1. You can review the data here https://www.kaggle.com/c/sliced-s01e05-WXx7h8/data
    
    The data consist of: 
    train.csv - the training set: training is understood in a broader content for developing and evaluating your model. This means that you still need to split this data into training and testing subset. 
    
    test.csv - the test set: this data is for your submission. Once you prepare your model, you need to apply this to this data which is "new" for your submission 
    
    sample_submission.csv - a sample submission file in the correct format 
    
    **Focus**: data wrangling, data visualization, data pre-processing 
    
    1. Create a .rmd document with the title "Predicting Airbnb prices". Keep the output as html document and use your name and student id for the author. 
    
    2. Load the required libraries, including `tidyverse`, `tidymodels`, `skimr` 
        
    Make sure that all your R code for all the coding steps is included as R chunks 
    ```{R} ``` 
    
    3. Read the dataset with `read_csv()` from "https://raw.githubusercontent.com/maria-pro/bco6008/main/train_airbnb.csv" 
    
    Name the variable `data` 
    
    4. Review the datatypes in the dataset using `skim()` function 
    
    
    The task is to predict prices for rentals.
    What are the variables that you think may be useful?
    What are the variables which will use useless and you can get rid off? 
    
    5. Split the dataset for ML with 75% for training and 25% for testing using `initial_split()`. You may name the training data set a `train` and testing as `test` 
    
    Remember: for the next steps you are going to use the `train` dataset to do initial exploration of the data. 
    
    6. Review the variables that were "useful"? Use `tidyverse` functions to explore them. 
    
    Use https://ggplot2.tidyverse.org/ to visualize your exploration. 
    
    **Numerical variables**: 
    
    Do they vary or constant? 
    
    Are they any outliers? 
    
    Missing values? 
    
    `geom_boxplot()` and `geom_histogram()` are particularly useful 
    
    See 
    https://ggplot2.tidyverse.org/reference/geom_boxplot.html and https://ggplot2.tidyverse.org/reference/geom_histogram.html 
    
    **Categorical variables**: 
    
    Do they have enough categories (or just one)? 
    
    Are all categories equal (e.g. have the same number of data points - `count()`)? 
    
    Should some categories be "merged" into *Other* (e.g. `fct_lump`)? 
    
    Missing values? 
    
    Remember to use `fct_reorder()` to make your visualizations more appealing. 
    
    7 . Run following examples of data exploration and provide exploration of another 1 numerical and 1 categorical variable that you think may be useful for the task 
    
    7.1 Use `distinct()` to see how many unique locations the dataset covers (`name`) 
    
    Use `count()` to see how many occurances for each unique location (`name`) the dataset has 
    
    Use `count()` to see how many occurances for each unique `neighbourhood_group` the dataset has 
    
    Use `count()` to see how many occurances for each unique `neighbourhood` the dataset has 
    
    7.2 Use `ggplot()` with `geom_histogram()` to see the flactuation of the `price` - you will need to build a model that will predict it for your final submission in Assigment 2. 
    
    7.3 Group the observations in the dataset on the `neighbourhood_group` with `group_by()` and use `summarise()` to calculate the average price `avr_price` variable as `mean(price)` and median price `median_price` variable as `median(price)` and number of observations for each group `n` as`n=n()`. 
    
    7.4 Use `arrange()` with desc(n) to see the results in the descrending order. 
    
    7.5 Develop a graph that shows distribution of the `price` variable across `neighbourhood_group`. 
    
    Use `mutate` and `fct_reorder` to order the graph by `price` `mutate(neighbourhood_group=fct_reorder(neighbourhood_group, price)` 
    
    for the `aes` use `exp(price)` and `neighbourhood_group` 
    
    7.6 Use `geom_boxplot()` and `scale_x_log10` 
    
    Write 1-2 sentences why we use exp() and why we use `scale_x_log10` in this case. 
    
    Write a sentence on what insights you can get from the graph 
    
    7.7 Use the above step to show distribution of the `price` variable across `neighbourhood`. 
    
    Since we have MANY unique `neighbourhood`s we need to consider the top 40. 
    
    Use `mutate` and `fct_lump` `mutate(neighbourhood=fct_lump(neighbourhood, 40), neighbourhood=fct_reorder(neighbourhood, price))`
    Keep the rest of the previous step's code. 
    
    Write a sentence on what insights you can get from the graph 
    
    7.8 Use the above step to develop a graph that shows distribution of `price` according to `room_type` 
    
    Write a sentence on what insights you can get from the graph? 
    
    7.9 Use `ggplot()` to develop a `geom_point()` graph that shows `reviews_per_month` as `x` and `price` as `y` 
    
    Use `scale_x_log10()` and add `geom_smooth(method="lm")`. 
    
    Write a sentence what value `geom_smooth(method="lm")` adds. What insights can you get from the graph? 
    
    7.10 Use the above step to develop a graph that shows the relationship between `calculated_host_listings_count` and `price` 
    
    Write a sentence on what insights you can get from the graph. 
    
    7.11 Use the above step to develop a graph that shows the relationship between `availability_365` and `price` 
    
    Write a sentence on what insights you can get from the graph. 

Answer rating: 100% (QA)

I can help you outline the rmd document based on the instructions ... View the full answer