---
title: "explore"
author: "Roland Krasser"
date: "`r Sys.Date()`"
output: rmarkdown::html_vignette
vignette: >
%\VignetteIndexEntry{explore}
%\VignetteEngine{knitr::rmarkdown}
%\VignetteEncoding{UTF-8}
---
```{r setup, include = FALSE}
knitr::opts_chunk$set(
collapse = TRUE,
comment = "#>"
)
```
The explore package simplifies Exploratory Data Analysis (EDA). Get faster insights with less code!
explore package on Github: [https://github.com/rolkra/explore](https://github.com/rolkra/explore)
As the explore-functions fits well into the tidyverse, we load the dplyr-package as well.
```{r message=FALSE, warning=FALSE}
library(dplyr)
library(explore)
```
### Interactive data exploration
Explore your data set (in this case the iris data set) in one line of code:
```{R eval=FALSE, echo=TRUE}
explore(iris)
```
A shiny app is launched, you can inspect individual variable, explore their relation to a target (binary / categorical / numerical), grow a decision tree or create a fully automated report of all variables with a few "mouse clicks".
{width=600px}
You can choose each variable containing as a target, that is binary (0/1, FALSE/TRUE or "no"/"yes"), categorical or numeric.
### Report variables
Create a rich HTML report of all variables with one line of code:
```{R eval=FALSE, echo=TRUE}
# report of all variables
iris %>% report(output_file = "report.html", output_dir = tempdir())
```
{width=600px}
Or you can simply add a target and create the report. In this case we use a binary target, but a categorical or numerical target would work as well.
```{R eval=FALSE, echo=TRUE}
# report of all variables and their relationship with a binary target
iris$is_versicolor <- ifelse(iris$Species == "versicolor", 1, 0)
iris %>%
report(output_file = "report.html",
output_dir = tempdir(),
target = is_versicolor)
```
If you use a binary target, the parameter ***split = FALSE*** (or `targetpct = TRUE`) will give you a different view on the data.
{width=600px}
### Grow a decision tree
Grow a decision tree with one line of code:
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=4}
iris %>% explain_tree(target = Species)
```
You can grow a decision tree with a binary target too.
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=4}
iris$is_versicolor <- ifelse(iris$Species == "versicolor", 1, 0)
iris %>% select(-Species) %>% explain_tree(target = is_versicolor)
```
Or using a numerical target. The syntax stays the same.
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=4}
iris %>% explain_tree(target = Sepal.Length)
```
You can control the growth of the tree using the parameters `maxdepth`, `minsplit` and `cp`.
To create other types of models use `explain_forest()`, `explain_xgboost()` and `explain_logreg()`.
### Explore dataset
Explore your table with one line of code to see which type of variables it contains.
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
iris %>% explore_tbl()
```
You can also use `describe_tbl()` if you just need the main facts without visualization.
```{r message=FALSE, warning=FALSE}
iris %>% describe_tbl()
```
### Explore variables
Explore a variable with one line of code. You don't have to care if a variable is numerical or categorical.
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
iris %>% explore(Species)
```
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
iris %>% explore(Sepal.Length)
```
### Explore variables with a target
Explore a variable and its relationship with a binary target with one line of code. You don't have to care if a variable is numerical or categorical.
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
iris %>% explore(Sepal.Length, target = is_versicolor)
```
Using split = FALSE will change the plot to %target:
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
iris %>% explore(Sepal.Length, target = is_versicolor, split = FALSE)
```
The target can have more than two levels:
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
iris %>% explore(Sepal.Length, target = Species)
```
Or the target can even be numeric:
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
iris %>% explore(Sepal.Length, target = Petal.Length)
```
### Explore multiple variables
```{r message=FALSE, warning=FALSE, fig.width=8, fig.height=2.5}
iris %>%
select(Sepal.Length, Sepal.Width) %>%
explore_all()
```
```{r message=FALSE, warning=FALSE, fig.width=8, fig.height=2.5}
iris %>%
select(Sepal.Length, Sepal.Width, is_versicolor) %>%
explore_all(target = is_versicolor)
```
```{r message=FALSE, warning=FALSE, fig.width=8, fig.height=2.5}
iris %>%
select(Sepal.Length, Sepal.Width, is_versicolor) %>%
explore_all(target = is_versicolor, split = FALSE)
```
```{r message=FALSE, warning=FALSE, fig.width=8, fig.height=2.5}
iris %>%
select(Sepal.Length, Sepal.Width, Species) %>%
explore_all(target = Species)
```
```{r message=FALSE, warning=FALSE, fig.width=8, fig.height=2.5}
iris %>%
select(Sepal.Length, Petal.Width, Petal.Length) %>%
explore_all(target = Petal.Length)
```
```{r message=FALSE, warning=FALSE}
data(iris)
```
To use a high number of variables with `explore_all()` in a RMarkdown-File, it is necessary to set a meaningful fig.width and fig.height in the junk. The function `total_fig_height()` helps to automatically set fig.height: ```fig.height=total_fig_height(iris)```
```{r message=FALSE, warning=FALSE, fig.width=8, fig.height=total_fig_height(iris, size=2.5)}
iris %>%
explore_all()
```
If you use a target: ```fig.height=total_fig_height(iris, var_name_target = "Species")```
```{r message=FALSE, warning=FALSE, fig.width=8, fig.height=total_fig_height(iris, var_name_target = "Species", size=2.5)}
iris %>% explore_all(target = Species)
```
You can control total_fig_height() by parameters ncols (number of columns of the plots) and size (height of 1 plot)
### Explore correlation between two variables
Explore correlation between two variables with one line of code:
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
iris %>% explore(Sepal.Length, Petal.Length)
```
You can add a target too:
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
iris %>% explore(Sepal.Length, Petal.Length, target = Species)
```
### Explore options
If you use explore to explore a variable and want to set lower and upper limits for values, you can use the `min_val` and `max_val` parameters. All values below min_val will be set to min_val. All values above max_val will be set to max_val.
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
iris %>% explore(Sepal.Length, min_val = 4.5, max_val = 7)
```
`explore` uses auto-scale by default. To deactivate it use the parameter `auto_scale = FALSE`
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
iris %>% explore(Sepal.Length, auto_scale = FALSE)
```
### Describing data
Describe your data in one line of code:
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
iris %>% describe()
```
The result is a data-frame, where each row is a variable of your data. You can use `filter` from dplyr for quick checks:
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
# show all variables that contain less than 5 unique values
iris %>% describe() %>% filter(unique < 5)
```
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
# show all variables contain NA values
iris %>% describe() %>% filter(na > 0)
```
You can use `describe` for describing variables too. You don't need to care if a variale is numerical or categorical. The output is a text.
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
# describe a numerical variable
iris %>% describe(Species)
```
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
# describe a categorical variable
iris %>% describe(Sepal.Length)
```
### Use data
Use one of the prepared datasets to explore:
* `use_data_beer()`
* `use_data_diamonds()`
* `use_data_iris()`
* `use_data_mpg()`
* `use_data_mtcars()`
* `use_data_penguins()`
* `use_data_starwars()`
* `use_data_titanic()`
* `use_data_wordle()`
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
use_data_beer() %>% describe()
```
### Create data
Use one of the prepared datasets to explore:
* `create_data_abtest()`
* `create_data_app()`
* `create_data_buy()`
* `create_data_churn()`
* `create_data_esoteric()`
* `create_data_newsletter()`
* `create_data_person()`
* `create_data_unfair()`
* `create_data_random()`
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
# create dataset and describe it
data <- create_data_app(obs = 100)
describe(data)
```
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
# create dataset and describe it
data <- create_data_random(obs = 100, vars = 5)
describe(data)
```
You can build you own random dataset by using ```create_data_empty()``` and ```add_var_random_*()``` functions:
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
# create dataset and describe it
data <- create_data_empty(obs = 1000) %>%
add_var_random_01("target") %>%
add_var_random_dbl("age", min_val = 18, max_val = 80) %>%
add_var_random_cat("gender",
cat = c("male", "female", "other"),
prob = c(0.4, 0.4, 0.2)) %>%
add_var_random_starsign() %>%
add_var_random_moon()
describe(data)
```
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
data %>% select(random_starsign, random_moon) %>% explore_all()
```
### Basic data cleaning
To clean a variable you can use `clean_var`. With one line of code you can rename a variable, replace NA-values and set a minimum and maximum for the value.
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
iris %>%
clean_var(Sepal.Length,
min_val = 4.5,
max_val = 7.0,
na = 5.8,
name = "sepal_length") %>%
describe()
```
To drop variables or observations you can use ```drop_var_*()``` and ```drop_obs_*()``` functions.
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
use_data_penguins() %>%
describe_tbl()
```
```{r message=FALSE, warning=FALSE, fig.width=6, fig.height=3}
use_data_penguins() %>%
drop_obs_with_na() %>%
describe_tbl()
```
### Create notebook
Create an RMarkdown template to explore your own data. Set output_dir (existing file may be overwritten)
```{r eval=FALSE, message=FALSE, warning=FALSE}
create_notebook_explore(
output_dir = tempdir(),
output_file = "notebook-explore.Rmd")
```
{width=600px}