Filip Schouwenaars
Filip Schouwenaars
Currently, it's based on string checking, but, as @sumedh10 rightfully pointed out, it can just inspect the `search()` path and see if `"package:"` is in there. Would make it more...
e.g. linear models can not be compared robustly. A way to work around this is to use `check_expr()` in combination with `check_equal()`. Explain this in the docs.
Currently, in `if` tests, `for` loops and function definitions, it is not possible to rerun the piece of code that is zoomed in on like it is possible in `pythonwhat`...
Might be worth looking into replacing the use of `all.equal()` in `testwhat` with the `compare` package. https://cran.r-project.org/web/packages/compare cc @richierocks
Still TBD what is bad and what needs to improve, but it's 100% sure that _something_ can be better. I checked the following out briefly (on the suggestion of @richierocks),...
The 'trust' in these examples is low, as they are not being actually run. Running the examples takes some work as well. Investigate whether we can actually run the code...
In this [exercise](https://campus.datacamp.com/courses/data-visualization-with-ggplot2-part-3/plots-for-specific-data-types-part-2?ex=5), Using the solution is obviously accepted, but the following fails: ```R germany There is something wrong in the following function call rgdal::readOGR(dsn = “shapes”, layer = “DEU_adm1”):...
```r pec % check_function('cor') ``` Fails, but shouldn't, because it's valid magrittr code! Pretty big bug, can affect a lot of courses.
The old markdown testing functions are used in only two courses: - Reporting with RMarkdown ([GitHub](https://github.com/datacamp/courses-rmarkdown), [Teach](https://www.datacamp.com/teach/repositories/428), [Campus](https://www.datacamp.com/courses/reporting-with-r-markdown)) - Communicating with Data in the TidyVerse ([GitHub](datacamp/courses-communicating-tidyverse), Teach, Campus) In short:...