tiledbsoma-write-nulls.r

#!/usr/bin/env Rscript

library(tiledbsoma)

uri <- 'tiledbsoma-write-nulls'

unlink(uri, recursive=TRUE)

schema = arrow::schema(
  #arrow::field("soma_joinid", arrow::int64(), nullable = FALSE),
  #arrow::field("mystring", arrow::large_utf8(), nullable = FALSE),
  #arrow::field("myint", arrow::float64(), nullable = FALSE)
  # Arrow seems to make this nullable = TRUE by default
  arrow::field("soma_joinid", arrow::int64()),
  arrow::field("mystring", arrow::large_utf8()),
  arrow::field("myenum", arrow::dictionary(index_type = arrow::int8(), value_type = arrow::utf8())),
  arrow::field("myint", arrow::int32()),
  arrow::field("myfloat", arrow::float64()),
)

tbl <- arrow::arrow_table(
  soma_joinid = bit64::as.integer64(c(0,1,2,3,4,5)),
  mystring = c('a', 'b', NA, 'd', 'e', 'f'),
  myenum = factor(c("B cell", NA, "T cell", "B cell", "T cell", NA)),
  myint = c(10,20,30,NA,50,60),
  myfloat = c(10,20,30,NA,50,60),
  schema=schema
)

sdf <- SOMADataFrameCreate(uri, schema)
sdf$write(tbl)
sdf$close()