A brief foray into two less-often-inspected SKLearn scenarios: custom features within a Pipeline and how to Pickle that Pipeline. If you’ve used and read up on SKlearn, there’s a good chance you’ve seen Pipelines mentioned as a powerful tool for ensuring streamlined workflow, avoiding data leakage, and creating a process flow suitable for serialization using Python’s pickle module. There are some incredible comprehensive guides already out there that give…

When designing a single or multiple linear regression model, a number of assumptions that support the integrity of that model. …

Recently, I was faced with a collection of several different, related .csv (comma separated values) files, which were zipped as .gz files. Though they hadn’t been pulled from a relational database, they shared several common columns that I wished to relate, then aggregate using SQL queries. After a little searching…

Dec 7, 2017: MongoDB setup and first impressions continued: I waded through a few MongoDB setup runthroughs, but found these to be most useful: https://www.codeproject.com/Articles/828392/Basics-of-MongoDB and https://docs.mongodb.com/getting-started/shell/installation/. Those sites have some good resources to what MongoDB is all about and why it’s relevant/useful compared to its peers. Setup on MacOS…

December 6, 2017. As of today, I start with: a 10+ year old comp sci major (plus 2 more majors); a mathematics masters thesis researched in Matlab, 7+ years of actuarial science work including SPSS, SAS, SQL, and Access/Excel, and about 10 months of refreshing coding skills by learning Python and doing CodeWars (https://www.codewars.com/users/mechafish)(and teaching a comp sci class at high school level).