learnr-demo.Rmd

---
title: "Learnr Demo"
output: 
  learnr::tutorial:
    progressive: true
    allow_skip: true
    css: "css/style.css"
runtime: shiny_prerendered
---

```{r setup, include=FALSE}
# load packages
library(learnr)
library(gradethis)
library(sortable)
library(tidyverse)
library(learnrhash)
library(showtext)
library(googlesheets4)

# load data
ikea <- read_csv("data/ikea.csv")

# don't echo chunks
knitr::opts_chunk$set(echo = FALSE)

# apply theme to ggplot
ggplot2::theme_set(theme_bw())
```

## Introduction

*This is a demo tutorial for the [Designing the Data Science Classroom](https://rstudio-conf-2022.github.io/teach-ds/) workshop at rstudio::conf 2022.*

### Packages

We will use the **tidyverse** package in this tutorial. Run the code below to load the package.

```{r load-package, exercise = TRUE}
library(tidyverse)
```

### Data: Ikea furniture

The data set is all about [Ikea](https://www.ikea.com/) furniture. It was obtained from the [TidyTuesday](https://github.com/rfordatascience/tidytuesday/blob/master/data/2020/2020-11-03/readme.md) data collection.

Run the code below to see the data.

```{r load-data, exercise = TRUE}
ikea
```

We will primarily use the following variables in the analysis:

-   `category`: The furniture category that the item belongs to (Sofas, beds, chairs, Trolleys,...)
-   `price_usd`: Price in US dollars based on the price on the Ikea website on 4/20/2020.

Run the code to see the names of all variables in the data set, and [click here](https://github.com/rfordatascience/tidytuesday/blob/master/data/2020/2020-11-03/readme.md) to see the full data dictionary.

```{r colnames, exercise = TRUE}
names(ikea)
```

## Part 1: Question types

### Multiple choice questions

Below are side-by-side boxplots comparing the price of three categories of furniture at Ikea.

```{r}
ikea |> 
  filter(category %in% c("Tables & desks", "Sofas & armchairs", "Beds")) |>
ggplot(aes(x = category, y = price_usd, fill = category)) +
  geom_boxplot() + 
  guides(fill = "none") +
  labs(x = "Furniture category", 
       y = "Price (in $)") 
```

```{r variability}
question("Which furniture category has the least variability in the price?", 
         answer("Beds"), 
         answer("Sofas & armchairs"), 
         answer("Tables & desk", correct = TRUE), 
         answer("The variability is the same for all three categories")
)
```

You can add options to allow retries, random order, and hints for incorrect responses.

```{r variability-hint}
question("Which furniture category has the largest IQR?", 
         answer("Beds"), 
         answer("Sofas & armchairs", correct = TRUE), 
         answer("Tables & desk"), 
         answer("The IQR is the same for all three categories"), 
         incorrect = "The IQR is the difference between the first and third quartiles.", 
         allow_retry = TRUE, 
         random_answer_order = TRUE
)
```

### Coding exercises

What is the most common furniture category?

```{r count-category, exercise = TRUE}

```

### Coding exercises with starter code

You can use starter code to provide scaffolding for coding exercises, and there are [multiple options](https://rstudio.github.io/learnr/articles/exercises.html) to customize the exercise input box.

What is the median price of furniture in the most common category?

```{r median-price, exercise = TRUE, exercise.lines = 5}
ikea |>
  filter(____) |>
  summarise(___)
```

### Exercises with hints

You can add hints to coding exercises.

Consider the categories `Sofas & armchairs` and `Tables & desks`. Which category has the highest mean price?

```{r avg-price, exercise = TRUE, exercise.lines = 5}
ikea
```

```{r avg-price-hint-1}
ikea |> 
  filter(category %in% c(____)) |>
  ____ |>
  ____
```

```{r avg-price-hint-2}
ikea |>
  filter(category %in% c("Sofas & armchairs", "Tables & desks")) |>
  group_by(___) |>
  ___

```

```{r avg-price-hint-3}
ikea |> 
  filter(category %in% c("Sofas & armchairs", "Tables & desks")) |>
  group_by(category) |>
  summarise(mean_price = mean(price_usd))
```

### Sorting questions

Use the [**sortable**](https://rstudio.github.io/sortable/) package to create sorting and ranking questions.

```{r sorting-exercise}
# Define the answer options
baking_steps <- c("Mix the ingredients.", 
           "Bake the cake.", 
           "Decorate the cake.", 
           "Slice the cake.")

# Write the exercise 
question_rank(
  "Sort the steps for making a cake in the correct order.",
  answer(baking_steps, correct = TRUE),
  answer(rev(baking_steps), correct = FALSE, message = "Other direction!"),
  allow_retry = TRUE
)
```

## Part 2: Your turn!

::: your-turn
Create a tutorial with three questions:

-   one multiple choice
-   one coding exercise
-   one sortable question

<b>Stretch goal:</b> Add hints to the coding exercise.

You can write your questions in this section of `learnr-demo.Rmd`.
:::

## Part 3: Automated feedback

So far students have to check their solutions to coding exercises by assessing whether the output is reasonable or comparing their code to the hints. You can add automated feedback on coding exercises using the [gradethis](https://pkgs.rstudio.com/gradethis/) package.

### Checking code

Count the number of items in the data set that cost more than \$300.

```{r gt300, exercise = TRUE}

```

```{r gt300-solution}
ikea |>
  filter(price_usd > 300) |>
  nrow()
```

```{r gt300-check}
grade_this_code()
```

### Checking output

Count the number of items in the data set that cost more than \$300. Submit the response as a single number, not in a data frame.

```{r gt300-pt2, exercise = TRUE}

```

```{r gt300-pt2-solution}
ikea |>
  filter(price_usd > 300) |>
  nrow()
```

```{r gt300-pt2-check}
grade_this({
  pass_if_equal(
    ikea |> filter(price_usd > 300) |> nrow(), 
  message = "Great work! You found the number of furniture items greater that cost more than $300!")
  fail_if_equal(nrow(ikea), 
                message = "You found the number of observations in the data, but did you forget to filter for those that cost more than $300?")
  fail("Not quite but keep trying!")
})
```

## Part 4: Your turn!

::: your-turn
Add automated feedback to your coding exercise.

**Stretch goal**: Consider 2 -3 common errors students may make. Write specific feedback for those errors.

You can write your responses in this section or modify the code in Part 2 of `learnr-demo.Rmd`.
:::

## Part 5: Minute paper

Provide a "minute paper" in an embedded online form for students to reflect on their understanding of the material in the tutorial.

```{=html}
<iframe src="https://docs.google.com/forms/d/e/1FAIpQLScsV-_m3unjj1S-NSsMjBir-yxrObaLbdS3ypzTe1tPoYv8gQ/viewform?embedded=true" width="640" height="550" frameborder="0" marginheight="0" marginwidth="0">Loading…</iframe>
```
## Part 6: Recording responses with learnrhash

Use the [**learnrhash**](https://github.com/rundel/learnrhash) package to collect student responses from learnr tutorials. Students submit their responses using a "hash" submitted to an online form.

Create a histogram of the distribution of Ikea furniture prices, then answer the following questions based on the histogram.

```{r price-histogram, exercise = TRUE}
ggplot(___, aes(x = ___)) +
  ___() +
  labs(
    ___ = "Price (in $)",
    ___ = "Frequency",
    ___ = "Distribution of Ikea furniture prices"
  )
```

```{r price-histogram-hint-1}
ggplot(ikea, aes(x = ___)) +
  ___() +
  labs(
    ___ = "Price (in $)",
    ___ = "Frequency",
    ___ = "Distribution of Ikea furniture prices"
  )
```

```{r price-histogram-hint-2}
ggplot(ikea, aes(x = price_usd)) +
  ___() +
  labs(
    ___ = "Price (in $)",
    ___ = "Frequency",
    ___ = "Distribution of Ikea furniture prices"
  )
```

```{r price-histogram-hint-3}
ggplot(ikea, aes(x = price_usd)) +
  geom_histogram() +
  labs(
    ___ = "Price (in $)",
    ___ = "Frequency",
    ___ = "Distribution of Ikea furniture prices"
  )
```

```{r price-histogram-solution}
ggplot(ikea, aes(x = price_usd)) +
  geom_histogram() +
  labs(
    x = "Price (in $)",
    y = "Frequency",
    title = "Distribution of Ikea furniture prices"
  )
```

```{r price-histogram-check}
grade_this_code(
  correct = "You correctly made the plot! Note that there is a warning that we will discuss in class."
)
```

```{r mc1}
question(
  "Which of the following describes the shape of the distribution of Ikea furniture prices? Check all that apply.",
  answer("Right skewed",
    correct = TRUE
  ),
  answer("Left skewed",
    message = "Skew is on the side of the longer tail"
  ),
  answer("Symmetric",
    message = "If you were to draw a vertical line down the middle of the x-axis, would the left and right sides of the distribution look like mirror images?"
  ),
  answer("Unimodal",
    correct = TRUE
  ),
  answer("Bimodal",
    message = "How many prominent peaks do you see?"
  ),
  answer("Multimodal",
    message = "How many prominent peaks do you see?"
  ), 
  random_answer_order = TRUE
)
```

```{r mc2}
question(
  "Which of the following is false?",
  answer("The mean price is expected to be lower than the median price of Ikea furniture.", correct = TRUE),
  answer("Less than 25% of the observations have prices greater than $1000."),
  answer("The IQR is less than $1000."),
  answer("The median price is less than $1000."), 
  random_answer_order = TRUE
)
```

### Submit answers

```{r context="server"}
encoder_logic(strip_output = TRUE)
```

```{r encode, echo=FALSE}
encoder_ui(
  ui_before = div(strong("Once you're done with your quiz, click on Generate Submission below, copy the hash generated and paste it in the form below."), br(), br()),
  ui_after  = iframe_ui(
    src = "https://forms.gle/w6hUr4j7tSdvRYtr7",
    width="700px", height= "600px"
  )
)
```

### Decode hash

The following would not be included in a student facing tutorial, but has been included here to demonstrate the decoding functionality of the [**learnrhash**](https://github.com/rundel/learnrhash) package.

```{r context="server"}
decoder_logic()
```

```{r decode, echo=FALSE}
decoder_ui()
```

### View data

You can view the collected data [here](https://docs.google.com/spreadsheets/d/1aEJxhOTHzUXdngrSOrntoyK8Suptj7Z_U7-8B4k1XBU/edit?usp=sharing).

<iframe width="900" height="400" src="https://docs.google.com/spreadsheets/d/e/2PACX-1vQjut2uPmkXsRpXSbbjSjQ68oGtw596Gxmb6kS079ZFxMoKqlc_xY3l66egsAvcXLP-XzDyPKwXpdqf/pubhtml?gid=333624434&amp;single=true&amp;widget=true&amp;headers=false">

</iframe>

```{r, warning = FALSE, message = FALSE}
gs4_deauth() #deauthorize to read googlesheets

# read in responses 

submissions <- read_sheet("ttps://docs.google.com/spreadsheets/d/1aEJxhOTHzUXdngrSOrntoyK8Suptj7Z_U7-8B4k1XBU/edit?resourcekey#gid=333624434") |>
  rename(student_name = `Name`, 
         student_id = `Student ID`,
         hash = `Copy and paste the hash generated at the end of your tutorial below.`) |>
  select(-Timestamp)
```

### Process student submissions

You can process the student data in R to review student answers, calculate assignment scores, identify common

```{r}
submission_data <- extract_hash(submissions)
submission_data
```

<br> 

Example of the data in the `answer` column. 

```{r}
submission_data$answer[1]
submission_data$answer[2]
submission_data$answer[3]
```