Introduction

Things we want to look at

• Examine the type of businesses in each city. Do different cities have markedly different proportions of businesses or are they similar?

Using Pandas for Quick Visualizations

There are many options in the Python ecosystem for visualization. When first exploring the dataset, I find it useful to use the visualization API that Pandas uses. These lack a certain amount of flexibility and presentability but it allows us to quickly examine the data.

Other Options for Visualization

In the next section, we will consider Tableau for our visualization needs. My favorite options using Python include

1. Bokeh
   • The visualizations are somewhat ugly but it makes use of the Grammar of Graphics and we can use it with large datasets.
2. Vincent
   • Good for making simple visualizations that are both presentable and easy to use. Uses D3 as a backend.
3. GGplot.
   • A Python port exists but I have always had difficulties with it. Ggplot is a great library and it is often worth the extra effort of loading the data in R to use it.

Getting Started