Numeric NA values read in as -2147483648 on Windows

Operating system

Windows 10

Description of the problem

When reading in data, numeric variables that have NA values are not read in as NA but are instead read in as -2147483648

Expected behavior

Columns are numeric, with missing values coded as NA

Actual behavior

No errors, warnings, or messages. Columns are numeric, but missing values are coded as -2147483648

Output of `sessionInfo()`

R version 4.1.2 (2021-11-01) Platform: x86_64-w64-mingw32/x64 (64-bit) Running under: Windows 10 x64 (build 19045)

Matrix products: default

locale: [1] LC_COLLATE=English_United States.1252 LC_CTYPE=English_United States.1252
[3] LC_MONETARY=English_United States.1252 LC_NUMERIC=C
[5] LC_TIME=English_United States.1252

attached base packages: [1] stats graphics grDevices utils datasets methods base

other attached packages: [1] synapser_1.0.0

loaded via a namespace (and not attached): [1] Rcpp_1.0.9 pillar_1.8.1 compiler_4.1.2 plyr_1.8.8 tools_4.1.2 digest_0.6.30
[7] jsonlite_1.8.4 evaluate_0.20 lifecycle_1.0.3 tibble_3.1.8 lattice_0.20-45 pkgconfig_2.0.3 [13] png_0.1-8 rlang_1.1.1 Matrix_1.5-3 cli_3.4.1 rstudioapi_0.14 yaml_2.3.7
[19] xfun_0.37 fastmap_1.1.0 withr_2.5.0 dplyr_1.1.0 knitr_1.42 generics_0.1.3
[25] vctrs_0.6.3 rprojroot_2.0.3 grid_4.1.2 tidyselect_1.2.0 reticulate_1.28 glue_1.6.2
[31] here_1.0.1 R6_2.5.1 fansi_1.0.4 rmarkdown_2.23 magrittr_2.0.3 codetools_0.2-18 [37] htmltools_0.5.3 utf8_1.2.3 rjson_0.2.21

Test code from meeting with @thomasyu888 today

# pull code other than as.data.frame()
test1 <- synTableQuery(sprintf("SELECT * FROM %s", paste0("syn21446700", ".", synGet("syn21446700")$properties$versionNumber - 1)))

View(as.data.frame(test1))

View(test1$asDataFrame())

# look at where file was temporarily saved
test1$filepath

# schema is NULL
test1$schema

# hybrid_death_int is recorded as an integer
test1$headers

# CsvFileTable is correct
class(test1)

# with read.csv it's read in correctly
test2 <- read.csv("c:/users/laveryj/.synapsecache/284/129372284/synapse_table_query_129372284.csv")

# with pandas read csv it's also read in correctly
library(reticulate)
pd <- reticulate::import("pandas")
test3 <- pd$read_csv("c:/users/laveryj/.synapsecache/284/129372284/synapse_table_query_129372284.csv")

# this is the problem - for some reason there is an issue converting to df
test1$asDataFrame() %>% select(contains("hybrid_death")) %>% view()

# no issues - numeric correctly NA
test4 <- read.csv("c:/users/laveryj/.synapsecache/284/129372284/synapse_table_query_129372284.csv",
                  encoding = "UTF-8",
                  stringsAsFactors = FALSE,
                  check.names = FALSE,
                  na.strings = c(""),
                  colClasses = "character")

test4 %>% select(contains("hybrid_death")) %>% view()

# these have correct NA
as.integer(test4$hybrid_death_int)
as.numeric(test4$hybrid_death_int)

# reproduces the error
test5 <- synapser::as.data.frame(test1)

test5 %>% select(contains("hybrid_death")) %>% view()

# this doesn't replicate the error, NAs are correct
# sourced functions from: https://github.com/Sage-Bionetworks/synapser/blob/8b2d310ca11c56d2514fb8d719edeb2a1cecc475/R/table.R#L120-L131
test6 <- .readCsvBasedOnSchema(test1)

test6 %>% select(contains("hybrid_death")) %>% view()

Sage-Bionetworks / synapser