Simon P. Couch
Simon P. Couch
I'm not able to reproduce at the moment, though I'm using : ``` r library(DBI) library(odbc) con [1] '1.2.2' packageVersion("odbc") #> [1] '1.4.2' con #> dbo@sql2 #> Database: master #>...
The changes in https://github.com/r-dbi/odbc/pull/553 went out in 1.3.5, which was released after this issue was filed. Is it possible that PR resolved this issue, @detule?
Closed in #424.☃️
Addressed in https://github.com/r-dbi/odbc/pull/656!
Related to #680, but not closed by that PR, I think.
Fixed in https://github.com/r-dbi/odbc/pull/692. :)
Thank you for the thorough issue description! Noting that I've edited the call to `read_csv()` in the OP to read directly from GH.
Ah, it appears this may be intentional. In [`?.use_case_weights_with_yardstick()`](https://tune.tidymodels.org/reference/dot-use_case_weights_with_yardstick.html): https://github.com/tidymodels/tune/blob/668eec2b747443d7b374a0fb79e38af92ed46960/R/case_weights.R#L3-L8 That is, internally, tune intentionally doesn't use importance case weights when calculating metrics. That's about as far as I can...
Mm, yes and no. tune contends that the most safe/reasonable default for importance weights is to apply them when fitting but not calculating metrics, so it doesn't allow overriding when...
This is an interesting idea! We'll come back to this after rstudio::conf. :)