Skip to contents

Join all data frames in an environment with optional late joins

Source: R/storage.R
join_env.Rd

Performs a dplyr::full_join() across all objects in env (in alphabetical order), excluding any listed in late. Late objects are then joined sequentially (via dplyr::left_join() by default) in the order given. Full-join clashes use suffixes c(".x", ".y"); late joins use c(".early", ".late").

Usage

join_env(
  env,
  by = "date",
  late = NULL,
  join_precedence = NULL,
  coalesce_suffixed = NULL,
  late_join = dplyr::left_join
)

Arguments

env

Environment containing only data frames (incl. tibbles) to join.

by

Character vector of join keys (passed to dplyr::full_join()).

late

Character vector of object names in env that should be excluded from the initial full join and instead left-joined afterwards. Defaults to NULL.

join_precedence

Optional character vector of length 2 giving join suffixes to coalesce (for example, c(".early", ".late") or c(".late", ".early")). When non-NULL, any pairs of columns whose names end in these suffixes (and share the same base name) are replaced by a single unsuffixed column containing the coalesced values, preferring the left suffix when both values are available. If NULL (the default), no automatic coalescing is performed.

coalesce_suffixed

Deprecated; use join_precedence (same meaning).

late_join

Function to use for joining late objects, e.g. dplyr::left_join (the default). Must accept dplyr-style suffix and by arguments.

Value

A tibble: the full join of all non-late objects, followed by sequential left-joins (or other joins specified by late_join) of the late objects. If join_precedence is supplied, suffixed join columns are coalesced into unsuffixed base columns as described above.

Details

Optionally, column pairs with specified suffixes can be coalesced into unsuffixed base columns via join_precedence.

See also

Other fund/index workflow functions: add_data_loader(), build_all_series(), clear_data_loaders(), clear_storage(), get_storage(), load_fund(), run_data_loaders(), store_timeseries()

Examples

  e <- new.env()
  e$members <- dplyr::band_members
  e$instruments <- dplyr::band_instruments
  e$other_instr <- dplyr::band_instruments |>
      dplyr::mutate(plays = dplyr::case_match(name,
                                              "John" ~ "banjo",
                                              "Paul" ~ "mellotron",
                                              "Keith" ~ "harpsichord")) |>
      dplyr::add_row(name = "Mick", plays = "harmonica") |>
      dplyr::add_row(name = "Stu", plays = "piano")

  full <- join_env(e, by = "name")
  late <- join_env(e, by = "name", late = "other_instr")
  late_coalesced <- join_env(e,
                             by = "name",
                             late = "other_instr",
                             join_precedence = c(".early", ".late"))
  print(list(full = full, late = late, late_coalesced = late_coalesced))
#> $full
#> # A tibble: 5 × 4
#>   name  plays.x band    plays.y    
#>   <chr> <chr>   <chr>   <chr>      
#> 1 John  guitar  Beatles banjo      
#> 2 Paul  bass    Beatles mellotron  
#> 3 Keith guitar  NA      harpsichord
#> 4 Mick  NA      Stones  harmonica  
#> 5 Stu   NA      NA      piano      
#> 
#> $late
#> # A tibble: 4 × 4
#>   name  plays.early band    plays.late 
#>   <chr> <chr>       <chr>   <chr>      
#> 1 John  guitar      Beatles banjo      
#> 2 Paul  bass        Beatles mellotron  
#> 3 Keith guitar      NA      harpsichord
#> 4 Mick  NA          Stones  harmonica  
#> 
#> $late_coalesced
#> # A tibble: 4 × 3
#>   name  band    plays    
#>   <chr> <chr>   <chr>    
#> 1 John  Beatles guitar   
#> 2 Paul  Beatles bass     
#> 3 Keith NA      guitar   
#> 4 Mick  Stones  harmonica
#>