How to create eye-candy graphs with python

Posted by MARTIEN LUBBERINK - Nov 06, 2019
0
236

Having attended many presentations over the years, I started noticing that tables are such a drag. Unfortunately, we are used to a presentation format that makes us want to show tables.

Why?
Quite often presenters only reveal the parts of a table that support their story. So, when was the last time you saw a row with low R-squares? And the coefficients? Right, these often join stars.
Presenters often rely on animated circles and boxes to highlight the important parts of a table. Doesn’t this demonstrate that tabular information is not at all clear and obvious? Right.
So why present tables in the first place? Tables disrupt the flow of the presentation, hardly prompt questions, occupy lots of space, and crowd out time.
Skilled presenters move tables to the end of a presentation, they expect to skip discussing them altogether.
Having also attended many presentations by investments banks, I realized that there are ways to improve your presentation.
Use graphs!
The problem is that statistical packages offer limited graph support. Also Excel has a limited choice of graphs. However, programming languages such as Python and R are almost unrestricted.
The link below offer Python code that converts New Zealand bank data into nice graphs, see above for the end-result. The libraries that I use are matplotlib (https://matplotlib.org) and seaborn (https://seaborn.pydata.org). The former is feature rich, a bit of overkill. Seaborn is great for statistics.

RBNZ_Dashboard – and nice graphs

See my github link to the Reserve Bank data: https://github.com/blucap/RBNZ_Dashboard
WordPress Cookie Plugin by Real Cookie Banner