modelr | Helper functions

I would seem that your package depends, at least, on raster and sf; and your script does not install them:

Add this after the YAML and then the method between ::: will work:

You have a very old version of R. You should update to the current version if you want things to go smoothly.

If you can't do that, here's what you'll have to do:

• Install the version of Rtools suitable for R 3.5.x.
• Install the packages you want. Some of them won't work, because they will need later versions of R, but won't declare that. So when you find one that fails, try installing the previous version of that package. If that also fails, try an even earlier one. Etc. Your R version was current in 2018, so you may need to go that far back in time to find compatible packages.
• Once you finally have everything working, try to update your packages. Maybe some of them could be more recent versions, maybe not. Do them one at a time. Typically tidyverse needs about 90 of them, so this will take a while.

So I recommend that you update your R version.

Need to run tools::write_PACKAGES before executing install.packages.

Even though the tar.gz files are in the same folder, the PACKAGES file must be populated to properly account for dependencies. Otherwise, the install.packages function doesn't know how the packages are related. Would be nice if that error message was more specific.

@importFrom recipes prep bake had to be added to the .R file

Errors like this in Shiny apps are almost always a result of missing packages, which the logs confirm.

Turns out in this case I think the missing package is kernlab, which I only found by reading the documentation given here: https://topepo.github.io/caret/train-models-by-tag.html#Support_Vector_Machines. It's a suggested package, not imported, so the command suggested in the comments by heds1 would sort this out.

One option would be to make use of geom_hline instead of geom_polygon:

You definitely were passing the vector of variables correctly to the recipe -- no problem there!

You were running into other problems with your model fitting. An xgboost model requires all predictors to be numeric, so if you convert something like zip code to factors, you need to then use step_dummy(). If you have something of high cardinality like zip codes, you probably will need to handle new levels or unknown levels as well.

This is tricky, and the answer below doesn't solve this 100%. Now there are a number of different ways assertr lets you handle errors/stops, just see ?error_stop (which is the default).

You need to not only filter out rows that fail, but also collect them (all) for later inspection.

Below I wrote my own error handler. It fetches those rows that fail, filter them away, and stores them in the global environment under the varibale my.failed.rows.