup to transect file
    mutate(
      drop_track = ifelse(is.na(drop_track), FALSE, drop_track),
      transect_id = ifelse(!is.na(rev_ids), rev_ids, transect_id)
    ) %>%
    filter(drop_track != TRUE) %>%
    select(-c(rev_ids, rev_names, drop_track))


  #
  #
  new_df
}

# track2 = roll_up_track(track, lookup_table)

#### Function to edit observation table same as aove two except we do not drop rows
# kept. Observations simply reassigned .
roll_up_observation <- function(
                                .df = obs.pre,
                                lookup = lookup_table) {
  new_df <- .df %>%
    left_join(., lookup, by = "transect_id") %>% # Join the lookup to transect file
    mutate(
      transect_id = ifelse(!is.na(rev_ids), rev_ids, transect_id) # is the !is.na needed?
    ) %>%
    select(-c(rev_ids, rev_names, drop_track))


  new_df
}

# observation2 = roll_up_observation(observation, lookup_table)


# Example Run
AMAPPS.target.ids <- c(113, 118, 137, 138, 140, 141, 146, 147)
target_id <- AMAPPS.target.ids
lookup_table <- Create.transect.revision.lookup()
cts.dat <- roll_up_transect(cts.dat, lookup_table)
shp.pre <- roll_up_track(shp.pre, lookup_table)
obs.pre <- roll_up_observation(obs.pre, lookup_table)