splitting tagset utility

closes #60

utility/migrate-variablest-new-tagset.R

@@ -0,0 +1,97 @@
+# knitr::stitch_rmd(script="./manipulation/te-ellis.R", output="./stitched-output/manipulation/te-ellis.md") # dir.create("./stitched-output/manipulation/", recursive=T)
+# For a brief description of this file see the presentation at
+#   - slides: https://rawgit.com/wibeasley/RAnalysisSkeleton/master/documentation/time-and-effort-synthesis.html#/
+#   - code: https://github.com/wibeasley/RAnalysisSkeleton/blob/master/documentation/time-and-effort-synthesis.Rpres
+rm(list=ls(all=TRUE))  #Clear the variables from previous runs.
+
+# ---- load-sources ------------------------------------------------------------
+# Call `base::source()` on any repo file that defines functions needed below.  Ideally, no real operations are performed.
+
+# ---- load-packages -----------------------------------------------------------
+# Attach these package(s) so their functions don't need to be qualified: http://r-pkgs.had.co.nz/namespace.html#search-path
+library(magrittr            , quietly=TRUE)
+library(DBI                 , quietly=TRUE)
+
+# Verify these packages are available on the machine, but their functions need to be qualified: http://r-pkgs.had.co.nz/namespace.html#search-path
+requireNamespace("readr"        )
+requireNamespace("tidyr"        )
+requireNamespace("dplyr"        ) # Avoid attaching dplyr, b/c its function names conflict with a lot of packages (esp base, stats, and plyr).
+requireNamespace("testit"       ) # For asserting conditions meet expected patterns/conditions.
+requireNamespace("checkmate"    ) # For asserting conditions meet expected patterns/conditions. # remotes::install_github("mllg/checkmate")
+requireNamespace("RSQLite"      ) # Lightweight database for non-PHI data.
+# requireNamespace("RODBC"      ) # For communicating with SQL Server over a locally-configured DSN.  Uncomment if you use 'upload-to-db' chunk.
+requireNamespace("OuhscMunge"   ) # remotes::install_github(repo="OuhscBbmc/OuhscMunge")
+
+# ---- declare-globals ---------------------------------------------------------
+# Constant values that won't change.
+directory_in                    <- "data-public/metadata/tagsets-79"
+path_in_tagset_new              <- file.path(directory_in, "Gen2FatherFromGen1Death.NLSY79" )
+path_in_tagset_old              <- file.path(directory_in, "Gen2FatherFromGen1.NLSY79"      )
+path_out_tagset_old             <- path_in_tagset_old                     # In this case, overwrite the tagset. Make sure things are committed to Git first
+
+variables_to_duplicate          <- c("R0000100")                      # For Nlsy79 Gen1
+# variables_to_duplicate          <- c("C0000100", "C0000200")          # For Nlsy79 Gen2
+# variables_to_duplicate          <- c("R0000100")                      # For Nlsy97
+
+
+col_types <- readr::cols_only( # readr::spec_csv(path_in_oklahoma)
+  `id`          = readr::col_character()
+)
+
+# ---- load-data ---------------------------------------------------------------
+# Read the CSVs
+ds_new <- readr::read_csv(path_in_tagset_new   , col_types=col_types, col_names="id")
+ds_old <- readr::read_csv(path_in_tagset_old   , col_types=col_types, col_names="id")
+
+rm(path_in_tagset_new, path_in_tagset_old)
+
+ds_new
+ds_old
+
+# ---- tweak-data --------------------------------------------------------------
+# OuhscMunge::column_rename_headstart(ds) #Spit out columns to help write call ato `dplyr::rename()`.
+ds_to_duplicate <-
+  tibble::tibble(
+    id  = variables_to_duplicate
+  )
+
+row_count_new_start       <- nrow(ds_new)
+row_count_old_start       <- nrow(ds_old)
+row_count_duplicate       <- nrow(ds_to_duplicate)
+
+# ---- anti-join ----------------------------------------------------------
+ds_old <-
+  ds_old %>%
+  dplyr::anti_join(ds_new, by="id") %>%
+  dplyr::union_all(
+    ds_to_duplicate
+  ) %>%
+  dplyr::arrange(id)
+
+# ---- verify-values -----------------------------------------------------------
+# Sniff out problems
+# OuhscMunge::verify_value_headstart(ds_old)
+checkmate::assert_character(ds_new$id , any.missing=F , pattern="^[A-Z]\\d{7}$" , unique=T)
+checkmate::assert_character(ds_old$id , any.missing=F , pattern="^[A-Z]\\d{7}$" , unique=T)
+
+row_count_old_end         <- nrow(ds_old)
+row_count_old_expected    <- row_count_old_start - row_count_new_start + row_count_duplicate
+
+testit::assert(row_count_old_end == row_count_old_expected)
+
+# ---- specify-columns-to-upload -----------------------------------------------
+# dput(colnames(ds)) # Print colnames for line below.
+columns_to_write <- c(
+  "id"
+)
+ds_slim <-
+  ds_old %>%
+  dplyr::select_(.dots=columns_to_write) %>%
+  # dplyr::slice(1:100) %>%
+  dplyr::mutate_if(is.logical, as.integer)       # Some databases & drivers need 0/1 instead of FALSE/TRUE.
+ds_slim
+
+rm(columns_to_write)
+
+# ---- save-to-disk ------------------------------------------------------------
+readr::write_csv(ds_slim, path_out_tagset_old, col_names=F)