AutoViz: Automated Exploratory Data Visualization

Overview

AutoViz provides a transparent, rule-based workflow for exploratory analysis. It profiles variables, quantifies missingness, screens numeric outliers, and creates visualization recommendations.

Basic workflow

av <- autoviz(iris)
av
#> <AutoViz>
#> Rows: 150  Columns: 5 
#> Recommended visualizations: 7 
#> 
#>  priority    plot            x            y                 reason
#>         1     bar      Species         <NA> Categorical frequency.
#>         2 scatter Sepal.Length  Sepal.Width  Numeric relationship.
#>         2 scatter Sepal.Length Petal.Length  Numeric relationship.
#>         2 scatter Sepal.Length  Petal.Width  Numeric relationship.
#>         2 scatter  Sepal.Width Petal.Length  Numeric relationship.
#>         2 scatter  Sepal.Width  Petal.Width  Numeric relationship.
#>         2 scatter Petal.Length  Petal.Width  Numeric relationship.

Profiling

vb_profile(iris)
#>                  variable   class        type n_unique missing missing_pct
#> Sepal.Length Sepal.Length numeric     numeric       35       0           0
#> Sepal.Width   Sepal.Width numeric     numeric       23       0           0
#> Petal.Length Petal.Length numeric     numeric       43       0           0
#> Petal.Width   Petal.Width numeric     numeric       22       0           0
#> Species           Species  factor categorical        3       0           0

Recommendations

vb_recommend(iris)
#>   priority    plot            x            y                 reason
#> 1        1     bar      Species         <NA> Categorical frequency.
#> 2        2 scatter Sepal.Length  Sepal.Width  Numeric relationship.
#> 3        2 scatter Sepal.Length Petal.Length  Numeric relationship.
#> 4        2 scatter Sepal.Length  Petal.Width  Numeric relationship.
#> 5        2 scatter  Sepal.Width Petal.Length  Numeric relationship.
#> 6        2 scatter  Sepal.Width  Petal.Width  Numeric relationship.
#> 7        2 scatter Petal.Length  Petal.Width  Numeric relationship.

A plot

vb_plot(iris, "Species", type = "bar")

HTML dashboard

dash <- vb_dashboard(iris, title = "Iris Dashboard")
dash

Iris Dashboard

Automated exploratory data analysis generated by AutoViz 1.0.0.

150

Rows

5

Columns

0

Missing values

7

Recommendations

Visualization recommendations

priority plot x y reason
1 bar Species NA Categorical frequency.
2 scatter Sepal.Length Sepal.Width Numeric relationship.
2 scatter Sepal.Length Petal.Length Numeric relationship.
2 scatter Sepal.Length Petal.Width Numeric relationship.
2 scatter Sepal.Width Petal.Length Numeric relationship.
2 scatter Sepal.Width Petal.Width Numeric relationship.
2 scatter Petal.Length Petal.Width Numeric relationship.

Dataset profile

variable class type n_unique missing missing_pct
Sepal.Length numeric numeric 35 0 0
Sepal.Width numeric numeric 23 0 0
Petal.Length numeric numeric 43 0 0
Petal.Width numeric numeric 22 0 0
Species factor categorical 3 0 0

Missing values

No missing values detected.

The dashboard layer is independent of Shiny. Shiny can be used when a full application with user-defined inputs and reactive server logic is required.