3patternsForDataAnalysis.qmd

---
title: "R_Basic_Patterns"
format: html
editor: visual
---

## 3 Common Patterns for Data Analysis Workflows

Here are 3 basic patterns using `dplyr` functions that you can use for practice. These patterns combine different verbs into common data analysis workflows.

For these examples, we will use the built-in `mtcars` dataset, which contains information about various car models. You can load it and the `dplyr` library with the following code:

```{r}
library(dplyr)
library(tibble)

data(mtcars)

summary(mtcars)
```

## **Pattern 1: Basic summary and filtering**

This is a fundamental pattern for exploring and summarizing a dataset. It involves:

1.  **Filtering** rows based on a condition.

2.  **Selecting** a few key columns.

3.  **Grouping** the data.

4.  **Summarizing** the groups.

**The task:** Find the average miles per gallon (MPG) and horsepower for 4 and 6-cylinder cars.

**The pattern:**

```{r}
mtcars %>%
  filter(cyl %in% c(4, 6)) %>%
  select(mpg, hp, cyl) %>%
  group_by(cyl) %>%
  summarise(
    avg_mpg = mean(mpg),
    avg_hp = mean(hp)
  )

summary(mtcars)
str(mtcars)
```

**What is happening?**

-   `filter(cyl %in% c(4, 6))`: Filters the dataset to include only rows where the `cyl` (number of cylinders) is either 4 or 6.

-   `select(mpg, hp, cyl)`: Keeps only the columns for MPG, horsepower, and cylinders. This is a good practice to avoid carrying unnecessary columns.

-   `group_by(cyl)`: Prepares the data for aggregation by `cyl`, so the next operation will be performed separately for the 4-cylinder and 6-cylinder groups.

-   `summarize(...)`: Calculates the `avg_mpg` and `avg_hp` for each group, collapsing the data into a single row per group.

## Pattern 2: Creating new variables and sorting

This pattern demonstrates how to transform data by creating new columns and then ordering the results. It involves:

1.  **Creating** a new column with `mutate()`.

2.  **Filtering** for a specific condition on the new column.

3.  **Arranging** the results.

4.  **Renaming** a column for better clarity.

**The task:** Create a new column for "weight per horsepower," and find the top 5 cars with the lowest ratio for 8-cylinder engines.

**The pattern:**

```{r}
mtcars <- tibble::rownames_to_column(mtcars, var = "car_model")

mtcars %>%
  mutate(weight_per_hp = wt / hp) %>%
  filter(cyl == 8) %>%
  arrange(weight_per_hp) %>%
  select(car_model, weight_per_hp, hp) %>%
  head(5)

```

**What is happening?**

The first function is to create a column out of the names of the type of car model.

-   First we used the assignment operator to call the tibble library which was loaded at the top of our page. Then we used the rownames_to_column(mtcars, var = "car_model") function to convert the row of names into a column and variable we called 'car_model' from the mtcars data set.

-   `mutate(weight_per_hp = wt / hp)`: Creates the new `weight_per_hp` variable.

-   `filter(cyl == 8)`: Focuses the analysis on 8-cylinder cars.

-   `arrange(weight_per_hp)`: Sorts the rows in ascending order based on the new variable.

-   `select(...)`: Selects the new variable and other relevant columns to present in the result.

-   `head(5)`: Displays only the top 5 rows of the sorted result.

**Pattern 3: Summarizing with unique values**

This is a common workflow for checking the unique combinations of categories in a dataset and then summarizing a numerical variable within those unique combinations. It involves:

1.  **Grouping** by multiple variables.

2.  **Summarizing** to get a count of unique combinations.

3.  **Grouping** again, this time for a deeper summary.

4.  **Summarizing** with another statistical function.

5.  **Arranging** to display the result clearly.

**The task:** Find the number of car models for each unique combination of cylinders and gear, and then calculate the average MPG for each of those groups.

**The pattern:**

```{r}
mtcars %>% 
  distinct(cyl, gear) %>% # to find the unique combinations of cylinders and gears
  arrange(cyl, gear) %>%
  print()

mtcars %>%
  group_by(cyl, gear) %>%
  summarize(
    model_count = n(),
    avg_mpg = mean(mpg),
    .groups = "drop"  # Drop grouping for subsequent operations
  ) %>%
  arrange(cyl, gear)
```

**What is happening?**

-   `distinct(cyl, gear)`: The first block of code finds and prints all unique combinations of `cyl` and `gear` in the data, which can be a good way to understand the structure of your categorical variables.

-   `group_by(cyl, gear)`: In the second block, it groups the data by both `cyl` and `gear` for the deeper summary.

-   `summarize(...)`:

    -   `model_count = n()`: Uses the special `n()` function to count the number of rows (i.e., car models) within each `cyl` and `gear` group.

    -   `avg_mpg = mean(mpg)`: Calculates the average MPG for each group.

-   `arrange(cyl, gear)`: Sorts the final table by `cyl` and `gear` for easy readability.