Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/norcalbiostat/math_350_tutorials


https://github.com/norcalbiostat/math_350_tutorials

Last synced: 23 days ago
JSON representation

Awesome Lists containing this project

README 

        
# 350_tutorials



Each person modifies their own branch and submits pull requests (PR) into the `main` branch. **Robin will handle all PR's**



## Using github



* make sure you are on the `main` branch

* do a `git pull` to make sure your local files are up to date with the main branch

* Then start a new branch. `git checkout -b ` to start a new branch

* do the things, commit often, and push up to your branch on github often also

    * `git add -A`

    * `git commit -m "message"`

    * `git push`

* When task complete, open a pull request from your branch to `main`. 

* After successful merge of your item back into the `main` branch delete your branch

     * Switch back to main branch type `git checkout main`  

     * Delete a local branch type `git branch -d `



### How to use tutorial

1. install package(s)

2. click on start tutorial

3. do stuff. 

4. Click stop button to stop tutorial. Progress will be saved. 

    - A timer is running at top of console. I wonder if we can time it out?

    - Doesn't prevent you from using the console 

    - or creating/knitting a new rmarkdown file

    

### Recording events

https://pkgs.rstudio.com/learnr/articles/publishing.html#events



## References

* https://pkgs.rstudio.com/learnr/articles/publishing.html

* https://community.rstudio.com/t/how-to-get-user-id-in-the-website-created-by-learnr-and-published-in-rstudio-connect/44067

* https://education.rstudio.com/blog/2020/09/delivering-learnr-tutorials-in-a-package/

* Info from the R Community on using a package to deploy tutorials, and how to get student data out.

https://community.rstudio.com/t/does-updating-learnr-tutorials-installed-via-package-overwrite-user-progress/124804/4 (

* Moving a class from being human graded to computer-graded http://www.citizen-statistician.org/2020/08/data-science-tutorials-with-learnr-and-gradethis/



## R/Exams http://www.r-exams.org/ 



* Each question is a separate .Rmd file

* no code based answers



----



# Meeting Notes



### 12/21/21

* Robins goals for this week: Revise Ch 1 and Hw 1.  

* Sent ch1 notes in slack

* Faith is messing around with chapter 1 notes sent in Slack --thinking learnr could create a nice page.

    * Might work to use checkr with learnr to allow students to check their code as they go

    * Setting up the table of contents in learnr allows students to return to whatever section they were working on Learnr stuff

    * Set progressive field to true in the yaml header to reveal sections as students work so there aren’t any spoilers (from link above)



### 12/26/21

Still working on converting notes. I’m not sure how the exercises where students use the console and are seeing how objects are stored will work in learnR -- looking into this

Might work to just insert images



### 12/29/21



* Purpose of `learnr` tutorials: Provide students opportunity to practice writing code outside of homework. 

    - hopefully collected for a grade (working on this)

    - distributed through a package

    - Examples, you try it & a little explanatory/context text. 

    - computer graded, not human graded. Looks like `gradethis` will be used here. 

  

* Types of activities: 

    * Multiple choice quiz to check knowledge. Has a correct answer.

    * Writing correct code (`code based grading`)

        - partially filled out code chunk with [exercise.blanks](https://rstudio.github.io/learnr/articles/exercises.html#checking-blanks)

        - fully blank code chunk

    * Check if the results of a code chunk is equal to or within a numeric tolerance of an answer. 

        - submit answer to check against a correct answer (`result grading`)



Next steps: 



* Robin figure out how to create and distribute tutorials via package :heavy_check_mark:

    - `devtools::install_github("norcalbiostat/ChicoLearnR")`

    - Package will be installed in the users library like this; `C:\Users\Robin\Documents\R\win-library\4.1\ChicoLearnR` 

        - We can access this path programmatically from the .REnviron if need be. 

* Faith: clean up ch1 tutorial for appearance & minor verbiage

* Aldo: add to `ch1notes_learnrtrial1.Rmd` a multiple choice quiz

* both: figure out how to check the results of a code chunk against an answer. 

    - is the code right

    - is the number right

* @ffatchen & @aprez1999 install the package, and go through the tutorial as if you were a student to generate user data. 



Next next steps: 



* Record events. Export to..? How to send to instructor?

    - Idea: Shiny engine allows you to knit/render/download files. Will go to their 'downloads' folder. 



### 12/30/21

Aldo Updates: Some notes. 

* `checkr` package is a nightmare. Last time it was updated with 2018.

	- Issues: Calling function arguments with this package is a pain in the ass

	-The package conflicts too much with 'rlang' and 'dplyr' packages

		-some of the functions that include ‘_arg’ (which are a lot)

		- Error given: ‘Error : Can't take the names() of the .data pronoun’

* Spent a while trying to figure out trying to get the checkr package up to date by modifying some of its files and using those as an updated version of the package but it’s just outdated

* Not sure why I didn’t look into the `gradethis` package a longggg time ago but that’s definitely the one we should go with regards to using it using to check student’s code as they move through the exercises

	-The package is continuously updated and much easier to use

* Progress so far: Basically wasted a lot of time editing ‘checkr.’

	-but I learned a lot about github. Also, a lot of the same concepts carry over to the `gradethis` package  

	-Was able to create some example quizzes including those that include code-chunks  

	-Also included some hints and was able automate some questions and the checking progress by reusing the same checking function for similar questions.  

	-I’m still waiting on cleaning it since I have a ton of junk and test files up so I haven’t pushed it to my folder yet  

* Also learned how to create exams that students can complete with an autograder without access to source code



#### Robin's updates ####

Refining our goal



1. Create a learnr tutorial with a few different types of questions. We're already on a good path for ch1 here.

2. use `gradethis` for feedback & answer checking

3. use `learrhash` to generate a hash code of the student responses -- Faith

4. embed a google form into the learnR tutorial where a student can submit their hash -- Aldo



@Faith Fatchen - since you have a copy of the ch1 tutorial i'll let you finish that up.

@Aldo Perez It looks like you have example quizzes that use gradethis for answer checking.



Who wants to work on adding a "generate hash" token using learnrhash and who wants to work on embedding a google form into a tutorial?



I will look at what Faith has created & the examples Aldo added and combine into a single file, then add it to my tutorial package and make sure it works. Then I'll add in what you both generate from learnrhash + google forms and then redeploy my package.



Then we can do some end user testing.



### 1/2/22



Course packet completely updated. What if we put all the coding examples as fill in the blank examples and you try it's as tutorials? 



* Faith get hash - send to robin via slack

* Aldo get google forms - send to robin via slack

* robin 

    - get updated ch1 tutorial back into this repo, clean up folders

    - add new ch1 & 2.1 finished course notes for new material

    - Incorporate new learnrhash and google form code into ch1 tutorial

* aldo & faith pull new copy of repo - create branches

* Faith update content/sections of ch1 - then start adding gradethis

* Aldo extract examples and you try it for ch 2.1 - add gradethis

* Robin will setup google form submissions for each ch1 & ch2



--- Useful functions



```{r}

compare.tables(

  mock_this_exercise(

    .user_code = {nheads},

    .solution_code = {nheads.soln}

  )

)

```