Skip to contents

sasquatch (development version)

rOpenSci Review Changes

  • configure_saspy() no longer asks that users type in their SAS ODA username and password within the terminal. See secrets vignette.
    From thisisnic:

    I did feel a little uncomfortable entering in my SAS online password into the command line - I would have preferred to be given information about the config file and where it’s stored to enter it there myself as that seems more typical of how other packages handle this kind of thing.

    The password remains in the R terminal unobscured after being entered, which isn’t great security practice.

  • Dataset used to show sas_to_r() in README.md changed to warpbreaks and the variable name of the version of "sashelp.cars" was renamed to sas_cars. From thisisnic:

    I tried another example from the README

    sas_from_r(mtcars, “mtcars”) Warning message: In sas_from_r(mtcars, “mtcars”) : x rownames will not be transferred as a column.

    It took me a moment to understand that x refers to the source table - this could be rephrased to be clearer, or a different example used which doesn’t result in a warning message. It’d also be handy to have it explained explicitly that this creates it as “mtcars” in the current SAS session - this wasn’t clear to me and I had to check the docs to find this out.

    cars <- sas_to_r(“cars”, libref = “sashelp”)

    As there is already a dataset in R called cars I might be tempted to call this variable something else.

  • Fixed typo in configure_saspy(). Changed Europe 2 to Europe 1.
    From thisisnic:

    When I went to configure it, the setup script asked “Which server is your account on?” with the following options:

    1. United States 1
    2. United States 2
    3. Europe 2
    4. Asia Pacific 1
    5. Asia Pacific 2

    Mine is “Europe 1” but I was unable to select this. This didn’t prevent it from running in the end though.

  • Specified rlang imports.
    From thisisnic:

    Running devtools::check()

    76f checking package subdirectories … NOTE
    Problems with news in ‘NEWS.md’:
    No news entries found.

    76f checking R code for possible problems … NOTE
    .rlang_check_is_string: no visible binding for global variable ‘na_chr’
    .rlang_stop_unexpected_typeof: no visible global function definition
    for ‘caller_env’
    check_logical: no visible global function definition for ‘caller_arg’
    check_logical: no visible global function definition for ‘caller_env’
    stop_input_type: no visible global function definition for ‘caller_arg’
    stop_input_type: no visible global function definition for ‘caller_env’
    Undefined global functions or variables:
    caller_arg caller_env na_chr

    Looking at the second of those, it looks like the notes were generated by the signature for check_logical()

    https://github.com/ryanzomorrodi/sasquatch/blob/18cd7ca03a107be78700fdd3febf6908f28563c6/R/import-standalone-types-check.R#L508-L509

    I might be tempted to write my own wrapper function or ask rlang to export their check functions, but this is just personal preference.

  • Added a Code of Conduct.
    From thisisnic:

    Contribution guidelines

    These guidelines exist, but the code of conduct that they link to is missing.