Tutorials

Notebooks

The vizdataquality package includes five types of notebook.

Calculations

• Calculate a set of data quality attributes and output them to a file

Visualizations

There is a separate notebook for each type of visualization. Some of them first do the above calculations and then visualize the output. Others create visualizations directly from a dataframe containing the dataset.

• Visualizing missing values with a bar chart

• Visualizing numerical and datetime distributions with a box plot

• Visualizing datetime distributions at multiple levels of detail

• Visualizing value lengths (the number of characters in values) with a dot & whisker chart

• Visualizing value counts (the number of times each value occurs in a variable) with:

  • lollipop chart

  • stacked bar chart

  • line chart, showing any gaps in a sequence of values

• Displaying attributes such as data types in:

  • table

  • text

• Visualizing numerical distributions with:

  • boxplot

  • histogram

  • violin chart

Report

• Create a report while you investigate data quality and profile a dataset

Workflow

• Apply a six-step workflow to investigate data quality and profile an open parking fines dataset

Missing data structures

• Investigating a simple monotone pattern of missing values

• Investigating interwoven patterns of missing values

• For large datasets, reading data in chunks from a file

• For saving your analysis, importing and exporting