Distributions should always be the first aspect to be inspected in your data. Boxplots draft the key figures in the distribution and help you spot outliers. Just use the boxplot method on your DataFrame for a quick overview:
In: boxplots = iris_df.boxplot(return_type='axes')
Here are the boxplots of all the numeric variables of the dataset:
If you already have groups in your data (from categorical variables, or derived from unsupervised learning), just point out the variable you need data to be represented in the boxplot and specify that you need to have it separated by the groups (use the by parameter ...