L28_DataFiltering_Final.Rmd

---
title: "Data Filtering"
output: 
  html_document:
    toc: true
    toc_float: true
    code_folding: hide
    number_sections: true
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = T, message = F, warning = F)
```

# Packages

```{r}
library(dplyr)
library(tidyr)
library(ggplot2)
```

# Data Preparation

## Load Data

```{r}
data("diamonds")
```

```{r}
summary(diamonds)
```

## Filtering Rows

You can filter for rows with **filter()**.

You can use all comparison operators, e.g. ==, !=, <=, >=, ...

You also can check for categories with %in%.

```{r}
diamonds[2, ]  # 2nd row
diamonds[c(1,3), 'cut']  # 1st, 3rd row, column 'cut'

diamonds_filt <- diamonds %>% 
  filter (carat >= 3) %>% 
  filter (cut %in% c("Ideal", "Premium"))
```

Other row-based filters are **sample_n()**

```{r}
diamonds_filt <- diamonds %>% 
  sample_n(100)
```

or **slice()**.

```{r}
diamonds_filt <- diamonds %>% 
  slice(10:20)
```

or **top_n()**.

## Filtering Columns

You can filter for columns with **select()**.

There are two ways to approach this:

1. White-List: define the columns that you want to keep

2. Black-List: define the columns that you don't want to keep

```{r}
diamonds_filt <- diamonds %>% 
  select(carat, cut)  # white-list

diamonds_filt <- diamonds %>% 
  select(-carat, -cut)  # black-list
```