SF Data Weekly - A New Lipstick Index and Analysis and Visualization of a Debate

The Lipstick Index was a term coined by the CEO of Estée Lauder, who noticed that lipstick sales spiked during times of economic downturn since it was an affordable luxury for women. This piece uses statistical analysis and data science to find a replacement for the Lipstick Index since you can’t see lipstick behind a mask.

Amazon Elasticsearch Service is a managed cloud deployment of the Elasticsearch search engine. This piece describes how to use ingest pipelines to normalize incoming data and to create indexes in a predefined fromat.

This piece shows how to do event correlation and enrichment with Amazon Kinesis Data Analytics, and it includes a specific implementation of the generic serverless unified streaming architecture that can be deployed in your own AWS accouunt.

The issue of which of the tables in your data warehouse contain vetted, current data, and which have outdated or incorrect data, is as old as data warehouses themselves. This piece describes a Key Assets architecture as way to address this age-old problem.

Five solid tips to help you get started with Google’s BigQuery analytical database.

This piece focuses on textual analysis of the first US Presidential Debate to try to tease some meaning from 90 minutes of fast-paced, chaotic argument.

Another approach to analyzing the Presidential Debate uses analysis of Twitter data to visualize how different users of that social media platform clustered together in their responses to what the two candidates said during the debate.

The author of the R package correally shows some examples of how the combination of his library, R and Plotly can produce beautiful correlation graphs.

Github engineer Hamel Husain explains how to use the open source Great Expectations project along with GitHub Actions to automatically test, document and profile your data pipelines as part of a traditional Continuous Integration workflow.